[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k] [s4s] [vip] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Board
Settings Home
/diy/ - Do It Yourself


Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File
  • Please read the Rules and FAQ before posting.
  • There are 98 posters in this thread.

05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
06/20/16New 4chan Banner Contest with a chance to win a 4chan Pass! See the contest page for details.
[Hide] [Show All]



File: 1482459142752.jpg (72 KB, 585x953)
72 KB
72 KB JPG
bump limit reached on old thread >>1251608

https://www.wiki.printf.pl/index.php?title=Pasta (fresh)

>I'm new to electronics, where to get started?
There are several good books and YouTube channels that are commonly recommended for beginners and those wanting to learn more, many with advanced techniques. The best way to get involved in electronics is just to make stuff. Don't be afraid to get your hands dirty.

>What books are there?
Beginner:
Getting Started in Electronics Forrest Mims III
Make: Electronics Charles Platt
How to Diagnose Fix Everything Electronic Michael Jay Greier

Intermediate:
All New Electronics Self-Teaching Guide: Kybett, Boysen
Practical Electronics for Inventors: Paul Scherz and Simon Monk

Advanced:
The Art of Electronics by Paul Horowitz and Winfield Hill

>What YouTube channels are there?
mjlorton
paceworldwide
eevblog
EcProjects
greatscottlab
AfroTechMods
Photonvids
sdgelectronics
TheSignalPathBlog

>What websites feature electronics projects or ideas?
http://adafruit.com
http://instructables.com/tag/type-id/category-technology/
http://makezine.com/category/electronics/

>Where do I get components and lab equipment from?
ladyada.net/library/procure/hobbyist.html
Search the web for "hobbyist electronics sources" to find plenty.
In the US and elsewhere, mouser.com, digikey.com, arrow.com, newark.com are full-line distributors that entertain small orders.

>What circuit sim software do you use?
This mostly comes down to personal preference. These are the most common ones though:
NI Multisim
LTSpice
CircuitLab
iCircuit for Macs
CircuitJS (quick, dirty, interactive)

>What software should I use to layout boards?
Circuit Wizard
ExpressPCB
EAGLE
KiCad

>Q1. Lithium batteries
Will explode if you abuse them. Read and understand all relevant datasheets and be prepared for catastrophe. See
batteryuniversity.com/learn/
>Q2. My circuit doesn't work. Halp?
Check wiring, soldering, part pinouts, and board artwork if applicable, then post schematic.
>>
File: 1480986287559.jpg (25 KB, 300x241)
25 KB
25 KB JPG
This thread brought to you in part by the NE544 servo amplifier.
>>
File: 1499346602285.png (230 KB, 472x460)
230 KB
230 KB PNG
... and by the MCP6544 quad comparator.
>>
Started a electronics and computer engineering program this September.
Electronics is fun /g/!
>>
>>1256470
>the only voltage reference I saw in store was a 2.5V diode and that was too low a voltage
How much precision (not accuracy) do you really need? Consider that you're probably using 5% resistors throughout the circuit and that you'll have temperature coefficients throughout the circuit to deal with anyway.
>Do not tie
Right, ±9V means V+ - V- > 18V.
>>
>>1256556
Well in the end I'll probably be fine with the (3.2V to 4.2V)*3 from the lithium cell, provided other drains on the system don't drop the battery voltage too suddenly. This is but a prototype. Making the thing on perfboard feels a bit stupid now, but so would putting 400V through a breadboard.
>>
File: mosfet.png (793 KB, 569x801)
793 KB
793 KB PNG
>>1256209
>>1256269
Well I can't get a circuit diagram but I had a little session with the multimeter. In the pic you see the footprints of two identical mosfets probably a totempole configuration to drive the transformer.
The top one reads .3Ohm across source and drain. .2 if I switch polarity.
If the lower one activated while the top one is failed the resulting short circuit might be what blew the bead, as it goes right across that.
Should I desolder the top mosfet and check it out of circuit to verrify?
And if that is indeed what happened is there a fair chance that replacing the mosfet and the bead will be enough to get it working again?
>>
>>1256544
What's the best circuit simulator?
>>
>>1256667
Mostly comes down to personal preference. Try them all.
>>
>>1256667
>>1256671
Free ones first, of course.
>>
File: bugs.jpg (825 KB, 1920x1080)
825 KB
825 KB JPG
my small bug collection.
>>
File: memories.jpg (766 KB, 1080x1920)
766 KB
766 KB JPG
>>1256715
Yes, I'm an oldfag.
>>
File: nonpotatocam.jpg (988 KB, 1080x1920)
988 KB
988 KB JPG
>>1256717
oldfag with potatocam it seems.
>>
>finished my quadricopter motherboard
>still have to wait for the frame to be 3d printed

in my boredom I wrote a vector library for arduino
>inb4 it already exists
>>
>>1256718

Can you even get schematics for that still?
>>
>>1256746
Yep.
I have a big ass book somewhere in the pile. But that schematic is still online. datasheet4u.com has it.
>>
>>1256667
Most Spice ones will quite often fail to converge unless you understand how they work at the most basic level and tweak accordingly.

