File: marge smith chart.png (205 KB, 500x385)

205 KB PNG

fresh OC, hot off the press

File: 0006294D-36A4-4F5C-AEE4-7(...).jpg (36 KB, 328x936)

36 KB JPG

The UT219DS looks really nice. Should I?

>>2945643
at a quick glance, I thought it could measure uA with the clamp lol

>>2945648
If you pay 2500 buckeroos for a fluke, you can get a leakage current clamp meter that only measures current up to 50mA or so. While that would be a really nice feature, I mainly want this for measuring inrush current. I have a cheaper clamp meter but it has no min/max mode, just a hold button. The hold button is pretty useless for bench use like I’m using it for. All the other cool features like the dual display aren’t a requirement, but I figure if I can replace the shitty clamp meter and my other meter (dual display Hioki meter) in my toolbox why not.
I’m still not entirely sure on buying Uni-T over a less Chinese brand. But Brymen don’t make many clamp meters, and Fluke and Hioki are pricy.

File: Capture.png (37 KB, 393x580)

37 KB PNG

>>2945584
has anyone ordered from LCSC in the last month or two? is the quoted poverty shipping and handling service (global direct standard mail) inclusive of tariffs -- or will i somehow be jewed into paying an additional tariff fee by USPS or whoever actually delivers it?

>Just order fully assembled board from JLC
Just ten years ago at university, we etched and hand soldered our own boards. Damn.

>>2945790
I made an order a month ago, and I wasn’t charged any extra duty, but I’m not a burger so maybe that’s not very useful.

>>2945793
i prefer etching my own boards for proof of concept. i don't need qty 5 and i don't want to wait 2 weeks to figure out if i created something that will never actually work. also etching is cool and fun.

retarded question but overhead three phase lines aren't ground referenced, right? there are three conductors and maybe an overhead line that is grounded for lighting protection, but there is no neutral line. so the voltages are all floating with respect to ground and saying e.g., "that's a 69kv line" means the 69kv is with respect to the other phases. so if it's floating, how would i be shocked to death by touching just one of the phases while being ground referenced?

>>2945801
I like breaking projects up. Instead of having one board with a lithium ion protection circuit, microcontroller, buttons, and motor driver, I’ll order 5pcs general purpose lithium protection boards and button boards. Then I’ll put the buttons and microcontroller on vero-board, or if it’s SMD I’ll do some etching. Hand-painted etch-resist because I’m autistic.

>>2945833
the "ground" is just the difference in potential relative to the phases, so pretty cooked if you were to close that potential with your body

>>2945843
just buy off-the-shelf boards. for example in an instrument I'm building (and possibly selling) I'm using an arduino clone mounted upside down rather than even bothering putting footprints for an AVR, USB etc on the main board

>>2945889
I use them where applicable, but I often use less common microcontrollers, or want different supporting components, or whatever. Like my lithium charge board, I’m using an alternative to the TP4056 that can instead charge a LiFePO4 if you set the divider up right, and it uses a different DW01 chip for different voltage thresholds, because 2.3-2.5V is too low for a Li-NMC cell.

Ahhh I WILL BECOME A MASTER of making stable, robust LC oscillators for any frequency conceivable AND NOTHING WILL STOP ME NOTHING. I will only use jellybean BJTs (and RF BJTs for higher frequencies)
Yes i also struggle with basic algebra but i don't care I don't care

I want you guys to help me. Any book, any video series, anything will help

>>2945584
if i want to measure the input impedance of a MOSFET amplifier how would i do that? something like a nanoVNA only outputs 500 mV so it wouldn't be enough to drive the MOSFET to fully on.

>>2945936
isn't it biased/terminated to give 50 Ohm?

>>2945937
i have a MOSFET pre-amplifier that feeds into a MOSFET amplifier. i want to impedance match the pre-amp output and the amp input.

>>2945937
>>2945941
put another way, i think i want to avoid resistive elements like a 50 ohm termination resistor?

Best thermal glue for heatsinks?

>>2945936
Does whether it’s getting turned on or not affect its input impedance? And for turning the MOSFET on, shouldn’t you be using a bias stage to put the MOSFET into the right mode of operation regardless of signal amplitude? Phantom power or otherwise.

>>2945953
The thermal epoxy made by Tech Ingredients. It’s probably overkill.

File: IMG_20250921_011535-2.jpg (140 KB, 1632x918)

140 KB JPG

>>2945941
post schematic
also just buy a monolithic amp IC
>>2945953
I use picrel. it works well enough for my needs

I want to do a small project with LoRa hardware and meshtastic, nothing fancy, just a proof of concept and to help teach some people about electronics and stuff. I do want to be able to send text messages back and forth between homes a few houses apart, but that's about it.

Can anyone recommend the cheapest hardware I could use to do this?

Mmfh, Infineon’s IGC019S06S1 is making me want to swap from silicon FETs to GaN. $3 each at 10x quantity, 1.9mΩ on resistance, 0.5C/W to the big thermal pad on top, rated at 99A continuous or 700A pulsed. Only 13nC total gate charge. Some 25V Si FETs come close, but at 30V and above they’re at least 29nC, let alone 60V

>>2946008
you're better off buying some kind of module either from semtech or rak wireless. then it's just 2 spi pins iirc to hook it up to any microcontroller you want.

>>2946008
Consider a LilyGo module or whatever they’re called, they’re boards with a LoRa module, ESP32, and often a screen and buttons all together. And you can easily put Meshtastic or whatever on them. But then that’s not a hardware or software project, it’s just a tech demo. So you might want to roll it back by designing a custom PCB for the same firmware, or writing custom firmware for the same hardware, or even both.

>>2946008
You don’t need meshtastic to send messages a few homes apart. Or lora, really.

Most Lora is artificially limited to ridiculously low aggregate bandwidths, it was designed to send periodic gps coordinates (in the burst).

>>2946024
Well, I wanted that sending messages back and forth to be a fun proof of concept for a few other people, and then I wanted to use that as a stepping stone to more complex meshtastic projects. But I want to start small first.

>>2946024
lora can run pretty fast if you want to, up to around 100 kbps I think. but to get range with low power you need low bandwidth. it's all down to how many joules you put into each bit

>>2946081
I think he’s referring to the transmit duty-cycle, you can only transmit like 5% of the time to reduce the chance of message collisions. Makes you wonder if there’s a (de)modulation method that can more easily discriminate overlapping signals.

>>2946087
>I think he’s referring to the transmit duty-cycle, you can only transmit like 5% of the time to reduce the chance of message collisions
that's more of a "being good neighbors" kind of thing. if you added something to measure the activity in the region then you could probably up the rate. maybe concentrators could broadcast how much traffic they're seeing
>Makes you wonder if there’s a (de)modulation method that can more easily discriminate overlapping signals.
if you know exactly how the signal was modulated, and you can align the clocks sufficiently well, then you can subtract the signal and try demodulating again
concentrators tend to have multiple receivers working with different bandwidths. something like 8 or 16 of them. there may even be a band plan for this, I'm not sure

>>2946087
oh also if you know part of the signal then you can use correlation to extract part of it from the noise. WSPR does this, since most messages are very stereotypical

>>2945584
is qspice any more user friendly than ltspice? im so fucking sick of forgetting how to use third party libraries once every 6 months i actually bother to use it, and then i have to spend an hour FUCKING AROUND with the retarded fucking filepaths and "Could not open library" errors with no indication as to what folder it's trying to look in and then erroring out.

File: 1735023765117344.jpg (384 KB, 1920x1080)

384 KB JPG

Do you use goggles when soldering?

File: NERD-GLASSES.jpg (215 KB, 589x550)

215 KB JPG

>>2946108
just these but i suspect it's not good for my eyes like the residue sticks to my eyeballs

>>2945584
which soldering station?

>>2946116
you VILL buy a pinecil V2 and you VILL be happy

>>2946108
> use goggles
Those aren’t goggle in picrel.
Also, yes, when soldering tiny stuff and my eyeball is close to what I’m doing.
Flux sometime spatters and pops, and wires come free an flick solder.
I didn’t think I needed a ventilator fan or safety glasses until I saw manual soldering station at a factory. They were using black PC fans to ventilate but they were all covered in this thin silver coating which was this ultra fine solder dust clumped onto the fan with flux residue.

File: 0CB0F426-1007-4A7E-AF50-E(...).jpg (117 KB, 983x620)

117 KB JPG

>>2946116
You can get by with just a regular soldering iron.
The temperature control is mostly bullshit if you’ve done a lot of soldering.
In a factory some stations do have hakko, but we keep the temp at around 400°C because generally we have to work fast.

>>2946098
> could not open
1. Start a little readme to yourself in ltspice-notes.txt in your home directory. Consult again in 6 months when you need it.
2. Use procmon from microsoft sysinternals to find out what files imagename ltdpice.exe is trying to open.

>>2946081
> 100 kbps
More like 50.

>>2946089
> being good neighbours kind of thin
Exactly. A lot of high level modules have firmware that limits you to something like, say, 100 characters per hour.
If you go low-level and saturate the available shared bandwidth you might annoy someone.

This is true of any radio based system. My neighbour, for example, is using an over-boosted WiFi that multiplexes over 7 channels and it’s driving me crazy.
However, I just bought 10 old wifi units and an 1—10 GHz RF amplifier and his 8k movie watching over wifi is about to degrade significantly.

File: 1700815019827.jpg (26 KB, 473x500)

26 KB JPG

>>2946143
just remove the shielding from a microwave and point it as his apartment

>>2946116
Something with a decent tip-to-grip distance. Know that a 40W temperature regulated iron is purring as much heat into a joint at 300C as a ~60W unregulated iron, because they have to dissipate heat into the air by design, and will be more compact. If you plan on soldering XT60s, I’d recommend at least 60W regulated. Regulated can mean temperature controlled, but also curie-point or PTC. If you plan on working on SMDs, I wouldn’t go above 50W if you’re not picking a regulated iron, the temperature will float pretty high and the iron will be pretty bulky.

Pick a tip ecosystem with lots of available tips of different shapes, especially D12 and D24.

Anyone here build audio stuff? I'm about to start slowly getting into synth building. Partly to investigate certain facators such as why some analog synths sound massive and others sound compressed, basically playing with output stages. Excited and hope it leads to designing something of my own

>>2946163
I’d start off looking at the waveforms and their Fourier transforms, maybe there’s something distinctive that makes a synth sound massive or compressed. Lantertronics has a series of online lectures about analog and digital synthesisers, which may be useful.

File: 1754966783603651.jpg (2.7 MB, 4446x3558)

2.7 MB JPG

>>2946118
this.
i got one and a set of tips for cheap, and the thing is insane with a proper 100w pd usb c power supply.
with the c4 tip that little iron shits on xt-60s
really kinda amazed me desu
big ass soldering stations are for boomers

File: IMG_6505.jpg (1.5 MB, 2268x4032)

1.5 MB JPG

Not sure if this belongs here or the stupid question thread. I set up an LED strip for my stairs using a microcontroller and a distance sensor (sensor’s not in the pic). I’m running into two issues. 1) The LED strip isn’t mounted. Right now it’s just kind of sitting there because I don’t want to drill into the railing. If I screw the railing down and glue the strip on, I’d have to tear the whole thing apart just to troubleshoot later, which has happened a few times. But the bigger issue is that it because stairs are slanted, the wires going to the microcontroller ended up getting pinched by the pillar which results in the entire system not working correctly. 2) I have no idea how to mount the microcontroller. It isn't shown here, but I tied the thing to the stair pillar with rope. Like it is bare circuit, no 3D printed case (lol I don't know how to use CAD yet) so it is janky, but at the same time, I also don't want to drill into the stair pillar either.

Any ideas for how I can mount both the strip and the microcontroller in a non-permanent way?

>>2946199
Still more than 40mm tip-to-grip distance, wouldn’t be me.

>>2946202
Command strip, velcro tape, whatever. Maybe put the MCU at the base of the strip directly, instead of having LEDs go all the way to the bottom.

>>2946202
Alien Tape type stuff is good. It can possibly pull the finish off of wood when you want to take it off, but if you go slowly it's usually OK. It's still better than most because it doesn't really leave any residue.

>>2946202
a more zoomed out picture would help
you might be able to get away with thinner wires

File: 3B167D60-B337-4FE0-A018-C(...).jpg (54 KB, 948x335)

54 KB JPG

>>2946206
> tip-to-grip distance
Here’s your dream soldering iron.

>>2946136
>they were all covered in this thin silver coating
Picture/video?

>>2946199
I just got a Weller 80th anniversary WE1010 all black and it looks beautiful on my bench. Take the boomerpill.

>>2946259
Yes.

Man, the 2ED300C17 gate drive module pushes 30A in and out of your IGBT gate. But it’s $200.