Simetrix is one of the more forgiving ones IME, but only free for small circuits and simple models.
>>
>>1256667
pen and paper :^^^^)>>1256667
>>
I'm following an instructables guide that unfortunately doesn't go into much detail and I'm retarded so any advice you guys have would be appreciated

According to the guide (http://www.instructables.com/id/Arduino-Based-Stranger-Things-Lights/), I'm supposed to connect the "green to pin 6" and "blue to ground" on my arduino, but the wires on the actual string of lights are all the same color.

The loose end of the string of lights did have some stickers attached to lose wires (one for +, -, and "data" whatever that means) but I'm not really sure what to do here
>>
>>1256667
It really depends on what you want, if you need something to simulate a circuit and you want accurate numbers but it is not quite as useful for beginners, If you just want a simulator to tinker around with electronics then you can just use tinkercad circuits
>>
File: BCP.jpg (2.15 MB, 4032x3024)
2.15 MB
2.15 MB JPG
Working on my boiler pump and air handling control panel. Just got components mounted so far and made a few cables up to try out my new crimpers. Will end up controlling 5 pumps, and my air handling unit with about 7 temperature/humidity sensors mounted around the house. The memeduiono talks to the tablet via OPC over USB, tablet will get a 3-D printed flush mount upstairs if I ever get around to it.
Tested a single sensor and coms are working I can watch readings on the display. Lots to do yet but getting closer.
>>
File: s-l1600[1].jpg (167 KB, 1600x1600)
167 KB
167 KB JPG
>>1256544
quick question from a beginner
do the number in the first line mean anything? trying to repair something and one of the transistors's legs are pretty corroded, so I want to replace it, but I'm confused

pic is from ebay of what I think is the same part, but the top line of letters/numbers are different than the part to be replaced

googled the spec sheet and it had no mention of this line: http://www.ti.com/lit/ds/symlink/tl431a.pdf

thanks in advance
>>
>>1256544
>Getting Started in Electronics Forrest Mims III
>Make: Electronics Charles Platt
>How to Diagnose Fix Everything Electronic Michael Jay Greier
so which of these/other beginner books not mentioned is the best?
>>
File: 148404708443.jpg (73 KB, 640x463)
73 KB
73 KB JPG
I need help designing a comparator that that tell me if a number A is bigger than a number B, but each number have 6 bit long, and I need to do it with logic gates. I tried to make a karnaugh map or a true table, but, yeah, there are 2e11 possible combinations, its hard to do.
Question: Do anyone know or have a software where I can do that? Or, any menmotecnic rule, or anything that could help me to do this shit?
>>
>>1256914

not wise to be using an arduino in any kind of serious application like that, coz you normally expect at least 10 years reliable service out of a control panel like that. but those PCBs could be made obsolete tomorrow; the connectors are weak and prone to oxidation; the arduino is liable to go nuts and run amok under brown-out conditions (or spikes from motors or lighting); the USB is probably chinese and could die at any moment; the switching power supply is the same. and if you're not around, who's gonna service that custom-built shit.
>>
How does one select a specific op-amp or transistor? There are so many that have virtually the same core characteristics (Uf, beta, etc.). What does it come down to? Is there an actual selection method, or it's just about personal experience / What you currently have in your components box? What about in a less DIY and more professional situation?
Particularly in regards to audio electronics.
>>
File: full subtractor.gif (24 KB, 678x542)
24 KB
24 KB GIF
>>1256933

it's been a million years since i did these stupid theoretical questions, but it seems to me one approach would be to use 6 full subtractors, then determine which is greater by looking at the sign bit. youtube has tuts on adders/subtractors, like this one: https://www.youtube.com/watch?v=JSa6BtQRjkI
>>
>>1256949

the engineer who makes the most profits rises fastest. so, you pick the cheapest good-enough part for the job, and await your next promotion.
>>
>>1256918
It is probably just a serial number or something like what factory it is from, I am pretty sure you don't have to worry about it.
>>
>>1256949
If you see a deal with 100 opamps for $5, just go for it, they basically all work the same, it is often nice to look at datasheets though you might for example find out that the max input voltage is lower than that of your circuit. Just buy the cheap once and if you want to do something special, just Google for am opamp with the specifications you want. Places like digikey are also pretty nice to find specific components.
>>
File: file.png (42 KB, 1268x726)
42 KB
42 KB PNG
I'm putting together a high current linear supply and I want to cut my bridge losses so I'm using ideal diodes. Can anyone think of any edge cases that would compromise the pictured circuit? The intent is to have four of those in a bridge. It runs off a bootstrap that's filled on the reverse bias period, and uses an op amp as a comparator because 358s are cute.
>>
>>1256937
meh, if it shits the bed it shits the bed, I've got spare parts and I built it to be able to swap shit out and fix it fast and easy. If I'm not around that means I sold the house and don't give a crap.

It can't really do much damage as the boiler is self contained and only operated by a set of dry contacts. Worst it can do is run my loop pumps and air exchanger.
>>
Has anyone here ordered from arrow.com? How was it?
>>
>>1257011
>if I'm not around that means I sold the house and don't give a crap

you'll care if the buyer lowers their bid by $10K to pay for replacing the toy-grade temperature control system.
>>
>>1256918
>(4) There may be additional marking, which relates to the logo, the lot trace code information, or the environmental category on the device.
>(5) Multiple Device Markings will be inside parentheses. Only one Device Marking contained in parentheses and separated by a "~" will appear on a device. If a line is indented then it is a continuation of the previous line and the two combined represent the entire Device Marking for that device.

>>1256937
>but those PCBs could be made obsolete tomorrow
bruh, they were obsolete the day they were born

>>1257011
if you sell the shithole, consider some industrial-strength lolduino board with screw connectors so it looks all real and stuff

>>1256949
You figure out what parameters are important for the application and, as the other anon said, pick the cheapest good-enough part, which, in a hobbyist or even professional situation, often does depend strongly on what's in house stock, then tweak the design to fit if needed.
For op-amps, supply voltage, THD, gain-bandwidth product all generally matter. For some applications input offset currents are important, for others, don't care. And so on.

>>1256933
Divide the problem up into six stages. Start by comparing A(n) and B(n) to get greater(n) and less(n). Add a second stage that compares A(n-1) and B(n-1) the same way but also depends on greater(n) and less(n). Duplicate this second stage four more times.
>>
>>1257032
>10k
>for the 2 hours and $100 worth of parts to install some shitty thermostats and relays
It’s a hobby project man, lighten up
>>
>>1256667
LTSpice is very use to use and mathematically one of the better ones out there.
>>
>>1257071
easy to use*

>>1256870
Check the datasheet of the lights you got to figure out which wire is which.
>>
>>1256964
ok, thanks
>>1257034
where did you get that info? was it in the datasheet and I missed it?
>>
>>1257103
hidden in the footnotes in the ordering information table, but common practice in the industry. ctrl + f "marking" found it
>>
What's the quickest way to understand opamp theory?
>>
>>1257140
Ignore the wierd music, this is actually a great explanation. https://youtu.be/_o4ScgRZtNI
>>
>>1257140
i think it's a bad idea to try to grasp "op amp theory" as a whole. the generic op amp circuits like differential amps are actually pretty clever (viewed without hindsight bias) so you're not just going to derive them by yourself given enough theoretical knowledge. imo it's much better to instead get a grasp on when they're useful, and then copy the generic circuit that matches your use case. after doing that a few times it'll click anyway.
>>
File: 1482758192789.jpg (1.82 MB, 4608x2592)
1.82 MB
1.82 MB JPG
>>1257140
What >>1257153 said. Also read some of the old National Semiconductor (now TI) app notes on op amps and anything that presumes to call itself an "op-amp cookbook".

>>1256715
berry gud
>SN76477
dart machine pulls?
>>
>>1257140
https://youtu.be/QLQrLO0zvDI
Really good video explaining opamp oscillators. I agree with >>1257153, but it can be a bit of a handful if you don't have some decent learning resources.
>>
File: kemobr-m204br-230-v.jpg (31 KB, 520x489)
31 KB
31 KB JPG
I have a huge electric heater unit, 5kW id like to use to heat something to a specific temperature. But at full blast it is too much, would a pic related heater controller do the trick? Its rated for 16A, the heater pulls 20A, but i would probably only control the heater around 10A, could this work? Wouldnt the current difference be a problem?

Also id like to do a PID control, can i replace the pot with a digital potentiometer? I wouldnt think the pot handles too much current, but it might be live or even at mains voltage, not sure. Worst case id use a servo to turn the knob.
>>
>>1257185
Assuming that is a normal PWM device, you will be pulling 20A max (actually 28A peak I guess), though this shouldn't matter as what causes fuses and FETs and the rest of the electronics to blow up is because of heat buildup. Since the heat buildup on each 60Hz cycle won't be significant (it could take up to several seconds to blow the 16A fuse at 20A), the average current being limited to 10A fairly linearly corresponds to the average heat produced being limited to what it would be if you were running a 10A load at maximum duty cycle. If you want to prevent the knob from being turned too far to the right, you could probably add a resistor in series with one of the potentiometer legs.
>>
>>1257185
Solid state relay, thermocouple, opamp?
>>
File: 1504192369706.jpg (51 KB, 582x457)
51 KB
51 KB JPG
>>1257185
No way to know whether a digital pot would work without opening it up or testing the control voltage the control circuit passes through it. If it's like Pic related, you're fucked.
>>
>>1257185
It does not split phases it skips one if needed. From what you are saying it wont work, it will pull max current in the enabled phases, blowing a fuse or the triac inside.

>>1257193
SSRs are shit in power control from what i heard, only turning stuff on and off (not rapidly) what they do well

>>1257195
Looking at that i feel its a no go, thanks.
>>
>>1257197
SSR are typically used for PID control of electric heating elements, the controller output is a percentage of a time index of something like 5 seconds, so the relay is only turning on and off once every 5 seconds. You can get away with this on heating applications because the control loops are extremely slow(usually)
>>
>>1257252
Okay then.
So they are not susceptible to "stuck on" when driven at a slow switching period? Idk where i heard this...