File: 20A445F4-C041-495F-A342-1(...).jpg (1.85 MB, 3024x4032)

1.85 MB JPG

I cannot for the life of me desolder this mouse switch. Any tips?

I used flux, added fresh solder, turned up the heat, and nothing. My wick just isn’t wicking or something. 375c with a pinecil. Considering just getting my heat gun and prying it off.

>>2946384
If you're replacing the switch, use side cutters to destroy it so you can desolder each pin individually. Otherwise, use a Soldapullt (full size), Chipquik, infrared or hot air. A vacuum desoldering station is the best way to do this.

Have you tried disassembling the housing in situ so you can clean the internal contacts?

>>2946388
> disassembling the housing in situ so you can clean the internal contacts
One of the most fascinating theories of the double click failure was the oxide builds up on contacts causes capacitance and the double click.
I’m definitely going to try and fix my old switches,
Also going to ultimately replace with optical and/or hall-effect re-using the old switch housing.
I wonder why they don’t fix the bounce in the USB driver for the mouse… or used to until I realized it’s intentional.

File: deoxit.png (201 KB, 591x489)

201 KB PNG

>>2946408
>I’m definitely going to try and fix my old switches
Do yourself a favor and get some DeoxIT. A little goes a long way. Thank me later.

>>2946384
Solder wick pulls solder off the workpiece using surface tension. You know where else solder gets sucked into via surface tension? The inside of a plated through-hole, especially if there’s a pin inside it. There are two basic modes that might work for you, breaking the solder joints one by one, or reflowing all joints at once.
For the former you could try:
>a good solder sucker, if the joint hasn’t solidified by the time you move the iron and get the sucker in place
>a hollow-tipped vacuum desoldering iron
>hollow desoldering needles, if they fit
For the latter you could try:
>hot air or infrared
>a low-melt desoldering alloy like chipquik, make your own by mixing bismuth and leaded solder
>a wide custom shaped soldering iron tip milled out of copper
>wrapping sufficiently large solid core copper wire around all the pins to transfer heat to them
>getting a soldering tweezer iron, or two more irons, to heat all three leads at once

>>2946384
Fuck, I just came here after having a miserable time trying to desolder an old op-amp on a 35 year old board. I ended up snapping it off and still have to clear a few legs out. I hate desoldering but I also just ordered a solder sucker.

File: edsyn-soldapullt-ds017.jpg (76 KB, 2000x1000)

76 KB JPG

>>2946426
Edsyn Soldapullt model DS017

File: Proco Rat.png (58 KB, 560x667)

58 KB PNG

I built a RAT guitar pedal clone on stripboard based on someone elses design just because I had the 100% of the parts laying around already. The normal schematics don't include the DPDT bypass switch with LED circuit. I built one using a jfet, it seemed like it worked, put everything in a box.
Then I catch it oscillating when I turn up the gain knob, it was even audible in bypass mode for some reason. And the LED would turn on from it.
I guess step one is to take it back out of the enclosure and start checking the ground paths by the gain section?

>>2946476
What was oscillating, the JFET you added, or the one already in the layout? What kind of frequency? Just adding a capacitor from its gate to its drain might kill the oscillation, or increasing that op-amp’s 100pF cap if it’s the op-amp that’s oscillating.

>>2946476
put a capacitor between the output and the negative input on either or both of the op amps. say 10 pF

>>2946384
Might want to consider using a desoldering gun, it makes life a lot easier. With that out of the way, though, I'd try to use a desoldering gun to get more of the solder off. Once I can't get any more that way, my usual approach for stubborn components is to use a dental pick (with VERY gentile pressure) and hot air to get the component the rest of the way out. Just be sure to cover any plastic or displays near/on the component with capton and aluminum tape.

>>2946384
you can't wick double-sided stuff. you need a desoldering gun or hot air+pliers

File: Screenshot 2025-09-24 055806.png (420 KB, 703x449)

420 KB PNG

Tourist here, my mouse started double clicking and the scroll encoder has started to fail but the rest of the mouse is fine (I know I can just solder new ones) but I was curious if its possible to design my own breakout PCB to work with the mainboard? Pic shown are the clicks which exist on a different PCB. Any idea how I should start reverse engineering the traces (or is it even possible) to have my own foreign PCB work with the original?

>>2946583
Instead of following the traces, you can likely just infer connections via the terminations. With a multimeter, you could probe for continuity between the USB socket, switch pins, encoder pins, etc. all the way back to the pin header. It’s a 10 pin header, USB will be 4, three switches will be 3, and the rotary encoder will be 2, chances are there’s just an unconnected pin on that header, or they’re using two for ground. Then for the passives. Those ferrite beads will likely be in series with the positive voltage rail from the USB, those diodes are probably TVS diodes for protecting the USB data pins and voltage rail, then there will be a bypass cap or two, probably pull-up resistors for the switches, and maybe denouncing caps on the switches too.
Ensure that when you route the USB data wires, you route them as a differential pair with a continuous ground path nearby.

>>2946588
Thanks for the details anon, what is the risk of me doing something dumb with voltage or shorting the entire mainboard during this process? Should I hire someone to do it instead?

>>2945584
is anyone aware of a wire ampacity chart for high frequency? i'm looking at home-built ham radios and despite the high frequencies and the skin effect, most of them just use solid core (rather than litz etc.) for inductors/transformers.

>>2946594
Use a current-limited power supply and you probably won’t kill anything. Before trying anything experimental or making a PCB, I’d just get a rotary encoder and some buttons and a USB a breakout on a breadboard, to ensure you can replicate the original function of the mouse.
You should make your mouse have USB B socket, I love USB B.

>>2946601
Didn't occur to me that I could change the USB micro-b to something else like C. Yeah good idea for doing a proof of concept for it actually working first I was thinking to upgrade them to optical switches/encoder so they would last significantly longer but would this implementation introduce an entirely new set of those passives you mentioned? I'm guessing they are there to maintain the fastest latency since this thing is a gaming mouse.

>>2946610
>would this implementation introduce an entirely new set of those passives you mentioned
Definitely. Optical or hall-effect switches will need different supporting components. It's not trivial to alter the circuit to use these switches instead, but it's hardly impossible. You can likely just measure the voltage on each header pin while it's powered up, click the switches, and measure how the voltages change. Then so long as whatever you do with an alternative switch has the same voltage change, you're good.
Passive components have little to do with latency, they're just the required supporting components. You'll find them in any mouse. Debouncing can be fast or slow, but I suspect the fastest method for a mechanical switch is to use no pullup resistor and just have the SPDT switch pull both up and down, with a small buffer capacitor and maybe current limiting resistor to prolong switch life. The gate of a MOSFET might even be enough. As far as I know, no mouse manufacturer does this.

>>2946611
Oh shit I can add SPDT switches and have them work like Endgame gear's OP18K mice?
Would there ever be a chance of a future firmware update from the manufacturer being incompatible or bricking the entire mouse because of my foreign optical components (even though they are dumb)? I'm guessing I have to reverse engineer all the circuitry exact to spec.

>>2946599
I know it’s common to use copper tubing
(which is hollow, obviously) for loop antennas.

>>2946615
>Endgame gear's OP18K
No they're doing something fancier, since they can adjust the effective lift-off distance in software. No clue what they're doing there if it's mechanical (measuring bounce?), if it's optical or magnetic you could definitely just read an analogue output from each switch. All the 3-pin microswitches you see are already SPDT, though I believe they do use pullups and software dead-time or RC hysteresis for debouncing.

>>2946637
>>2946615
>Would there ever be a chance of a future firmware update from the manufacturer being incompatible or bricking the entire mouse because of my foreign optical components (even though they are dumb)?
If it's a smart mouse trying to measure switch bounce, or the existing switches are some sort of analogue variable switch, then maybe. If it's something dumber like my ancient G303 then likely not.

>>2946480
I cut the LED circuit back out, it still oscillates. It's not audible until after about halfway on the gain knob. Then it increases in frequency until the knob is maxed out. The volume and tone knobs work normally. I guess I need to look at the opamp, around the feedback section. It's weird though how I can hear it even when it should be in True Bypass mode
>>2946482
will try

File: RAT.png (781 KB, 1920x839)

781 KB PNG

>>2946482
Anon, it already has a 100pF cap across its single op-amp. Another 10 wouldn't do much.

>>2946677
A problem you may run into, is that the 100pF is already a significant enough impedance to influence the frequency response of the circuit. At 20kHz 100pF is 80 kiloohms, so using a 1nF capacitor will almost certainly change the tone by rounding off the sharper edges.

But I can't really see why it would oscillate, there shouldn't be any positive feedback possible, and the phase delay in the feedback loop is simple and small. Assuming it's built as designed, and worked as it was designed. Maybe the high impedance power rail is causing feedback, in which case you could try adding a bunch of bulk capacitance, or reducing the series resistor. I see some schematics with a 100R and some with a 47R, but it seems you've got a 47R so that likely isn't the source of the issue.

I guess I'd just clean the board thoroughly for any flux residue or little bits of swarf or solder bridging pads. If you have a scope you can check the inverting and non-inverting inputs to see what's going on to some extent, I guess.

>>2946682
I was looking at the stripboard >>2946476 where the 100p doesn't seem to go back to the opamp. it goes from pin 5, to gain 2&3, (presumably) to gain 1 and then back to pin 5?
oh wait no I'm retarded - the vertical lines aren't cuts, they're wires. the red squares are the cuts
>>2946677
>the gain knob
do you mean R DISTORTION or R VOLUME? (presumably you don't mean R TONE)
what exactly is oscillating? U1 or Q1? show us some scope traces. say the output, negative input and positive input on the op amp and all three pins on the JFET

>>2946697
It's almost certainly R DISTORTION he's referring to, because that controls the gain of the op-amp circuit, and it's a lot more likely to oscillate than a source-follower.

Could also be a shitty fake LM308 from alibay. I got fake TL072s from a "reputable" seller over there, so I wouldn't put anything past those guys.

>>2946704
oh yeah that's a possibility
C8 could be a suspect if anon doesn't have the chip he thinks he has. driving capacitive loads is always tricky. a bit of extra resistance in series or a smaller C8 might fix the issue if that's the case

File: currents.png (19 KB, 915x904)

19 KB PNG

I want to use a voltage of 0-5 volts to output two matched currents between 0 and 1 mA (to control a pair of OTAs)
This circuit works in Falstad, but will differences between the transistors mean that it doesn't work IRL? I understand that I could match transistors based on VBE, but is there any kind of feedback mechanism here that would cause all the current to come out of one transistor only, for example?
I think the "current mirror" concept is related to this but I am a complete noob

>>2946720
just use a second op amp. in fact your currents are low enough that you could skip the BJTs entirely and just use two op amp current sources. just regulate the voltage drop across a reference resistor

>>2946721
I'm trying to optimize on part count as much as possible, because I will need to build 6 identical circuits on stripboard.
I'm controlling LM13700 OTA's which expect an input bias current of 0-1 mA, and the bias pin sits at -9.6 V. I've seen some designs where a simple voltage is applied and the bias current is determined by a resistor, but I understand that using a current source like this will be better.

>>2946723
so you're biasing two OTAs (the two in one LM13700 I guess)? and it's the amp bias input, not the diode bias current? I'm guessing you want to control the gain using this current. but the OTAs in the LM13700 are matched, so if you wanted the same gain on both that would be easy - just tie them together like fig 22 in the datasheet. the fact that you say that you want two different currents says to me that you want a voltage controlled pan instead. is this correct?
it might be possible to do this by having just one OP, and "stealing" current from a fixed bias on one of the OTAs
>using a current source like this will be better
probably. it depends on how important linearity is. the voltage on the I_ABC pin probably isn't constant, but varies with the signal level

File: Screenshot 2025-09-24 114648.png (39 KB, 998x300)

39 KB PNG

>>2946720
>will differences between the transistors mean that it doesn't work IRL?
depends on how well it needs to "work". if youre cool with a 20% difference between the two then this would probably work. it would work even better if you put resistors R_E >> r_E on the emitters so they dont share a common Vbe. 100 ohms should be more than enough. 1 kohm would get you well under 1% matching.
>is there any kind of feedback mechanism here that would cause all the current to come out of one transistor only, for example?
for a given Vbe, current decreases with increasing temperature, so youre good.
>>2946723
>I'm trying to optimize on part count as much as possible
they sell dual op-amps.
>I understand that using a current source like this will be better.
the matching between the OTAs is almost assuredly better than your present circuit; datasheet says 0.3 dB which is either 7% or 3.5% depending on how they did the math. if you really want it tighter than that, i recommend either pic rel left with a dual op amp or pic rel right with a quad. you may have to adjust the gains with a pot to get them to track well.