This might be okay since i already have a temperature controller but didnt have the switching mechanism to control.
>>
>>1257185
This is what you want.
>>
>>1257270
>So they are not susceptible to "stuck on" when driven at a slow switching period?
If anything I would think mechanical relays are more susceptible to that failure mode, by way of fused contacts.
>>
>>1257512
>>1257597
Right, ill go for an SSR then.
>>
File: arm-comparison.png (380 KB, 1017x437)
380 KB
380 KB PNG
So I've been wanting to try my hand at making a robot arm with some servos and an arduino for a while now, but I'm a bit confused by the many different designs.
Pic related, I'm guessing the metallic one would be a bit sturdier and it looks simpler to assemble, but would the black one be able to lift heavier stuff with the "reinforcement rod" or whatever that thing sticking out behind it is?
>>
>>1257614
the left one is a lot stronger
look at the right one, it has blue servos. those are sg90s, and are tiny. the left one has mg995 servos, which are bigger and stronger
>>
>>1257614
not electronics problem
lern2mechanical
>>
>>1257619
Robot arms are mechatronics, they qualify.
>>
>>1257614
Those bars on the right side look like mechanical links. One of them controls the angle of the forearm, the other controls the angle of the 'hand.'
>>
>>1257173
>>SN76477
>dart machine pulls?
Nope, I bought that at a local chip shop decades ago. It was one of my first builds from an old Popular Electronics. Still have a box full of those zines. I don't have many pulls except for FETs
>>
>>1257614
Depends on where you want the servos anon.
The right arm will be lighter but lack the dexterity of the left. Figure the right one is more like a backhoe in range of motion.
>>
>>1257660
Ah, cool. I used to live near a salvage yard that took in lots and lots of computer, typesetting, medical, and other such equipment. I could occasionally find boards half a meter on each side just packed with 74xx logic of almost every flavor, and pull chips off of them for $5/lb.
>popular electronics
ayyyyy oldfag
I remember the pretty glass piggy with the waveforms on the front, illustrating a simple digital storage attachment for an o-scope, was what originally goaded me into electronics in the first place. It wasn't until late in high school I found a mentor to actually teach me the finer points of circuit analysis and transistors.
>>
>>1257662
>>1257652
Oh, alright, I see that now. Would either of them be stronger than the other? If I wanted to, say, lift a can of soda, would that be too much for the right one?
>>1257617
Are you sure? Looks like only the servo at the claw part is blue on the right one, it has black ones like the other at the base.
>>
>>1257665
It should be easier to cobble together a solution to a new problem for the right side. Too much weight on the left means get beefier servos. Too much weight on the right could use a small gear driving a large gear to increase the torque for one of the arm segments. If you already have a 3d printer, you can dick around with gears for the cost of filament
>>
>>1257668
You don't actually need beefier servos, just to change the gearbox in the servos to a higher ratio. These servos were originally made to drive remote controlled cars and planes, they're made to be pretty quick and to shift things that aren't that heavy.
>>
>>1257668
I don't have a 3D printer, but since it doesn't seem like the right one is notably stronger than the left, I'm probably buying one of the left. They're a lot cheaper, if you compare. Arms like the right one for the cost of the left one look super flimsy, and they're cheap enough that it's not the end of the world if the arm I buy turns out to be shit anyway. I'll learn a lot either way.
For controlling it, people seem to either be using pots on breadboards, or playstation controllers. Am I right in assuming that it would be easier to control by making a coordinate system on the arduino, where the x,y,z and clamp rotation position are set by moving the joysticks, than by using pots to control each joint directly? I'm assuming the arduino is powerful enough to save a dozen or so sets of cartesian coordinates and convert them into angles for the servos? I was hoping something like right joystick for x and y, left for z and clamp rotation, with x to save a position and b to loop through the positions I've saved, it would be a lot like the robot arms in Take On Mars that way, and those are pretty intuitive.
>>
>>1257673
Forgive me, the last servo i touched was from a vex robotics kit in high school.
>>
>>1257732
It's alright. It's in my mind because I recently spent an absurd amount on an overpowered motor for a project, only to have my manager shout my ear off because I never considered a gearbox.
>>
>>1257711
>Am I right in assuming that it would be easier to control by making a coordinate system on the arduino
Try it. It's just software.
>>
guys, can someone explain to me when I should use ceramic and when I should use electrolytic caps, please. I understand that ceramics occupy lower value ranges and electrolytics higher values ranges.

but is there some kind of a rule of thumb when picking caps in the middle?

thanks!
>>
>>1257818
the rule of thumb is that ceramics are always the correct choice unless overruled by price, size, or availability. electrolytics are technically inferior other than in energy density.
>>
>>1257819
Thanks anon! Appreciate it.
>>
>>1257819
That's a decent rule of thumb, but I'd like to remind that high value ceramics can have absolutely horrible tolerances, tempco, voltage coefficient and microphonicity. So you might still end up with electrolytics if you want some sort of stability.
>>
File: electrometer 1.png (3 KB, 640x400)
3 KB
3 KB PNG
I want to quantitatively measure electric charge building up on a metal surface from impacting ions/ionising radiation, how do? I assume that using a capacitor to ground (whatever ground is in this context) and measuring the voltage across it is some way of measuring the surface's charge.