File: filtor.png (129 KB, 2095x1152)

129 KB PNG

>>2946729
This is for a 2-pole voltage-controlled filter (God help me). So, the OTAs control the cutoff frequency of each stage by sending current to the amp bias input, yes. I hadn't seen Figure 22, but that would seem to be the simplest option because the bias currents should always be the same. I was working from other designs where one PNP transistor per bias pin was used, but that was for a 4-pole filter. Maybe I can get away with it because, as you say, the amplifiers on the same IC are matched.

I think I have it all mainly working in Falstad, although the characteristic resonance peaks on my input square wave appear at the wrong place in the circuit. I also wonder whether I really need to invert the signal before feeding it back in the resonance loop.

>>2946742
I can't believe another reply happened just as I was writing my own but thanks for the details anon, it seems that I should really just use one transistor and split the control current across both of the matched bias pins.
As for part count, I will probably be building 6 of these, so even if I can save one op amp per circuit, thats 1.5 TL074s I don't have to incorporate.

How can I probe the little legs of IC chips with my oscope, the reference ground leg is too tiny for an alligator clip

File: clip.png (385 KB, 1024x1024)

385 KB PNG

>>2946754
There are special clips that are small enough to grab onto individual pins of chips. Lots of variations of them are available (not just pic related), they're pretty handy.

File: Screenshot_20250925-00015(...).jpg (556 KB, 1440x1444)

556 KB JPG

Opinion on chinese low temp solder pastes?

File: 6021b41593c0ef55e01d591a9(...).mp4 (3.65 MB, 532x532)

3.65 MB MP4

>>2946815
Guy's acing it

File: Zrzut ekranu 2025-09-25 024935.png (4 KB, 258x209)

4 KB PNG

>>2945584
I've been googling and asking AI for the past hour and I just can't get the damn answer, bros. Even though it's really basic;

Consider the simplest high-pass filter in picrel. Obviously, capacitor loads up and stops the flow on DC/low frequencies, but grok claims some of the low frequency signal passes and sinks through the resistor to the ground. I just can't grasp why it would prefer to go via resistor instead towards V_out, which seems to have less resistance. I'm missing some piece of knowledge that keeps me in retard zone

>>2946839
>capacitor loads up and stops the flow on DC/low frequencies
>some of the low frequency signal passes and sinks through the resistor to the ground
youre going to have to phrase your question differently; a "signal" isnt a physical thing that moves around.
this is the jist:
>the right side of the capacitor wants to stay at ground
>in order for the right side of the capacitor to stay at ground while the left side rises/falls (due to incoming signal), the capacitor has to develop/discharge a charge across it
>the only path to due this is through R
>at low frequencies, the current through R is enough to keep the right side of C close to ground
>at high frequencies, the current through R isnt enough to charge/discharge C fast enough, so the signal begins appearing at the right side of C

>I just can't grasp why it would prefer to go via resistor instead towards V_out, which seems to have less resistance.
why do you think V_out is low resistance?

>>2946847
>>at high frequencies, the current through R isnt enough to charge/discharge C fast enough, so the signal begins appearing at the right side of C
So resistor is just a chokepoint for excessive current?

>why do you think V_out is low resistance?
I guess I'm just assuming. V_out has to go somewhere so the signal goes either idk where with R unknown or through a resistor. Am I supposed to assume the next stage after this filter has to have higher resistance than that resistor for this filter to work?

>>2946851
>So resistor is just a chokepoint for excessive current?
sort of.
>V_out has to go somewhere
...why?
>Am I supposed to assume the next stage after this filter has to have higher resistance than that resistor for this filter to work?
yes. think of it like a voltage divider (it basically is, its just frequency dependent). anything you connect to a voltage divider has to have an impedance much higher than the output impedance of the divider (which, fyi, is equal the two divider impedances in parallel).

File: lipo-100-302025-3w_n.jpg (209 KB, 1440x1080)

209 KB JPG

How trivial is it to turn these jst connector lipo batteries into hot swappable ones?

>>2946854
I think I got this. Thanks, anon.

>>2946859
what's not hot-swappable about a jst connector?

>>2946866
the hot part?

>>2946816
god i wish that were me

>>2946883
Solder paste is a scam.
It goes bad and has a shelf life.
I’m still using my grandfather’s solder from the 40s.

>>2946259
too fat

>>2946859
put it in a nice rigid housing? use an 18650?

File: 71vktNq1rYL._UF1000,1000_QL80_.jpg (77 KB, 1000x1000)

77 KB JPG

>>2946889
that's a good thing

>>2946206
>40mm tip-to-grip distance
Ah, so like holding a pen or pencil. I wonder what sort of soldering irons get used by those chinese guys who hold the chopsticks like 20cm up?

>>2946384
Higher power or more patience. There's some mass on the switch pulling off heat faster than your warmstick applies it.

File: E54DD6DA-07C5-4632-A73E-8(...).jpg (377 KB, 1422x709)

377 KB JPG

>>2946420
I think the ms data sheet said it’s just made of oleic acid.
Not sure why deoxit is so overpriced.

File: Untitled.png (207 KB, 1583x539)

207 KB PNG

the duality of man
some say R-core is better and people go for toroidal out of ignorance. some say toroidal is better and R-core is just a poorfag cope. then there's the guy on the right who says even split-bobbin is better but i thought split-bobbin was deprecated from before toroidal transformers became commonplace in pro audio. a split-bobbin transformer is probably going to have a lower wattage rating for a given physical size but i'm not sure if that compromises the performance in anyway as long as the power output is adequate and you have plenty of bulk capacitance. the true poorfag option is a wall wart like in the alesis 3630 or lindell lin76 so there's switching noise and no ground connection for EMI shielding. either R-core or toroidal can be had for like $20-30 so it's no big deal but how do you decide which one to get? even split-bobbin is valid according to the guy on the right in pic related?

correction: AC-to-AC wall warts aren't SMPS

>>2946927
I’m pretty sure toroidal is mathematically the best.
And maybe you can only do better by an outer toroid to encase the coils, but single toroids are already expensive to produce.

File: 59BC782D-C642-4513-A0C1-6(...).jpg (315 KB, 788x557)

315 KB JPG

>>2946286
We’re with you on that one.

>>2946927
R-core has the benefit that the primary and secondary windings are separated. Less efficient power conversion, but better common-mode noise rejection. R-core is just a toroid with windings that are separated, and with a straightened-out section to more easily put a flux ring around it, in principle you can do the same to a toroidal core. Core material may differ too though.
Marco Reps did an entire video on minimizing common mode noise with exotic transformers, it's worth a watch.

File: filtro.png (103 KB, 1966x1106)

103 KB PNG

>>2946744 (me)
Spent an entire evening wondering why I couldn't get my two pole filter to resonate before learning that you really need four poles to do that, now I built an extremely simple 4-pole version and it worked flawlessly

File: asdf1.png (122 KB, 1457x884)

122 KB PNG

>>2947049 (me)
The low pass reference design from the LM13700 data sheet has this capacitor tied to -15V rather than ground, does anyone know if it matters? I have found other designs online where each cap is just referenced to ground at the output of the OTA. Otherwise I think the topology is the same as what I built in Falstad

>>2946883
>>2946884
https://www.youtube.com/shorts/NPXI8ZpGPd0

>>2947050
That's tied to the output voltage follower, not to ground. Looks like they're getting feedback at high frequencies. Looks like positive feedback too, but I have no experience with the LM13700.

>>2947060
You are right, thanks.
I'm satisfied with the simulation for today, and will probably try to breadboard it at the weekend where I'll inevitably find the real world version doesn't work

File: Roll's Recliner of Rage.png (326 KB, 800x800)

326 KB PNG

I asked a few different AIs what voltage I'd need to get a needle to negatively ionize air without producing ozone and the answers I've gotten have been somewhere between 1 kV and 6 kV.

With an answer that vague, why even bother?

>>2947032
>Marco Reps
https://www.youtube.com/watch?v=9JinSfCKuNQ
intredasting, i haven't watched the whole video yet but huh the split-bobbin at 9:18-9:48 is oddly good. though i don't even known if common mode noise is the main thing to look for because i would use top-tier filtering (cascaded LDOs in some sensitive circuits) but EMI radiation and magnetic leakage flux could mess things up too.

>>2946950
hmmmmmm

>R-core transformers are superior to toroids. R-core win hands down when wound correctly as their stray field is incredibly low. You can use them millimeters from sensitive electronics without issue (still don't do that). I did a project here and tested many and haven't found a single one that hummed. They were used in the better Japanese amplifiers as well which may be convincing. That (insert famous name of brand or designer here) doesn't use them is simply a matter of cost.

>Toroids are often very wideband and therefor don't filter pollution on the main voltage and they can hum as they are often allergic to any small DC component. They are available everywhere which is the reason for their popularity. They are cheaper than E core/R core/C core transformers, they're available and easy to mount. They pushed other more costly types out of the nest and people repeat their positive properties which makes it a self fulfilling prophesy that they are king of the hill. Try to find anything else! There are quality difference so industrial types are preferred.

>>2947070
It isn't the voltage directly that makes ozone, it's the corona discharge. Corona discharge is a product of electric field strength, which is a function of geometry and voltage. I think any ionization of air will result in ozone, because ionization of air means mobile electrons moving between gas particles, which will naturally produce ozone. Even if you can ionize air with a meme catalyst while using barely the first ionization energy of oxygen or nitrogen per electron (13.6eV and 14.5eV respectively) you're far overshooting the enthalpy of formation of ozone, which is 1.42eV. Without such a catalyst you're overshooting the chemical bond energy by a factor of a thousand.

Maybe you can use a catalyst to break down the ozone before it gets in contact with an organism or other organics that could decompose from it, what do you want to use the ozone for?

>>2947083
> i don't even known if common mode noise is the main thing to look for because i would use top-tier filtering
LDOs and LC filters and such filter out normal-mode noise, but common-mode noise goes straight through them. Common-mode suppression chokes and good grounding makes a difference, but it isn't perfect, and sometimes you want something floating.
>Toroids are often very wideband and therefor don't filter pollution on the main voltage and they can hum as they are often allergic to any small DC component
Seems like you should instead be looking up and comparing the magnetic core materials. Because it's the core materials that determine these features.

File: img_1406-1-scaled.jpg (533 KB, 1920x2560)

533 KB JPG

>>2947084
mmmm so if i get what you're saying then a digital audio interface might use split-bobbin not just for cost-cutting and space-saving reasons but also for common-mode noise rejection. but then an analog audio rack unit might have a toroidal transformer because, although line-level audio doesn't take that much power, you might want plenty of headroom and low impedance in the bass frequencies (going off of what a less advanced AI told me months ago). the type of split-bobbin that would fit in a single rack unit (1U) would be like 9-15 watts whereas toroidals are like 30-36W.

and some analog units have many different features built in to process the audio with a large quantity of op-amps so they would genuinely need the wattage capacity

>Conclusion: Choosing the Right Transformer for the Job
Both split-bobbin and R-core transformers offer significant improvements in common-mode noise rejection compared to standard transformer designs. The choice between the two often comes down to the specific requirements of the application and budget constraints.

Split-bobbin transformers provide a robust and cost-effective solution for applications where good noise immunity and high isolation are crucial. Their simple yet effective design makes them a popular choice in a wide range of power supply applications.

R-core transformers, with their superior magnetic properties and balanced winding capabilities, represent a premium choice for the most demanding, noise-sensitive applications. The extremely low magnetic field radiation and excellent common-mode noise rejection make them the go-to option for high-end audio and precision instrumentation where preserving signal integrity is paramount.

In essence, while both are champions in the fight against common-mode noise, the split-bobbin transformer wins on practicality and cost-effectiveness for general-purpose applications, whereas the R-core transformer reigns supreme in scenarios where uncompromising noise performance is the ultimate goal.

>>2947084
I saw this video recently https://youtu.be/kqqkpfd7OCc and instead of trying to use it to convert atmospheric electricity into usable low-voltage power I had an idea I thought would be funny: step voltage up from a battery to charge a needle (or maybe some steel wool) to negatively charge the air, then have a second one wired in reverse to do the opposite, stepping the voltage down to power a battery-powered device and see how far apart I can separate both while still getting the device to light-up or whatever. I figured the higher the voltage the better, but not to the point it's producing ozone or UV.

I like the idea of doing it with only capacitors and diodes but this kind of voltage booster has a (+) input and a (-) output when what I'm trying to do would need both for the inputs and outputs.

File: siliconetubingtohelpwiths(...).jpg (66 KB, 830x499)

66 KB JPG

>>2946457
Anon, add silicone tube on the end I did this to cheap-o aliexpress and the reliability went up a ton. I think I stretched a 6mm OD 4mm ID on mine.

>pic related

Anyone question anyone can answer and I didnt want to make two separate comments. if I am looking for suggestion on a buck converter that maintain same output but accept range of voltage input would I be asking in mcg instead?

Toroidal transformers are constructed by winding the primary and secondary coils directly onto a doughnut-shaped, or toroidal, core. To achieve good magnetic coupling and a low external magnetic field, these windings are typically wound one on top of the other, covering the entire circumference of the core. While this design is excellent for minimizing magnetic flux leakage (low EMI) and achieving high efficiency in a compact size, the close proximity and large overlapping surface area of the primary and secondary windings result in a relatively high interwinding capacitance. This creates a more significant path for common-mode noise to couple from the primary to the secondary side. While techniques like electrostatic shields can be employed to mitigate this, the inherent design of a toroidal transformer makes it more susceptible to capacitive noise coupling.

R-core transformers, on the other hand, utilize a racetrack-shaped, gapless core with the primary and secondary windings placed on two separate bobbins on the parallel legs of the core. This physical separation between the primary and secondary windings is the key to their superior common-mode noise rejection. By creating a greater distance and reducing the parallel surface area between the windings, the interwinding capacitance is significantly lowered compared to the concentric winding arrangement of a toroidal transformer. This inherent design feature makes the R-core transformer a more effective barrier to common-mode noise without necessarily requiring additional shielding.

>>2947087
While I'm not well versed on your particular situation, you're right that it's a matter of compromise. Toroidals are probably the most power-dense transformers out there. Remember that you can send digital audio signals through optocouplers or toslink or whatever, which can basically completely mitigate any leakage capacitance. Couple the clocks and data at the same time instead of using a single S/PDIF signal, if you care about jitter.

>>2947088
Op-amps draw very little current. You could make a graphic equalizer with slide pots and two dozen op-amps run off <50mA if you knew what you were doing.

>>2947090
Interesting video, though you still need a switch that is capable of disconnecting the high voltage. I don't think your electric field is something you can harvest energy from because it's DC. You'd have to get it to the point where it's leaking enough current through the air, instead of just relying on capacitive coupling like a tesla coil can. You'd also need that current to flow in a circuit, likely relative to some common ground reference.

>>2947095
/ohm/ is better suited to SMPS design than /mcg/ is. What input range and output voltage? What's the maximum power it will need to source? Any noise requirements? Does it need to be isolated? Any other key considerations, like efficiency or reliability or mechanical strength or size?

>>2947089
>>2947096
Cease

I bought an electronics book and a breadboard a year ago and it's just been sitting around
Should I still spend some time to try it out or is it useless

>>2947103
>/ohm/ is better suited to SMPS design than /mcg/ is. What input range and output voltage? What's the maximum power it will need to source? Any noise requirements? Does it need to be isolated? Any other key considerations, like efficiency or reliability or mechanical strength or size?

It will be powering a esp32-s2 mini, a small 5v vacuum pump motor (according to m y bench power supply it stays 300-500ma and I believe stalls around 1-1.5 which I got from various datasheet with similar voltages so that's just a guess) I hacking together pcb thats controlling the motor, using esp32 wired to various things on the pcb and read sensors associated with the motor, because the motor use to be ran off single 18650 I would powering it the same as if there was a battery (4.2v) also a solenoid that works at 5v

I wanted to include compatibility to be driven off of 12v lipo for fps drones I have as a "just in case" idk how stupid it sounds if I input 5v it will output 5v if I input 12v it still stays at 5v maybe im asking for the ipossible, I have couple of power supplies I an use to power this if what I am asking is impossible or not readily available as a module, I'll just use mp1584en module or lm2596

>>2947083
> best
Best at what though? Lol.
> common mode noise
I think you can get rid of some common mode noise simply by twisting the wires together (aka “twisted pair” like you find in ethernet cables.
I’ve seen that before inside metal enclosures.
My theory is that, by twisting them, they induce the opposing magnetic fields in one another thus cancelling the noise signal in both wires.
Fairly brilliant idea. And doesn't need anything fancy.

>>2947154
best at anything performance related, not giving a shit about efficiency. the japanese are clearly smarter than you. not even going to reply about that other thing.

>>2947153
Generally a buck converter will have a small amount of dropout and so won’t be able to output exactly its input voltage, but with a half-decent MOSFET you can minimise this. The switching converter on the Pi Pico board can turn anything from 1.8-5.5V down to the 3.3V it runs at, because it’s a buck-boost converter. If you want to be able to run off a 3S or a 1S battery buck-boost would be the obvious solution, though if you’re fine with running the vacuum pump and solenoid at 3.2V you could simplify it a bit and just use a buck converter for everything. You could even just run the vacuum pump and solenoid directly on the input voltage, and PWM them so they don’t overheat at voltages above 5V. Then you could use an ultra-low-dropout linear regulator like the HT7333 for the ESP, no switching converters required.
It’s also an option to have seperate inputs for a 12V lipo and your 5V USB or whatever.

I’d test how they operate on lower voltages, and what their stall current is.

>>2947154
Common-mode noise through a transformer comes mainly via capacitive coupling between the primary and secondary, i.e. its electric field noise. Shielding, grounding, and separation is the antidote to this, twisted pairs are for magnetic field noise. Though in the video Marco shows how coaxial windings can have a useful shielding effect.

>>2947177
>Generally a buck converter will have a small amount of dropout and so won’t be able to output exactly its input voltage, but with a half-decent MOSFET you can minimise this. The switching converter on the Pi Pico board can turn anything from 1.8-5.5V down to the 3.3V it runs at, because it’s a buck-boost converter. If you want to be able to run off a 3S or a 1S battery buck-boost would be the obvious solution, though if you’re fine with running the vacuum pump and solenoid at 3.2V you could simplify it a bit and just use a buck converter for everything. You could even just run the vacuum pump and solenoid directly on the input voltage, and PWM them so they don’t overheat at voltages above 5V. Then you could use an ultra-low-dropout linear regulator like the HT7333 for the ESP, no switching converters required.
>It’s also an option to have seperate inputs for a 12V lipo and your 5V USB or whatever.
>I’d test how they operate on lower voltages, and what their stall current is.

Thanks I'll give it a shot, I also may just keep it simple and have it be rated for 12v instead of trying to accept voltage range. if I have to change power supplies later since im gonna use this old atx psu which seems to handle the pump fine its rated for 5 amps, once the psu kicks the bucket then simply adjusting output on the buck converters is all that is needed.

File: gfsessefse.png (302 KB, 1558x713)

302 KB PNG

good sources for surface mount to through-hole adaptors? I could do with breadboarding some SOIC chips. This pack seems almost too good to be true with how cheap it is, though

>>2947199
what do you mean almost too good to be true? Price looks right for what you get, did you verify the type you need is in there?

File: sch.png (21 KB, 1103x633)

21 KB PNG

Can you guys help me improve my design? I'm feeding this circuit a stepped down mains (9V RMS). I want a triangle wave at the output but at the moment I got a sawtooth waveform. How could I do that with the fewest possible components?

>>2947223
35 adapter boards for 10 dollarydoos seemed like a good deal, then again, I guess they are cheap to manufacture.
I need to adapt a SOP-18 chip so I figure I will use the 20 pin adaptors and just leave two unconnected.

>>2947236
Yeah I see what you mean I googled "cermant" brand and they're selling 37 piece like yours foe 30 bucks on ebay different seller than that one but I also saw 10 packs of different seizes for 0.75 to ~3.25

>>2947242
I just went ahead and ordered them, what's the worst that could happen

File: adapter-board-sop28-dip.jpg (125 KB, 800x800)

125 KB JPG

>>2947242
the brands are made up, you can get the same design from any number of places. aliexpress might not be worth it now because of customs idk i'm not american.

Hi /ohm/, hoping for some desoldering help.
I've been trying to remove a few 3 leg micro switches from a board, but had a lot of trouble last night. Had trouble getting the solder to melt without adding more, but it's just ended up with me sticking more on and having to remove it. I have copper braid but it wasn't wanting to stick. The solder sucker has been useless as it sets again before I can get the tip over.
I'm going to try again using a larger chisel tip, dip the braid in flux and try again. Anything else I can try?
I went looking for a desoldering gun today but there seems to be nothing locally, only irons.

>>2947234
If the wave frequency is constant maybe you can use a low-pass filter to get it looking like a triangle-wave. A current-mirror feeding into a capacitor might work for a better wave-shape, tied to the input wave via a diode. Whether you make a current source or sink, and whether the diode points in or out of the capacitor, depends on the polarity of the sawtooth. Also the current mirror needs to be making little enough current that the sawtooth input can easily overpower it.
Otherwise you could try and make a wave-folder, but you’d probably need an op-amp for that.

>>2947262
Get some thick solid-core copper wire (e.g. 1mm if not thicker) and wrap it around all three pins, adding solder to each pin. Then put your iron on the middle pin with as much surface area contact as possible, such that the heat from the iron is conducted to all three pads. If one of the end pads is tied to a ground-plane you may want to put the iron on that side instead.
Otherwise buy some aliexpress desoldering needles.

>>2947258
>aliexpress might not be worth it now because of customs
tell me about it
I heard the tariffs might go away in October, I sure hope so

>>2947279
>I heard the tariffs might go away in October
Where did you hear that? WW3 is about to go hot.

File: 1733861832165565.jpg (477 KB, 2150x1614)

477 KB JPG

I practiced a lot with junk boards to get the hang of SMD rework, but I still struggle with it: one thing I always fail at is not melting JST connectors. Today I had to change a TO-252 near a JST connector, and once again the plastic of the connector is melted and charred, pic related (the fibers are just remains of qtips when I cleaned all the flux). Any tips?

File: firmly grasp it.jpg (40 KB, 500x379)

40 KB JPG

>>2947323
hot air and pull on them at the same time

>>2947323
herp I didn't read your post properly. yes plastic parts near parts you want to remove is tricky. try using some kapton tape to shield the plastic from the worst of the heat

File: 1734102343715941.jpg (43 KB, 632x474)

43 KB JPG

>>2947329
>kapton tape
>mfw
I might be retarded

>>2947130
Yes you should. Maybe start with some LED circuits

File: 36t54.png (945 KB, 1046x638)

945 KB PNG

>>2947064
>will probably try to breadboard it at the weekend where I'll inevitably find the real world version doesn't work
against all odds, the circuit worked in real life too. My cutoff frequency is rather high though. I know that it's determined by resistor and capacitor values but are there any rules or heuristics for how to vary those component values? I guess you can't solely tweak the capacitors and never change the resistor, for example

>>2947323
I’ve noticed that jst (and other) connectors often use plastic that doesn’t tolerate heat, and in factories they’re often put on with a separate process.
I just use aluminum foil and blue tack to protect them.
> kapton tape
A bit overkill in my opinion.
Probably the best thing would be some thin shims of alumina fire brick like a shuttle tile.

File: IMG_20250921_133135599_HDR.jpg (3.6 MB, 4096x2304)

3.6 MB JPG

Behold, my VCO. I'm dipping my feet in modular synth stuff.

File: scope.jpg (2.84 MB, 4080x1884)

2.84 MB JPG

>>2946682
>>2946697
>>2946704
>>2946713
I was referring to the R DISTORTION knob. It's probably just caused by a fake chip. But I should also clean the flux better too. It's just shitty stripboard and I soldered it a long time ago already. Q1 should be biased okay.

>>2947479
try removing C8

File: 1_CRpJCqBuZynv0nMtmrB13Q[1].jpg (1.86 MB, 4496x3000)

1.86 MB JPG

Pic is just for attention I am not looking at that model specifically.

Curious on recommendation? I have seen people recommend handheld ones and seen some what looked like super slim but I am curious if theres any models that happen to have hidden mods or something that unlocks more features?

It may not be athing, I just heard about stories on diffferent electronics being torn open and companies using boards that are in higher tier product but with missing chips you can buy online and suddenly you are running most espensive model with exterior looking like cheaper model

Is that still a thing and any under the radar? Worth a shot asking.

>>2947482
What about removing C3 and then try a real NE5534 to see if the oscillation stops?

>>2947489
removing the compensation cap will make it less stable, not more
anon could also try using a not-obsolete chip

>>2947490
the distortion of the op-amp is part of the effect

>>2947486
>recommendation?
State your budget and needs.
I got the DSO5072P, the one in your picture, long ago as I was starting out and didn't have a deep wallet. The firmware is chink trash, sometimes I wish I had more than two channels, mostly for convenience, but I never objectively needed more than that and I have yet to outgrow it. The DSO5072P can be modded relatively easily into a 200MHz scope IIRC, it's all detailed in the eevblog forum.

>>2947492
>I have to have the exact same part as the other guy or I won't sound le identical!
why are musicoids like this? just develop your own sound

>>2947496
that's like telling someone to build their own piano and it's going to be as good as a steinway. you might as well use a software plugin at that point. the original manufacturer ditched the LM308 because of availability issues but still went with an old school OP07DP because the distortion is part of the desired effect. there's even a band named after the LM308.