Ideally I'd use something similar to a Kolbe electroscope, since it measures absolute charge instead of relative charge and lets essentially no current through, though I'm not sure how it stores its energy. Capacitance to the atmosphere? Or is it all potential energy of the needle?

For a solid state method I assume I'd need a very high input impedance and perhaps some way of measuring a log scale to have a similar relative accuracy at low and high voltages, since I could be measuring the charge building up over a few hours at a time. Perhaps pic related? If changing the capacitance changes the voltage at the FET gate through V=Q/C, will that also change the electric field emitted by the surface and potentially deter more ions from impacting differently? I can't see how that would happen since the device's net charge will be the same, leading back to the question of what the voltage is relative to.
>>
>>1257885
The usual approach (when accuracy is needed) is to use either very sensitive current meters or charge amplifiers. Both keep their input terminal potential constant. Of these, the latter can be made more sensitive.

For simple, easy to build circuits google diy ionization chamber electronics. They typically resemble either your circuit (fet with a floating gate) or measure current by adding a bleeder resistor.
>>
>>1257885
A particle striking the plate that can eject an electron will charge the plate relative to the ground, which you can measure if your equipment is sensitive enough. However you will need an amp between the plate and meter.
>>
>>1257907
Actually they all seem to use Darlingtons, but I see where you're coming from. I need the input charge to stay constant, but I don't have much idea what kind of charge build-up to expect. Hopefully I can keep the voltage below a certain value to not arc my circuit to pieces, so I'll probably want to go oversize on the capacitor. Both those circuits require a current to flow, which I would rather avoid because I don't know where it's flowing. Maybe I can just use a non-inverting amp if its input current is low enough, else make some sort of bang-bang controller with a couple of FETs as a makeshift comparator for all that insulated gate goodness.

>>1257909
That sounds like the photoelectric effect; to measure the charge separation generated when a high energy photon impacts the positive plate. As far as I know, no electron would be emitted if a hydrogen or helium nucleus or electron hit either plate.
>>
>>1257940
>I don't have much idea what kind of charge build-up to expect.
You haven't said anything about your radiation source or operating conditions. For not-immediately-dangerous radiation sources, the currents are really tiny and usually the potential shift of the meter doesn't matter. If it matters, ground it or connect it to your radiation source or do whatever to eliminate the problem.
>>
>>1257951
Low Earth orbit radiation, with an hour or so of exposure. The sensor is for a cubesat-style setup, so I can hardly strap a mechanical electroscope to it.

By tiny do you mean greater or lesser than the nA range?
>>
File: 4-Way-Grounding-Rod.jpg (26 KB, 640x480)
26 KB
26 KB JPG
>>1257885
Many modern multimeters have an open input at the most sensitive DC voltage range. The only termination is the internal ESD protection circuit. Use range-hold to lock it. Pic shows your zero reference.
>>
>>1257965
Uh. Is this that meme frog satellite?

For reference, the ion current seen by the electronics of a radioactive smoke detector is around 100fA, and that includes ion chamber's gain. Source is 40kBq or less.
Your satellite floats in any case, so if the collector plate's voltage has any effect on your readings, then the charge of your satellite is even bigger problem.
>>
>>1257976
>Pic shows your zero reference.
Not going to work at a few dozen kilometres. Not sure if the ADC will have a low enough input impedance.

>Is this that meme frog satellite?
Maybe...
That current reference should be higher than anything I'm likely to see, from what I hear about aircraft radiation levels, we shouldn't get more than 1000 times background, and 40kBq is pretty high comparatively. The point everything is relative to becomes the disperse matter surrounding the device, meaning it will become a small relative positive charge in a sea of electrons, since electrons have a shorter mean-free path than ions. But charge is still an absolute measurement in comparison to voltage, as a gold-leaf electroscope demonstrates, so what voltage my ground is at doesn't imply that this should equal a zero reading on my electrometer. Ideally I keep a reservoir of 0 charge in the centre of the payload and don't sink any external current into it to serve as my voltage reference, but that shouldn't be necessary if I find a way to measure charge directly, for which only mechanical devices come to mind. Maybe something akin to a hall effect sensor could work; detecting a voltage gradient in the presence of an electric field created by the incoming charge or something. The voltage could get up to 200V or so going off 840km orbit estimates for day time, so normal ADCs won't serve me there, and neither will voltage dividers, because they require current to flow to ground. If there's some way to ensure a linear output of a FET while retaining its insulated gate that would be a solution, but it's tough to avoid the threshold voltage problem.
>>
>>1258008
*2000V
>>
I build this circuit and my output signal should be a square wave with a 50% duty cycle but according to my scope I have a 58% duty cycle. (I am not using the exact same cap and resistor values as in pic related). I think it might be related to the fact that I am building the circuit on a breadboard but I don't really see how that would be the problem in this case.
>>
>>1258008
If you can force more charge in, then the existing charge and potential do not matter. On the other hand, if your ions have low energies, then your charged electroscope, satellite or whatnot might start repelling them at some point.
Current meters and charge amplifiers on a satellite would basically measure the movement of charges through the meter to the rest of the satellite. A capacitor + voltage amplifier would measure the accumulation of charges in comparison to the rest of the satellite body. All these can measure charge just fine, as long as you can make the charges enter.
If you're dead set on mechanical devices, then quartz fiber electrometer can be made to work at relatively low voltages.
>The voltage could get up to 200V or so going off
The current meter wouldn't have that problem to begin with and the other two just need sufficiently large capacitors to prevent voltage from soaring.