>>2947495
Sorry about that, lets say my budget is 200 USD maybe thats way too low lets say 200 is preferred but 300 is absolute max, and you mention channels wishing have mor than 2 I have a feeling I would be in the same boat and would prefer 4 channels, however I don't think theres 4 channel in that price range, but maybe theres a 2 channe lcheap enoug hwhere I can buy 2 adn still be within budget. Maybe not

https://equipboard.com/posts/lm308-chip-history

>>2947500
>needs
I forgot that too ugh I havent sleep in 3 days and im starting to see the results for it.

basically the main part is to help ID or solve issues that oscilloscope can achieve that, essentially, I am so wishy washy on this, and its piece of equipment i keep puting off on buying.

>>2947436
lack of reverse image search results suggests it's not a troll post so I'm curious about what ICs you are using. Why so many trimmers?

>>2947436
the TRS being mounted on the long side triggered my 'tism for a bit until I noticed this is a 100x80 mm (-ish) card

>>2947504
Lol, no trolling, there is just a lot of things to adjust. The design requirement of it is to produce triangle, square, sawtooth and sine waves at a consistent magnitude. Its frequency range is 20Hz to 40kHz. The control voltage is 1Volt per octave. The project started because I acquired literal thousands of TL084's from my old job for free, and wanted to do something with them. This vco is based on a triangle core using the lm13700. 4 of the trimmers are used for setting the output magnitude. An other 2 are for the trip to sine wave converter. 2 more used for the exponential converter and HF error correction. Another 2 used for for the zero and span adjustment for the envelope tracker for the sawtooth shaper circuit. 1 is just a test CV. It's ugly but it works and need to know that before I order proper PCBs from jlcpcb

>>2947507
based and building a prototype before ordering PCBs pilled

>>2947507
Is the sine also generated with an LM13700? I'm building an oscillator too, but I left out sine wave because I couldn't find a circuit that didn't require trimming. I want my system to be polyphonic so I have to build 6 of everything.

File: th_sine_shape.jpg (45 KB, 1024x680)

45 KB JPG

>>2947516
No, its triangle output.
I used attached image. its a set and forget kinda thing, my trimmer pots remained centered. I think you could probably try using fixed resistors for this.

https://www.youtube.com/watch?v=FzkYmsaucbc
the LM308 sounds glorious, though it might be difficult for a non-expert to tell the difference. you can't take most people's word for it with these things that involve subtleties and often a lot of cope from people who mainly use digital software plugins because they can't afford professional studio quality analog hardware. even a lot of professsionals in the music biz are bluepilled about the subtle technical details, so it's pretty rare to see mega successful acts like the beatles and ABBA and people believe it's almost all in the songwriting and performance and not the technology lmao. for example bruno mars hit a home run with APT even though it's inherently a pretty shitty song. the official proco rat might have modified circuitry to better accommodate the OP07, i haven't looked into it.

File: jerk3.gif (3.24 MB, 479x671)

3.24 MB GIF

>>2947497
>>2947558
>noo they're totally different!
>you have to have golden ears to really hear the difference!
of course there are zero figures to go along with this audiophile wankery. same shit as with toob guys
>the distortion is part of the desired effect
I mean yeah, the specs on the LM308 are pretty shit. 10 mV input offset voltage? absolute garbage, especially with line level signals. you're likely going to get lots of even harmonics as a result. you don't need unobtanium op amps to get that effect. better yet, you could tweak the precise amount of even/odd harmonics added to the signal

>>2947575
> 10 mV input offset voltage?
That’s what trimmers are for.
Things that come with precise, laser-trimmed specs are primarily for factories where they don’t have (or want) some guy (or nowadays, some elderly chinese lady) calibrating trim pots, caps, and IF cans.
If you’re diy-ing a one off, why not calibrate it instead of spending the extra $ on a part that you’te probably going to kill anyway with static, overheat, or a wiring eff up?

>>2947578
most recent parts today have way better offset voltage than that, without trimming. but that's kind of besides the point, namely that looking for le special sound in jelly bean opamps is dumb

>>2947584
> looking for a special sound is dumb
It depends.
Some of those effects pedals have very questionable schematics for sure.
However, we often need to look for specific parts.
Transistors with a specific hfe (so you don’t necessarily need negative feedback)
Transistors with matched characteristics for amplifiers or darlington pairs.
Noisy diodes for random number generators and white noise generators.
Sometimes I’ll go through a dozen 10% resistors looking for the best one in a voltage divider.
Even GPUs that won’t run at 1400 MHz are downclocked and re-sold at a lower frequency.

>>2947591
sure, but the amount of cargo culting among some audio guys is astounding

File: de.png (560 KB, 760x473)

560 KB PNG

>>2947521
hey that sine wave shaper circuit works pretty well, thanks anon
Wish I hadn't built my whole system so far around 5 V peak to peak when this needs 10 V though, lel

>>2947597
I’m surprised you didn’t go with a +5 0 -5 system at a minimum, the usual standard being +15 0V -15.
The reason is because a lot of the pre-amp area stuff (filters, shapers, generators, modulators, etc, etc, ad nauseam) use miniscule amounts of current and it’s easier to go down in voltage rather than go up.

>>2947601
There is somewhere that I can swap out a resistor and increase all my internal signal amplitudes, so I might just do that. It probably makes sense to stick to standards even though I'm not really making a modular system as such. The plan is 6 chains of VCO -> VCF -> VCA that will all get mixed together at the end, and then I might subsequently apply other effects to the mixed signal afterwards.
One problem I faced was finding multiplexer chips that could run off +/- 12 V because that's what I'm using to select the waveform out of the VCOs. I ended up finding a chip but they're going obsolete which I think is why I got them cheap.

>>2947594
many amateurs believe in the equivalent that they can take a free piano that would otherwise go in the trash and become lang lang tier successful with just a lot of practice and hard work. yeah good luck with that. the guitar tone is a big part of many famous musicians sound and you have no authority to bully someone like kurt cobain (if he were still alive) to use some random other pedal or some meme software emulation of a pedal, he's going to use whatever he likes and he happened to like the proco rat pedal.

just in case you guys didn't see this:
>>>/g/106720752
https://www.youtube.com/watch?v=8TlgrIAI8x4 [Embed]

from >>>/g/106725336: https://limewire.com/d/yTB5o#XKW5MW0C4g

File: iu.jpg (63 KB, 960x540)

63 KB JPG

>>2945903
>He doesn't know about pll controlled vcos
>he doesn't know about generic frequency generator peripherals
>he doen about cherrypicking generic germanium diodes to be used as a gunn diodes in ghz cavity oscillators
> he doesn't know about undocumented working modesof cheap FETs
> babbys first crystal oscillator when?
Srsly ngmi

File: oscill.jpg (169 KB, 1713x160)

169 KB JPG

>>2947479
replaced the opamp with a NE5534 and now it oscillates more clearly if anything
https://voca.ro/1nGKoBk3v86t
If I removed C8 wouldn't the tone control not work?

File: Darumi_Fullbody.png (397 KB, 590x1059)

397 KB PNG

I have an unopened breadboard and all of that stuff for about a year...is it still worth it to try it out?
It won't catch on fire or nothing right?

>>2947755
your breadboards are probably stale

>>2947755
if it starts to oscillate it's a point of no return no homo

File: Mari_Illustrious_Makinami.png (54 KB, 132x375)

54 KB PNG

Quick question: If i get a IR Laser diode that matches the frequency of the LED in a remote control and replace the remote's LED with it can i make a really long ranged remote control?

Pic unrelated besides being best girl.

>>2946388
>>2946388
>>2946421
>>2946516
>>2946574
It was as these anons said, getting solder out of a through-hole with just a wick is near impossible. The scroll wheel was even worse. I managed to get it done by cutting the pins, partially destroying and prying at the the switches, and carefully scraping away the solder with my iron. As a bonus, I didn't break the mouse in the process.

Would definitely recommend just buying a desoldering gun. I will the next time I need to do something like this.

>>2947844
it could be difficult to aim

>>2947847
>It was as these anons said, getting solder out of a through-hole with just a wick is near impossible. The scroll wheel was even worse. I managed to get it done by cutting the pins, partially destroying and prying at the the switches, and carefully scraping away the solder with my iron. As a bonus, I didn't break the mouse in the process.

Did you cut a piece of wick off or no? probably not and the wicks amazing heat transfer was just heating the rest of the wick while you trying to fight that, I wick through holes with ease but then again I cut a small piece off.

>>2947905
motherfucker
he holds a loose piece but doesn't mention cutting it, he goes straight to step 1. place the wick over the solder at 2:50. many other videos don't say anything and show a close-up shot so you don't see if they're holding a loose piece or if it's still on the roll.
https://www.youtube.com/watch?v=L6u5toKyFHk
i'm still skeptical about how good the braid would be compared to a good pump for through-hole. fuck those bottom tier chinese pumps but the relife RL-084 plus (soldapullt clone with silicone tube to put on the nozzle) works great with occasional cleaning of the gasket. it'll probably last longer between cleanings if i add grease and a second gasket to improve the seal (iirc it can fit two gaskets but only comes with one installed). for SMD i guess the braid makes the most sense for removing remaining solder after desoldering the component.

>>2947756
>>2947776
Should I actually go through the book and see if I like it? I'm scared of causing a fire

>>2947962
theyre fine. stale was a joke cause its a breadboard. year old bread would be very stale.

>>2947962
You'll be fine, just watch out for electromagnetic infetterence.

>>2947916
Well I only discovered cutting the wick like a few months agoo so im not smart I do have a pump cheap one from aliexpress but actually is decent. >>2947095 is my comment about adding silicone hose on to it. Not sure if you are going for style I am using or a electric one but it has worked much better with silicone piece on it.

File: Screenshot 2025-09-30 224125.png (1.21 MB, 973x1008)

1.21 MB PNG

>>2947247 (me)
>what's the worst that could happen
my first surface mount soldering attempt........

>>2948086
Anon good to see an update seems fine since its just being used for whatever, no pins bridged either

>>2947681
Hehe I actually do know about all of those but for whatever reason the "all in one" chips with VCO+DDS/PLL just seem way too integrated to really understand what's happening. Plus even discrete VCOs are really expensive for what they are, so I don't want to buy them.
I did make a 3.3MHz oscillator using a single BJT and a self wound centre tapped ferrite coil but honestly that topology isn't that impressive at all.
Its a different story if and when I learn to design simple harmonic oscillators (clapp, colpitt, vackar, hartley whatever) with the minimum of parts.

>>2947379
Is that an oscilloscope? Which one is this?

just got my new solid state relay. datasheet says they can be turned on at 3.3V but the label on the says 4V. which should I trust? QLT Power - SSR-2528ZD3

>>2947885
It is for a light a couple of hundred feet down my driveway, I'd probably put a round rearview truck mirror behind the receiver and use my nerf IR viewer to aim it then permanently fix the remote on a window sill pointed at the mirror.

>>2948152
Try it and find out for yourself. Adding a small charge pump is easy if you need a little boost.
>>2948187
You could do that with a remote controlled relay installed in the light.
https://www.amazon.com/long-range-remote-switch/s?k=long+range+remote+switch

>>2948201
I plan on controlling them with a Pi, so I have 5V available. just annoying to have to maybe to a level shift

>>2948201
Too much power needed, i'm off grid. Those things want 40 amps which isn't happening off of 4 12v batteries and a 500w inverter.

File: osilosope.png (385 KB, 1228x571)

385 KB PNG

>>2948148
It's not really an oscilloscope so much as a generic microcontroller reading a voltage and writing it to a display, but it's more than adequate for my purposes (audio frequencies) and the price was right. I'm happy with it, but my requirements are modest indeed

i need a current source/sink with <300 fA/√Hz of noise @ 50 Hz. ideas? its driving a large load at low currents e.g. 500 kOhm @ 150 uA (75 V) so an inductor would be unfeasible.

File: 1561405659092.jpg (20 KB, 640x640)

20 KB JPG

>>2948227
>Those things want 40 amps

>>2948232
>source/sink
to clarify, that’s either-or; only single quadrant operation is needed.

>>2948227
what

>>2948246
>>2948235
Oops, they can HANDLE 40 amps. Sorry, i have the dumb.

>>2948249
It would be impractical from what i'm reading, a RF keychain switch would work better for like $10.

>>2946108
Yeah, boiling metal makes me jumpy

Also, trimming leads and headers to keep the bottom of the board thin leads to tiny flying metal bits, which also make me jumpy

>>2948228
Nice! It's very suitable for me for quick debug sessions as well, as I don't have a lot of space on my desk. It reminds me a lot of the game boy digital sampling oscilloscope. Thanks a lot!

>>2948232
a low noise op amp? possibly with an output transistor that can handle the voltage. the voltage noise due to the current noise across the resistor is 150 nV/rtHz, so there's quite a few parts to choose from. the MAX4238 should work. it meets your specs while also having quite low input offset voltage and input bias current

bring back 5V

>>2948086
Also first photography attempt I see

>>2948228
Cute.
I think a standard iphone might be a better oscilloscope in the audio range (if such an app exists).
Nowadays you’d need the apple headphone jack dongle which was $40, and you’d probably want a “front end” circuit.

File: file.png (32 KB, 551x408)

32 KB PNG

>>2948314
im gonna try pic rel.
i probably cant use a BJT to block the HV since the shot noise of the collector @ 150 uA would be about 7 pA/rtHz which already blows the budget, but maybe a FET would work; ive seen reports that FETs dont suffer from shot noise, only Nyquist noise.
the other side of the load is allowed to be floating outside of the business band, so maybe pic rel can be part of a larger loop where the voltage at one end of the loop is controlled at low frequencies and the voltage at the other end is controlled at high frequencies. that way the amps in pic related dont have to worry about high voltages. thoughts?
thanks for the rely fren <3

>>2948363
to minimize noise you want to maximize transconductance. you can do this by biasing the gate such that the maximum voltage swing of your op amp corresponds as closely as possible to the voltage swing you need on the gate. you can do that entirely passively with a suitable resistance network. it's not terribly hard to work out the resistances necessary to turn the +-Vs output from the op amp to your desired Vgs(min) .. Vgs(max). doing so minimizes current noise due to voltage noise * transconductance
when I read up on this, the sources I found say a FET is always better than a BJT when it comes to noise. you can never get away from the inherent shot noise in any active device, but you at least don't get any extra shot noise from the gate as you do with a BJT
you'll get shot noise even if you drive the thing directly off the output of an op amp. the typical way to remove shot noise is to either increase power of the drive or use lock-in amplification

>>2948377
>to minimize noise you want to maximize transconductance
herp, I meant minimize transconductance of course

>>2948363
If you're going to use a JFET, the CPH3910 is probably the lowest noise model now that the BF862 is discontinued.

>>2948380
noise figure probably doesn't matter in this application, since the noise out of the op amp is likely far far above the thermal noise. especially if anon chooses the lowest possible values for bias resistors in what I described >>2948377
oh and I forgot: anon also needs a low noise voltage source like LT3045 (positive) and LT3093 (negative) to generate the appropriate bias with as long noise as possible

File: file.png (177 KB, 451x722)

177 KB PNG

k so turns out plugging Ic into the shot noise formula is *not* correct...
i guess im free to use a simple op amp + BJT current source as long as it doesnt have some horrible 1/f noise down at 50 Hz.
>>2948377
>when I read up on this, the sources I found say a FET is always better than a BJT when it comes to noise.
i think that might just be true for amplifying weak signals, where input current noise * source impedance dominates. but is that still the case with zero source voltage noise and zero source impedance? youll have to forgive me, ive never used a FET in my life.
>use lock-in amplification
actually the reason for this circuit is to measure the noise of a photoconductor vs current; you have lots of options to separate signal from noise, but fewer options to separate noise from noise. zero options, in fact. so no fancy domain-shifting noise reduction techniques for me :^)
>>2948386
>anon also needs a low noise voltage source like LT3045 (positive) and LT3093 (negative) to generate the appropriate bias with as [low] noise as possible
"as low noise as possible" would be chemical batteries. for larger voltages, ive had a good amount of success with just a 431 + pass bjt + 2 enormous metal film caps.

File: led_lamp.png (357 KB, 4683x2059)

357 KB PNG

Which boost regulator would be good for powering a few LED strings from 2 AA batteries? Most of them, such as the mcp1640, boost only up to 5.5 V. The chip LEDs I'm planning to use have two LED dies on one chip and Vf around 5.5V. I think I will need a bit higher voltage than 5.5 V for reliable LED driving. So far I've found the tlv61046 that boosts up to 28 V. It could also light up two or more series-connected chip LEDs on each string.

File: driver.png (42 KB, 1596x1009)

42 KB PNG

>>2948393
>i think that might just be true for amplifying weak signals
your signal is already amplified by the op amp(s), no?
the current noise density in a FET is essentially just shot noise + voltage noise density on gate * g_m * gamma * 2 where gamma is a number between 0.5 - 2.0 if I remember correctly
absent a biasing network you want a FET with inherently low g_m. best is one with a high Rds(on), just below the highest Rds(on) you can tolerate
>zero source voltage noise
the only way you can have zero voltage noise is with a dead short. the best op amps on the market have noise on the order of a 50 Ohm resistor
picrel is a WIP design for a project that didn't go anywhere (yet). here I try to keep noise down via three mechanisms:
>a high ratio between voltage swing and voltage noise density on the op amp
this pointed to OPA2277
>passing as much current as possible via fixed resistors
this ensures as little current as possible has to pass through Q1, minimizing shot noise
>Q1 with as low Ids(max) as possible while still able to dissipate the necessary amount of power
not done optimizing this
>biasing network on the gate of Q1 that enables as wide a voltage swing on U1A as possible
the resistance seen at the gate of Q1 additionally needs to be lower than the equivalent noise resistance of U1, which is (8/0.13)^2 ≃ 3.8 kOhm. this is another reason to pick an amp with wide supply rails
optimizing R3 and R19 is still TODO. the goal is to transform the +-15V swing of U1A into 1-4V for the gate of Q1. this attenuates the transconductance from U1A to 1/10th, turning a shitty 8 nV/rtHz into a respectable 0.8 nV/rtHz. it might still be the case that the LMH6629 is a better choice
>"as low noise as possible" would be chemical batteries
maybe. batteries are a hassle, and if your biasing network needs a lot of drive (which it will with an amp like the LMH6629) then they'll need frequent replacing. you have to use the right kinds of batteries also

File: file.png (64 KB, 610x261)

64 KB PNG

>>2948403
>your signal is already amplified by the op amp(s), no?
by "weak signals", i meant signals with high source impedances, where the current noise into the base is problematic. i was hypothesizing that FETs are actually noisier at lower source impedances compared to BJTs. pic rel from TAoE seems to confirm this to be the case, at least sometimes. and if my current source is being controlled from the output of an op amp plugged right tf into a transistor, i would say the current noise into the base probably isnt a huge concern.
a JFET could potentially work for what i want to do, but i would need to find one with a transconductance of <8, has a corner frequency << 50 Hz, and ideally works at high voltages. does this sound reasonable for a JFET? genuinely asking, i really dont have a lot of experience with FETs.

cool project anon, thanks for sharing.
>turning a shitty 8 nV/rtHz into a respectable 0.8 nV/rtHz
can i ask what youre doing that 8 nV/rtHz isnt good enough?
>batteries are a hassle
no, picovolt-noise regulators are a hassle. batteries are free.
>LMH6629
not sure if youre aware but thats a bipolar input amp. bipolar numbah 1 bby

>>2948417
>i meant signals with high source impedances
yeah but the gate of your transistor is driven by an op amp, which has low impedance. but I might be missing something. maybe BJTs have lower gamma
>pic rel from TAoE seems to confirm this to be the case, at least sometimes
well, it does imply BJTs have lower voltage noise so there's that
>does this sound reasonable for a JFET?
not sure. you'd have to look at the datasheets to know. look for the maximum slope of the Ids/Vgs curve. don't limit yourself to just JFETs. I believe there's two contributions: one from the channel itself (inherent g_m, #charges in the channel) and another from voltage noise on the gate. my design tries to minimize the latter. component choice can minimize the former
>can i ask what youre doing that 8 nV/rtHz isnt good enough?
metrology stuff
>not sure if youre aware but thats a bipolar input amp. bipolar numbah 1 bby
yes I'm aware. actually I think that's why I can't use it - too low input impedance. you need to drive it with like 1 mA I think

>>2948423
>yeah but the gate of your transistor is driven by an op amp, which has low impedance.
which means that the low bias current (and low bias current noise) of the FET isnt as useful. i think when people say that JFETs are low-noise, theyre specifically comparing them to MOSFETs, not BJTs. if you look at all the "ultra low noise" op amps from TI, all of them are bipolar input.
>metrology stuff
interdasting. how big is your signal and what SNR do you need? whats your source impedance? what bandwidth?
i dont know if youve actually built this thing and learned this lesson yet but in case you havent ill save you a headache later: important resistors NEED to be wirewound and important capacitors NEED to be metal-film, ideally polypropylene or polyphenylene sulfide. unless your signal band is >10 kHz in which case do whatever honestly.
>actually I think that's why I can't use it - too low input impedance.
OPA211 is 1.1 nV/rtHz with 125 nA max input bias. paralleling two of them gets that down to 0.78 nV/rtHz at the expense of doubling your bias current and increasing your bias current noise by root 2.

>>2948431
>which means that the low bias current (and low bias current noise) of the FET isnt as useful
oh yeah you're right. but there was some other reason why a FET was more appropriate in my use case
>interdasting. how big is your signal and what SNR do you need? whats your source impedance? what bandwidth?
it's a few MHz at most. the entire system is a laser lock-in amplifier. the signal is likely close to full scale, so SNR is a non-issue. that is, the noise added by the voltage-to-current stage is likely to dominate, which in turn is likely to be dominated by shot noise. which is where we want to be, because shot noise is easy to calibrate out - just vary the power
>important resistors NEED to be wirewound
probably too high inductance at MHz levels. I've use polyprop film caps before though, precisely because they're nice and stable. but beeg
>unless your signal band is >10 kHz in which case do whatever honestly
we're also doing temperature stuff which is at the other end of the spectrum. µHz to mHz. we use Vishay foil resistors as references. they're actually better than wirewound standard resistors, at least the ones we have. but perhaps they have less drift in the µHz range if we can put them in a better oven

File: file.png (75 KB, 2806x950)

75 KB PNG

>>2948441
>the entire system is a laser lock-in amplifier.
oh i see, the circuit you posted is a laser driver? lol im retarded
thats cool, spectroscopy?
>we use Vishay foil resistors as references.
huh, didnt even know those existed :/

here's a drawing of exactly what im trying to do. left is my current setup. we have some photoconductors with resistances of around 500 kOhms. the easy/usual way of using them is in a 1:1 voltage divider. in this configuration, we get around 600 nV/rtHz of noise on the output. this setup is good enough (and cheap) but we want to test devices that have a much larger range of resistances, from 50 kOhms to 2 MOhms, and manually swapping the load resistance for every device is a hassle. i proposed using a current source for biasing instead (pic rel right), but we need to make sure the noise it creates (i.e. the voltage noise generated by the current source driving a 2MOhm ideal resistor) is well below the noise of the device itself. reading up on things, im getting confused about *current* sources generating *voltage* noise. my best guess is voltage noise depends on the load while current noise doesnt, but im not 100% sure.
my original "<300 fA/rtHz" requirement came from taking the thermal noise of a 500 kOhm resistor (128 nV/rtHz) and dividing it by 500k. technically my "real" requirement is less (or equal) noise than thermal at the output of pic rel.

File: d5572cde-40d9-43fc-ace4-7(...).jpg (85 KB, 1000x1000)

85 KB JPG

Anyone using a owon sds1104 oscilloscope? I saw it on bangood for 217.00 USD.

>https://ban.ggood.vip/1mnC0

>>2948535
Great price for 4ch, 100MHz. I bet firmware and screen are cancer.

>>2948535
better than my old rigol, and cheaper too

File: opamp.jpg (9 KB, 303x517)

9 KB JPG

Noob question. Need the output of this integrator which is being fed by a constant current to go to zero after it reaches a certain (negative since it's a decreasing ramp, but whatever) voltage (a diode drop is fine if it helps - anything else is also fine) and then go back to integrating, so it's a sawtooth oscillator, except the integrated current is provided by another part of the circuit so I can't use the standard topology (or at least I can't seem to make it work), any help?

>>2948716
the resistor there makes this a low-pass filter, not an integrator
you could do it with another amp acting as a Schmitt trigger

File: Screenshot_20251004-153906.png (71 KB, 1080x621)

71 KB PNG

>>2948718
You're right, it is an integrator for the frequencies that I care about though; anyway, do you mean something like this? Because I tried but couldn't make it work. If it is supposed to I'll figure it out though, probably need to tweak a few values. Thanks

>>2948719
something like that yeah

any advice on probing 0.5 pitch fpc ribbon cables? i was thinking of sticking tape over the pins i'm not touching but the adhesive might eventually damage the ribbon contacts. i have a 3d printer with 0.2 nozzle if that also helps?

File: 1758821036201278.jpg (349 KB, 4080x2296)

349 KB JPG

I have a bank of 20 3.3 mF 25V capacitors wired in parallel and I want to be able to discharge it very rapidly by shorting two probes coming off of it. (It's a DIY spot welder, pic related) I also want a switching mechanism so that I can position both probes so they have strong electrical contact without discharging the capacitors, and only actually discharge them after hitting a switch of some kind. I'm currently thinking a MOSFET would be the best choice for this. (I'm open to suggestions however)

I imagine dumping a 66 mF capacitor bank charged at up to 25V through just any little transistor will probably blow it up before long, so I want to calculate the peak current that would be passing through the transistor so I have some idea of what might not blow up.