>>1258028
300 ohms is pretty small. Also, the CMOS version of 555 gets closer to 50%.
>>
>>1258033
not I am wondering why a different version of the 555 would be different. does this have something to do with internal resistance of the 555 being different when output is heigh than when low. which does seem to be indicated by the fact that the bottom of the square wave is at 0v and the top is not at vcc.
I don't really see why that 300 ohm resistor matters. in any way.
>>
>>1258037
*now
>>
>>1258037
The capacitor is charged from 555's output voltage and if the output goes closer to 0V than VCC, then the charging speeds of low/high states differ. CMOS has more symmetrical output swing.
If the resistor is overly small, then the capacitor charging current causes a noticeable voltage drop in 555's output stage. CMOS outputs typically have different output resistances when outputting high/low, causing the same effect as before.
>>
File: ferite bead.png (2.03 MB, 1222x719)
2.03 MB
2.03 MB PNG
>>1256632
So I checked the mosfets out of circuit.
one of them reads 0V on diode check between drain and source and the other actually has a scortch mark on the bottom.
I beleve this resulted in a short circuit that blew up the ferrite bead in pic related.

But how do I know if this is the cause and not just another symtom of something else gone wrong? Also how likely is it to work if I replace the mosfets and the ferrite bead?
>>
>>1258130
Replace the broken mosfets, power the thing up, and see if it works.
>>
Is it worth it to invest in an FPGA
>>
>>1258134
How do I find compatible mosfets?
The ones used were
http://www.datasheetlib.com/datasheet/200528/2sk2876-01mr_fuji-electric.html
What values do I have to pay attention to when searching for replacements?
>>
>>1258135
Do you have projects requiring FPGAs or does your employer use them?

>>1258149
Start from the values shown in the title: 500V 6A N-FET, 1.5 ohm RDSon. Try to find one in the same package.
You can exceed the voltage and current ratings and you can have lower RDSon, but try to not exceed the indicated capacitances a lot.
>>
>>1258156
Neither, I was thinking as something for prototyping and hobby development
My employer is completely software focused, but I've always been an electrical engineer at heart
>>
>>1258160
>prototyping and hobby development
Well, we get once in a while guys wanting to learn FPGAs. They ask what development board they should get. Then we never hear about them again.
Even if FPGAs are general purpose chips in theory, they aren't that in practice. Other options tend to be more convenient, cheaper, smaller and consume less power.
Consider again when you have found some actual reason for using them.
>>
>>1258179
Maybe we never hear from them again because FPGAs are so easy to use that they don't have any more questions
>>
>>1258201
Judging by the kind of shit on here I fucking doubt it lol. fpga isn't exactly code monkey programming, it's not difficult but its certainly different if go in expecting some procedural structure.
>>
>>1258028
Tell your video game to learn E-series values and to read data sheets. Both versions of the 555 have been optimised for TTL, sink is better than source. There is a simple way to adjust to exactly 50% though. Without that just divide output by two.
>>
>tfw you're too dumb to be smart
>>
>>1258205
well it's made to emulate more complex hardware to figure out optimal structure before you get boards fabricated
>>
>>1258205
Who needs programming when design entry can be accomplished through schematic capture? FPGAs really are that easy to use, once set up. Just not all that generically useful.