Unfortunately I don't have an oscilloscope, but I do have several multimeters, a wide range of resistors, and know the capacitance and voltage values of the capacitor bank. I don't know for certain what the resistance is for the bank but It's pretty small if i had to guess, less than 1 ohm collectively. I used solid copper wire for residential wiring for almost all of the connections and everything is securely soldered.

>TLDR: How do I calculate the peak discharge current of a capacitor bank being short-circuited?

Or, more specifically, how do I determine what sort of transistor I need to switch it? (or do I use something else entirely?)

>>2948751
Consider an SCR. It is unmatched for pulse application. Cheap and common to-220 devices like 2N6394, 2N6400, 2N6504 have pulse current ratings from 100 to 250 A. There are also TO-247 devices with >1 kA pulse ratings.

>>2948716
could you use another op amp as a comparator against the sawtooth voltage, that activates a transistor/FET that discharges the integrating capacitor when a certain threshold is reached?

File: B866A4C9-756C-4326-BF8F-1(...).jpg (172 KB, 930x560)

172 KB JPG

>>2948751
> 3.3 mF
Milli Farad?

> switching high currents
It’s kind of an unsolved problem. A lot of people make their own switch with a very rugged construction, kind of like a knife switch) with the main feature that you can replace the contacts that arc during the contact process since it will destroy itself.

>>2948716
Bring the output through a high value resistor, apply a voltage divider to bias it so that it’s level-shifted to zero, then re-buffer it with another op-amp.

>>2948751
If you can use a semiconductor switch then you won’t have to worry about arcing or bounce, but yeah figuring out the peak current draw is the hard part. If your multimeter has a fast enough min/max mode you could measure the voltage across a current sense resistor with it. If not, you could make an op-amp sample+hold circuit.
You could put a lower-bound on your series resistance by putting a DC current through your wires (e.g. 1A) and measuring the voltage drop from one end to the other. If you don’t want to make an op-amp or LM317 current-source, just use a series resistor with a normal DC source, and measure both the voltage across the wire and across that series resistor to calculate the current. You can also look at the datasheet for the capacitors to see what their effective ESR is, and even to figure out what wire gauge you’re using and analytically determine your total resistance. This will still just be a minimum because of the contact resistance, but it will be useful.

With a few 200A-rated MOSFETs in parallel it’s definitely possible to switch a spot-welder circuit. That’s what car jump-starters use. An SCR is also an option, but they’re generally rated for higher voltages and lower currents. Avalanche energy ratings are also important though, and SCRs may be better for those.

Take a look at existing schematics for spot-welders.

>>2948779
>Milli Farad?
Each one is 3300 uF

>>2948791
I ordered a pair of IRFP2907PBF mosfet's off ebay which claim to be able to handle 840 Amps of pulsed currents. Not sure if I should put them in parallel or use one of them at a time.

If you do it in parallel you have to put a really low value resistor (like less than one ohm) on the output of the transistor right? (otherwise the slight differences in characteristics makes one heat up faster and can blow it up from thermal runaway(?)

Another thing I might try is discharging them through various sized resistors (all large enough to time visually with a multimeter) and figure out the time constant and/or discharge curve. I know you can find out everything you'd ever want to know about a capacitor with an oscilloscope because I did it in an electronics lab. I'd have to look it up to remember how to do it though. Short of an oscilloscope I'm pretty sure the same method can be used by manually timing it with larger resistors though.

>>2948800
In steady-state MOSFETs have a positive temperature coefficient, and so shouldn’t need balancing resistors like bipolar transistors do. But for pulsed current I’m not sure, you’d have to look into that. Apparently you want seperate gate resistors either way, and probably a decently high-current gate drive circuit.

Your total series resistance will be on the order of 10mΩ, give or take an order of magnitude. Any resistor charge time measurement will just give you an estimate for the capacitance. To get an idea of the milliohm resistance’s impact on the time constant, to say nothing of the parasitic inductance, you’ll need to look at it far faster than a multimeter’s live readout.

Also fuck gookmoot for range-banning my ISP again so I have to wait 2 minutes each time I want to post via 4G.

File: flying test probes.png (526 KB, 900x424)

526 KB PNG

>>2948721
>probing 0.5 pitch fpc ribbon cables?

tape the cable down
make some flying probes using a needle, some springy wire, and a weighted base
if you have a length of nichrome, you can forgo the needle - cut the tip at an angle to make it pointy

File: GFrOFZsbAAAaOcw.jpg (64 KB, 996x1041)

64 KB JPG

I have an old bicycle with lights powered by a generator, I was wondering if I can just remove the generator and retrofit some DC LEDs into both front and back lights and ideally connect it to a USB C powerbank as +- using the original generator connections, the thing is it only uses one wire, the other is just powered through the bike frame itself, will the power loses bee too much? does this kind of setup make any sense or em I being retarded.

Anyone tried making a vacuum tube?
https://www.youtube.com/watch?v=-UEfqAWb3fE
I just watched this video and he didn't really use any special tools.

>>2948838
You could probably just keep the dynamo and use a drop-in LED bulb that doesn't care about polarity. No reason a power bank wouldn't work too, but you'd need to wire in the ground on the new plug.

>>2948841
Glasslinger?

>>2948751
That cap bank looks insufficient to do anything useful.
Get (or make) a large transformer with a low output voltage and a high output current and control it (the primary) with a household lighting switch.
This is how a soldering gun works.
And spot welders.
And arc welders.
And ac mains-powered battery chargers.

>>2948841
Tempted to try making them out of a coke bottle at come point.
There's nuance to making them actually good, but if you don't care that much about performance/lifespan it doesn't seem that hard.

>>2948841
I accidentally made one when we were experimenting with ways of measuring static E-fields for a project. but maybe it doesn't count because it was basically just the metal parts in a vacuum chamber rather than a free standing thing

>making a linear CC/CV supply for electrochemistry
>use an op-amp each for current and voltage
>use low-dropout schottky diodes from the op-amps into the node that controls the power transistor
>pull-up resistor on that node since op-amps can only sink current
>find that a sziklai pair can’t be turned off fully because op-amps aren’t RRIO, pull-up still leaks current into the transistor pair base
>decide to put diode in series with transistor base
>still leaks
>put 4 diodes
>still leaks current
>fuck it, I’m using a MOSFET, spiritio effect be damned
>find a non-logic-level excessively rated FET in my collection
>threshold voltage of 2V minimum, 3V typical @250uA
>finish wiring everything together
>put in box
>test
>with no load current the output voltage still drifts high
>even with 100uA load at 11V output it’s not regulating down to the 5V setpoint
>gate voltage down at 1V or so
>what the fuck
>add series diode and pull-down resistor to gate
>same problem
>but performs ok with a 1mA load now
Fuck my shitty circuit. Serves me right for using a low-side transistor and single-supply rails.

>>2948885
Dear Anon, I hope you realize greentext makes your wall of text particularly hard to read.

>>2948841
that drill lathe is pretty clever and cheap
bending the pins around 13:00 without any strain relief looks super risky. you want to grab the base of the wire with another set of pliers to protect the glass
the fact that grid spacing controls gain vs current is new to me. interesting
I would have put the evacuation stem on the top
>>2948857
>Tempted to try making them out of a coke bottle
soda glass sucks to work with. use some borosilicate tubing, or test tubes. that'll also give you the flare for free
>>2948885
post schematic

>>2948841
I know they’re just copying old tubes, but tungsten is terrible to use for the glass to metal seal. It’s super brittle and tends to develop microscopic cracks along the length of the wire that ruin the vacuum, especially if you try to bend them like the guy in the video. That usually happens anyway since the coefficient of expansion of the tungsten is so different from that of the glass that it produces a lot of strain at the interface. It’s way better to use kovar and an expansion matched borosilicate.
Also the outgassing procedure is super rushed in the video.

>>2948843
I kinda dont want to use the dynamo, its lout and slows you down. why would I need a ground for the LEDs and how can I even ground a bike?

will learning about electronics make me less afraid of it?

File: Capture.png (40 KB, 1490x163)

40 KB PNG

>>2948891
I thought tungsten with an oxide coating was a preferred choice without resorting to alloys

File: car ground.jpg (30 KB, 500x265)

30 KB JPG

>>2948895
>why would I need a ground for the LEDs

so you can use the existing wire, instead of adding 2 new wires

> and how can I even ground a bike?

connect power pack negative to chassis and LED negative to chassis

>>2945584
https://www.pa0nhc.nl/DirectionalCoupler/FactoryPCB.htm

as per this link, the coupling loss is inverse proportional with the RF frequency and it's not exactly linear. i have seen PCB directional couplers used on amateur 13.56 MHz and 15 MHz designs, where the trick is (apparently) increasing the trace length to increase the coupling factor. the trace lengths were still well, well below the "quarter wavelength" stuff i've read about.

is there any way of calculating the coupling based on trace length/spacing and dielectric? i don't care if it's particularly accurate, i just want to get a rough idea of what i can expect (rather than resorting to trial-and-error or just ripping off someone else's design.)

>>2948900
That’s a gross oversimplification. Tungsten is not used in any modern commercially produced tubes and hasn’t been for a long time. Tungsten seals are the least reliable unless you take a lot of special precautions, and the oxide layer between the metal and glass is water soluble and causes leaks when exposed to high humidity.
https://www.worldradiohistory.com/BOOKSHELF-ARH/Technology/Technology-Vacuum-Tube/Materials-and-Techniques-for-Electron-Tubes-Wlater-H-Kohl-1960-(657-pages).pdf

speaking of toobs and emission, what prevents making an ambient pressure "tube"? I'm guessing it's the filament burning out, but the wire used in furnaces lasts quite a long time and easily glows orange hot
another possibility we don't see often is using a flame as the "medium" in which amplification is done. I think it's possible to make something similar to gas-filled tubes with an open flame

Made this vco in falstad according to a youtube tutorial. It doesnt work. Why?

File: circuit.png (27 KB, 793x539)

27 KB PNG

>>2948934
forgot the image

>>2948935
https://www.falstad.com/circuit/e-vco.html

are you using a non-ideal op amp?

>>2948941
Yes

Any good guides for GaNsistors?

>>2948942
your input is 50V. it's probably spazzing out because that's 38V above your positive voltage rail (assuming it can even swing to +12V.)

>>2948944
I took the screenshot after trying a bunch of different voltages. It didnt work on 5V either

>>2948911
> open flame
MHD generator

>>2948885
> drift
Just use a couple of potentiometers and adjust over time like we used to do in the olden days.
Remember that guy in the lab coat and the clipboard reading dials? That’s you.

>>2948901
isnt that how the generator does it already? it has 1 wire + chassis

>>2948901
The image shows different methods for grounding a car audio amplifier, illustrating the worst, better, and best practices for making a secure electrical connection. A proper ground connection is crucial for the performance and safety of a car audio system.

Worst: A poor ground connection, often with corrosion or an unclean surface, can cause the amplifier to run hot, have poor output, or even fail.

Better: Using a clean, flat surface with multiple screws can improve the connection, but it is still not ideal.

Best: The best practice involves a clean, bare metal surface with a single, large bolt and nut to create a tight, secure connection with maximum surface area.

A bad ground wire is a major cause of amplifier problems and can lead to dangerous electrical issues.

>>2948904
At frequencies that low you can practically treat the lines as just having a lumped capacitance coupling them.
The equations are in this book 103.203.175.90:81/fdScript/RootOfEBooks/E%20Book%20collection%20-%202020%20-%20A/ECE/RF_and_Microwave_Coupled_Line_Circuits_Second_Ed.pdf
but they’re a pain in the ass because microstrip has fringing fields in regions with different dielectric constant, the coupled lines support both even and odd modes, etc.
>>2948899
Not if you learn about dielectric absorption.

Sup guys. New to all of this I don’t know much.

I’d like to attempt to attach (solder?) a Bluetooth board to my car radio that’s 30 years old. Is this doable and is it simply just finding the pins on the circuit board? Any special equipment needed? I saw a few boards that are pretty cheap but wondering if somebody has done this before I pull the trigger. Thank you.

Getting confused on this topic but a function generator is not a power supply so is the only power supply for measuring and testing simply a linear DC power supply? or at least the most commonly used kind?

I'm trying to fix a car stereo headunit right now. The volume control was causing really loud scratchy static noises and different speaker channels would cut out depending on how I touched the volume knob.

I've taken the radio out and I've sprayed the potentiometers with CRC electronic cleaner. I'll test this soon to see if it worked well but I'm also interested in lubricating the potentiometers for longevity. What should I use for this?

File: Untitled.jpg (113 KB, 1225x471)

113 KB JPG

are ring tongue terminals based or would you rather screw the conductor directly even if it's a stranded conductor?

File: ferrules.jpg (28 KB, 448x1200)

28 KB JPG

>>2949097
allow us to introduce ourselves

File: Untitled.jpg (255 KB, 1200x1200)

255 KB JPG

>>2949098
these cheap shits don't have the round hole for ferrules. can they be soldered since the circled parts in pic related more or less resemble those holes/cups that are meant for soldering?

>>2949103
Aren't you supposed to put the bare wires under the screws? It has a strain relief so there shouldn't be any problem with it come lose. Worst case, you could crimp on some ring terminals and put those under the screws.

>>2949118
the section that sticks out with a hole in it looks like a solder lug, what's up with that?
>Worst case, you could crimp on some ring terminals
that's what i was asking >>2949097
crimping is now preferred by the US military for reliability, but a good solder job used to be preferred for conductivity

>>2945584
Where the fuck do you even get a good bench supply anymore my dumb ass just got conned off a fucking jesverty chink shit off amazon that died same fucking week I got it how the fuck do you get proper tools anymore is it just digging through old technicians estate sales and relic companies selling the same products for decades?

File: 61xaBRE8VvL._AC_SL1100_.jpg (112 KB, 1087x1064)

112 KB JPG

at least some sellers state that either crimping or soldering is fine (welding is a bad translation but they mean soldering)

>>2949129
Also I'm experimenting with negative and high voltage stuff in a house with old groundless wiring I can tack stuff to plumbing to get ground though.

And bit more questions while at it
1 what happens when you try to run current through the core of an inductor?
2 what's a good piece of starting hardware for messing with FPGAs whether just emulation or something you can more easily prototype with
3 can you get negative voltage out of a raspberry pi and yes it has to be negative
4 is there a list for old computing experiments like ternary logic and svg 3d graphics I have gained a burning hate for the state of modern computing and want to see if doing it myself is possible

File: cs-tcom-sincgars-rt-1523-(...).jpg (67 KB, 625x570)

67 KB JPG

>>2949128
>US military
Having been in the US military as a 'radio mechanic' (Signal Support System Specialist) and been trained to solder I can tell you that it is 100% due to solder training being literally 2 hours long during my 6 month AIT and crimping being much more idiot proof than any issue with soldering itself. Also, generally, if something needed to be soldered as a repair we would just throw it out and get a replacement. We didn't even keep soldering equipment on hand.

>>2949192
>radio mechanic
>no soldering allowed
What does it entail then? I mean, how can a radio fail enough to warrant a person to fiddle with it, but not enough to require a soldering iron/hot air station?

>>2949129
how many volts and amps do you need? what's your use case?

File: 1.png (3.71 MB, 1144x1902)

3.71 MB PNG

>>2949198
When I was in the military it had a tiered system for repairs. Level 10 was operator level. So, a driver of a vehicle or a radio operator for a radio. Level 20 was unit level. So, a mechanic or me if it was a radio. Basically, help desk. Level 30 was something like a Forward Support Company, which was generally a specialized unit under the brigade or regiment level. 40 was 'general support/depot', which could be anything from an army depot to the original equipment manufacturer.

Even 25 years ago the insides of these radios was not something you wanted an 18 year old with 2 hours of training to take a soldering iron to. This is a single card out of a nearly 30 year old radio (TR-1523E). It is only a couple of inches across and they have a few of these in them. They are little computers and had a unit cost in the tens of thousands of dollars. Even 30 level would only replace cards. Broken PCBs went back to General Dynamics for repair.

So what did I do? Get screamed at by a bunch of sergeants and LTs because their radio wasn't working.

>We can hear other stations but they can't hear us!
Check the radio, power transmission was on low. Set to high.
>Other stations can here us but we can't hear them!
Turn volume up on speaker
>The antenna cable keeps breaking!
Instruct LT to not step on the radio when climbing out the top of a Bradley Fighting Vehicle.
>The radio works but the display is blank
Turn up brightness on display

Also, lots of installs. They would pull the radios out of vehicles when not in use because they are classified equipment (I needed security clearance just to be a 20 level support tech). Could these motherfuckers be trusted to put them back together properly? Well, it is literally just two cables: antenna and handset. So, that would be a hard no. That said, the radios weren't my biggest headache. They tended to just work when someone wasn't being a moron. Most of my time was spent on the FBCB2. Less said about it the better.

>>2949049
It could be even simpler, if your car stereo has RCA audio inputs on the back or an normal auxiliary input. But if the unit isn’t designed to take any external audio input at all, it might be quite challenging to wire up a Bluetooth receiver. As in, finding the signal wires between the CD/radio receiver and the amplifier, cutting them, and splicing in a DPDT switch and your signal. And it would be hard to ensure that the signal level would be correct.

These Bluetooth receiver units also have a power input that you’ll need to get from somewhere, that’s probably a 5V USB but if you can’t find a 5V rail you’ll want to add a voltage regulator I guess. A related issue is that of ground reference, you may find that the audio signals aren’t referenced to 0V but to some intermediate voltage for headroom on a 12V single supply. In which case, you’d need to AC-couple the output signal, or just to use an isolation transformer, or to use an isolated power converter. If I recall the DIY Perks YouTube channel made a (concrete?) Bluetooth speaker that ran into this kind of problem, and he found a fancier Bluetooth receiver that was much less noisy than the others.

>>2949050
Function generators usually have low power output capability (maybe a few watts if you’re lucky) and usually can’t output DC. Linear bench top DC CC/CV supplies (the ones with multiple transformer taps being automatically switched with relays) are the standard for electronics use because they’re high-ish power and low-noise. But these days more people are using switched-mode adjustable power supplies, I’m using an SK90W with a USB C PD trigger board because I’m a zoomer. These are fine for powering high-power loads. If you want low-noise, fixed rail supplies are usually fine.

>>2949129
Make your own. A big transformer, rectifier, and cap with a bunch of relays and comparators is pretty easy. Then it’s a few op-amps for the CC/CV linear circuit feeding a power transistor.

>>2949237
qrd me on those dsp based radios.
I recently pulled one from underneath the flood of a former german police(?) or army (circa 2006)vw sedan. Idk if it was for some cell network or intercom.
IDK about frequencies either. It had some small footprint faggot pcb antenna with some bitchass filters, it could be 433 MHz , it could be anything. Whole board was badly water-damaged with the green slime and some tracks corroded away. Even if i could connect its interface i doubt it would work.
Don't have a photo but it had TI adc/dac(?) and big deadass TI DSP chip besides some gayass makeshift arm processor, flash and ddr's for some homosexual linux clone most likely.
I'll appreciate your input.

>>2949260
>if your car stereo has RCA audio inputs on the back or an normal auxiliary input.
Old Pioneer units had AUX-in pins on the back that you could buy a simple adapter for. Dunno if anybody else did anything like that.

>>2949202
Most immediate is electrotherapy unlike most people in the field I actually read through the literature that built it (Robert Beckers "the body electric") and noted the body really fucking hates positive ions I've already gotten lasting benefits just applying that info in grounding to clear an effectively permanent stuffed nose and while it lasted going from active to passive grounding did show promise in my experiments.
I really only need something I know gives a usable negative current and is robust enough to poke around the body closest I've found so far is plasma balls which were shitty at it the power supply let me feel meridians something fierce but again it died after just giving me enough to know I damn well need a good one

You can find meridian pens online if you want to mess with this stuff yourselves

https://www.elektormagazine.com/review/zoyi-zt-703s-oscilloscope-multimeter-and-function-generator-review
https://www.youtube.com/watch?v=7h_uaO6DKzY
wtf does this oscilloscope/multimemeter combo look good? i was considering DSO154Pro for like $30 (18MHz single channel) but i'm starting to get satisfied with other things i've bought so i'm not as concerned about cheapskatemaxxing and i want a nicer multimeter too and this seems to have multimeter functions with reasonable speed and precision

File: cs-tcom-sincgars-rt-1523-(...).png (150 KB, 400x261)

150 KB PNG

>>2949267
Like I said, I was 'helpdesk' level. I never opened them up or soldered on them. If a radio was acting odd and it wasn't just some stupid operator issue then my troubleshooting was limited. We had a power output tester that could confirm it was broadcasting the wattage it was rated for. Other that that, I'd swap parts (antenna, handset), make sure the encryption fill was loaded, and confirmed it was time synced for frequency hopping and that was it. Anything else meant a trip to 30 level support. The RT-1523e was a super solid radio though. In the 4 years I delt with them I can count on one hand the number that they failed do to internal issues. My unit had 150+ of them. Generally it was "vehicle caught fire, radios in a puddle on the floor", "radio damaged by enemy fire, put all the pieces in a bag and sent to depot", or "LT still stepping on radios to climb out of the BFV, antenna connector/display damaged".

>>2949304
Punction is important anon. The content of your post already makes you sound crazy without it all appearing to be a run on sentence.

>>2949306
dear level 20 support
you claim to not want soldiers stepping on your radios yet they are built sturdy, boot-sized and positioned perfectly for stepping
curious
- Turning Point Hertz

>>2949305
Single channel really makes its usefulness pretty limited. There’s an EEVblog Forum thread about portable oscilloscopes. I bought the cheapest 2-channel unit, a 50MSps unit with an inline current measurement shunt too. I haven’t used it yet, just added it to my toolbox.

I say it’s better to have a decent multimeter separately from a scope, if anything I’d want a decent clamp meter as an all-in-one tool. With accurate DC amps and min/max option, maybe even a dual-display. I’m looking at the UT219DS for this. They also make the UT217 clamp meters with scope functionality, but it’s still a single-channel unit.

File: neg voltage generator.jpg (14 KB, 604x292)

14 KB JPG

>>2949132
>can you get negative voltage out of a raspberry pi

if you can generate a pulse waveform out of an I/O pin, you can use this old cap+diode trick
as shown it will generate -5 from +5 IN THEORY
in practice it'll be lower
but you can keep adding more caps and diodes to boost it
to 1,000,000 volts and beyond

File: asdf2.png (9 KB, 788x377)

9 KB PNG

Please forgive my extremely entry level question. I want to "gate" whether a square wave is sent to other parts of my circuit. Can I use a transistor like this to send the signal to ground when I don't want it to reach the output? (The voltmeter in this picture)
What's the "proper" way? I suppose it depends on the exact application

File: IMG_2384.jpg (110 KB, 1067x575)

110 KB JPG

>>2949398
i think your circuit technically works but for wrong the reasons. a better way would be to put the square wave (input) on the emitter, which turns it into a common-base amplifier. dropping the voltage at the base turns it off. using a BJT for this has a ton of issues so it would make more sense to use a FET. if you connect two complementary ones in parallel and drive them oppositely, now current and flow in both directions. this is how analog switches work.
>What's the "proper" way?
an analog switch/mux IC. for high power, you can use a solid state relay or an optocoupler.

File: 1757765639687382.png (23 KB, 1026x482)

23 KB PNG

>>2949398
You could try a PNP "in series" if the direction of current is strictly one-way. But resistor values need to be considered. As Anon already mentioned a FET switch is probably the best option

>>2949423
Does it make a difference if the signal will only ever be in one voltage domain? (0 to +3 V). My thinking is that the transistor just diverts the signal to ground when I don't need it.
It's a low power signal, just the output from a GPIO, but I need to control 8 such signals. So I figure it wouldn't need too many components if I can just have one transistor to gate each signal, and then have the transistor bases connected to a shift register whose 8 bit value determines which ones are on/off.

>>2949434
That makes sense to me, thanks anon. I still don't have a good intuition about where to use PNP vs. NPN but that circuit does the job.

>>2949398
>>2949436
For that kind of single-supply situation, I say that your original suggestion is best, so long as the thing that takes the square wave is fine with a highish-impedance square wave. Because when you turn it off, it will be a solidly 0V signal. The series transistor/analog-switch method is lower impedance when turned on and sinking current, but has an arbitrarily high impedance when turned off or when the square wave is logical low. Unless there’s already a pull-down resistor on whatever is taking this square wave signal, this undefined state is a poor idea. You could add an additional pull-down resistor, but it would mean more components. Ideally you’d have a push-pull output that follows the square wave when on, but just pulls low when off. A bunch of 2-input AND gates will do this, for 8 total you could use a pair of 7408 or 4081 quad AND gate ICs. Choose a logic family that suits your voltage range. The only supporting components would be a bypass cap or two, and a lot of people leave them out and it works fine.

Otherwise, NPN or NMOS transistors pulling down works fine, you could use a transistor array like the common ULN2003, though it only has 7 transistors for whatever reason. The Vce-sat of a BJT is higher than that of a FET, so you may see a small (0.1V) signal leaking. Shouldn’t be an issue if it’s digital.

File: 7-Segment_Display.png (369 KB, 1000x758)

369 KB PNG

>>2949449
>7 transistors for whatever reason