>>1258227
Sure, but there are situations where they shine, such as when you need any two of more, smarter or faster I/O than that lolduino or beaglebone can give you. Hard IP for high-speed USB or PCI-Express can be really nice.
>before you get your boards fabricated
I'm sure you meant ASICs, no boards, but alright.
>tfw the programmable logic + footprint is cheaper than discrete logic + footprint
>>
>>1258239
I've used CPLDs before just because they were cheaper than a couple of 74 series chips.
>>
>>1258135
Depends on what you want to do.
If you're developing asics, custom hardware, codecs or need something done faster than a cpu, dsp, or gpu, or doing real high speed communications, sure.
I'll be perfectly candid here and admit that I have an altera fpga dev board that suffered the same fate as hundreds of thousands of tarduinos: collecting dust on a shelf. I played with the fpga a bit and made some very basic circuits. I also toyed with the idea of using it to accelerate some heavy-lifting data tasks at my job (software and data), realized life is too short, then left it alone. Further, I didn't want to provide support for a niche piece of custom firmware for the rest of my life.
>>
I'm using my arduino to do pwm. Each loop in my main program calls the "analogWrite" function to set the pwm, but is it bad to call the analogWrite function so often? I only want to actually change the pwm every tenth cycle or so, does the constant calling of analogWrite have any effect on the output pulse? Like, if it's in the middle of doing a pulse when the new call comes, does it then change the pulse it's sending out? Like if it was doing low for 10 micros, and then high for 10 micros, if a new call (with the same pwm) comes in at the 7th high micro, would it go directly to low or would it finish the 10 high micros first?
>>
>>1258307
No. The arduino uses an output compare register, what analogWrite does is change the value in the compare register. You then have a separate hardware running a timer, which is compared against the value in the compare register. To use your example, if at micro 7 you change the value to 10, it will continue the current pulse until 10. If you change it to 7 or less, it will change over to low immediately. Redundant calls to analogWrite are a waste of cycles, but don't affect the output pwm.
>>
>>1258033
From what some searching on the internet gives me, there's a constant sea of negative potential out there, and only if you're at a lower potential than it will you collect electron charges. Once you reach just above that potential then electrons will be repelled by it before they get to the craft.

If I measure the plasma potential by using a low-capacity high-voltage capacitor that will charge up to max voltage in no time at all, probably just the capacitance of the craft itself, then I'll have values for the voltage of the inside and outside of the craft assuming I can measure it with a high enough input impedance and not blow up the ADC with 2kV.

The reason I don't want to allow a current to flow through to the main body of the craft is because I don't want to charge up the electronics to significantly different potentials to their surrounding insulators. I'd have to sink this current to another capacitor instead, and it would have to be large enough to hold the charge accumulated for the full duration. Using the 100fA current as an example, I'd need to hold 720E-12 coulombs, in comparison to the hundreds that standard super capacitors can hold, so this should be fine. But as I sink this current, the potential of the outside effectively remains at 0 net charge, which comes from storing charge in a high-capacity super capacitor instead of in the stray capacitance of the outside shell, and so won't reflect the actual accumulation of charge on insulated parts. If I feed this charge straight into the capacitor and simply measure its voltage, then I can calculate charge flow/ion current from C•dV/dt, and calculate the electron flux from that, which would be pretty easy if I had high enough input impedance.

I am somewhat worried that the antenna won't react nicely to having 2kV on it, but an isolation transformer should solve that problem.

I actually want to avoid using a mechanical electroscope because of vibrations, otherwise I'd just film one.
>>
>>1258215
>tfw too pathetic to live up to my potential
>>
What's hot regarding EE R&D?
Where could I inform myself about this (besides jewgling) so I don't bother with this kind of silly question?

Also reading the "getting started in electronics" book, I found it funny that the author made it by hand, pretty cool. ¿Should I complement with another book of the beginner list? Or it's preferable to jump onto the upper echelon?

t. soon to be EE freshmen scrub.
>>
File: mpv-shot0072.jpg (37 KB, 640x360)
37 KB
37 KB JPG
So I was watching Dave's old video on a diy electric load. But I don't understand why he's adding an opamp for the ADC here instead of just a simple voltage divider.
>>
>>1258533
How are you going to set up a voltage divider to get gain?
>>
>>1258535
Oh right, the voltage there is fairly small.
My bad.
>>
File: DSC_0002.jpg (914 KB, 3159x2290)
914 KB
914 KB JPG
How can I test to see if I have a bad cell in this power bank? They are all in parallel and spot welded together.

Current voltage is 2.9v and it has trouble charging and powering my phone sometimes (current going in and out).
>>
>>1258547
Also ignore the missing positive wire to the circuit board, I removed it for troubleshooting
>>
>>1258547
>see if I have a bad cell
You can't, but you can charge by other means to see if you get 4 volt or so and then discharge to gauge the capacity. Maybe all 5 are bad.
>>
>>1258586
Could a bad cell get hotter than the others?

What would be another way of charging them? I'm not geared up for these new age batteries, would a power supply and resistor work?
>>
>>1258412
Make: electronics is pretty practical if you want to get your hands dirty right away.
If you're about to start EE you might want to have a look at practical electronics already. It's a good mix of being easy to read and explaining fundamental concepts
>>
>>1258156
How important is the maximum Gate-Source-Voltage? Most I see are ±25V but my datasheet says ±35V
>>
>>1258638
As long as its the same or higher than it would be fine, don't go lower as the gate drive circuit could have ringing which could exceed that.
>>
Any problems if I use a 2mm thick wire for wiring these little things? And in general: When is a wire too thick for the job and what problems does it cause?
>>
>>1258646
You forgot to include a pic you silly goofball.
>>
>>1258587
>power supply and resistor?
Yes, plus multimeter plus you _must_ monitor the process. 5V 2A supply, 1Ω 5W resistor in series. Observe the current (via voltage across the resistor) and the voltage across the cells. Stop when cells are charged to 4V. This setup would temporarily replace a real (automatic) lithium battery charger. If you get more than the 2.9V you have now and the voltage does not decrease on its own (without load), the charger in your power brick may be faulty.
>>
File: sku_312253_1.jpg (34 KB, 600x600)
34 KB
34 KB JPG
Will 2mm thick wire be too thick for wiring these little switches? Also in general: When is a wire too thick for the job and what problems does it cause?
>>
>>1258671
>wire too thick
when your soldering iron is too weak
>>
>>1258674
>tfw your soldering iron can't handle 4" billet
>>
>>1258671
2mm? = 12ga? why so thick?

just use crimp quick connects. those micros can be flaky and it'll be easier to replace them if you dont need to resolder every time.
>>
>>1258671
too thick and you suffer losses to ac skin effect*
(*note: this will never, ever happen to you)
>>
>>1258643
DigiKey doesn't seem to carry anything with ±35V or more and the other caracteristcs I need. Am I fucked or is ±30V worth a try?
>>
>>1258684
Send me a datasheet to the part in question, you could use a lower rated part just fine probably but just check the peak gate voltage in case they are using the ±35V FET to fend against a spiky gate drive.
>>
>>1258691
Alright, so I'm considering
replacing: http://www.datasheetlib.com/datasheet/200528/2sk2876-01mr_fuji-electric.html
with: http://aosmd.com/res/data_sheets/AOT11S60.pdf
Thats the closet I found so far.
>>
File: SS-5GL.png (32 KB, 521x334)
32 KB
32 KB PNG
>>1258671
RTFD
>>
File: zero.jpg (3 KB, 106x172)
3 KB
3 KB JPG
>>1258537
>the voltage there is fairly small
>>
Anything interesting to get from Arrow that you can't easily get in Europe?
Want to get the most out of that free international shipping offer.
>>
>>1258720
Arrow as in Arrow Electronics? They have offices in many European countries.
https://www.arrow.com/en/support/contact-support/find-an-arrow-office
>>
>>1258736
Yeah them.
They usually charge something like 50$ to ship here.
>>
what is this large yellow disc thing in the middle?

pic related. its a opened up kWh meter next to the fusebox.
>>
>>1258747
a backup coin cell or supercapacitor
>>
>>1258747
It stopped metering. How did you do it?
>>
File: pyhä jysäys 001.jpg (384 KB, 1067x1600)
384 KB
384 KB JPG
>>1258761
>>
File: lump.png (10 KB, 945x91)
10 KB
10 KB PNG
>>1258762
neat
>>
>>1257885

For Faraday cups this is usually done with an integrator instead of a direct capacitor to ground, presumably so the increasing voltage of the surface doesn't mess things up.

http://www.keysight.com/main/editorial.jspx?ckey=2682025

The LMC662 is the standard cheap electrometer grade opamp.
>>
File: CEamp.png (14 KB, 763x528)
14 KB
14 KB PNG
How do I go about designing a circuit to have a specific input/output impedance? Like if I had an amplifier that I wanted to have a 50Ω output?

Take this common emitter amplifier for instance, with an open circuit voltage gain of ~22 at 1kHz. We know we can calculate the output impedance of this circuit, it's Zout = Rc||RL and if RL is high impedance then it's effectively Zout = Rc which in this case would be 6kΩ.

How do I go about making this 50Ω output impedance (flat 50Ω from DC up to 1MHz or so) without significantly altering the gain, frequency response, or any other major properties of the amplifier?
>>
>>1258856

R3 would obviously have to be 50 Ohm. To get 25x amplification the AC emitter resistance would have to be 2 Ohm.

The internal emitter resistance Ree=25mA/Ie ... this would have to be <<2 Ohm if you want to see any sort of linearity, so you'd be looking at quite a large bias current.

Lets make the bias current 50 mA, so make R5 100 Ohm and R4 1.5 Ohm. Decrease R1 and R2 to 10k and 1k. Increase C2 and C3 to 100 uF and 1 mF.

PS. for high linearity this isn't a good amplifier with low output impedance.
>>
>>1258870

I meant to say decrease R1 and R2 to 1K. No longer a need to care about headroom voltage of the collector, the dynamic range is tiny any way.
>>
>>1258871
C1 has to increase to 100 uF too.
>>
>>1258870
Well is there some other way? Can I set a fixed output impedance with buffers or somehow use feedback?

Like, how does an instrument with a 50Ω output impedance like a signal generator accomplish that? Surely they don't simply have a buffer and 50Ω resistor tacked on to their output?
>>
>>1258873
A low'ish frequency signal generator (ie. less than a GHz) will have a opamp with a 50 Ohm resistor tacked on.
>>
I ordered 60 $0.03 2.5mm heat shrink strips from lcsc.com along with some pcbs and a shitload of other stuff. They forgot to ship 30 555 timers I ordered. Within 1 day of my email, they had shipped $2.58 worth of 555's EXPRESS through DHL to the US. Also, the 10cm heat shrink "strips" I thought I had bought were actually meters.

My question is: this is the first website I've tried for ordering components galore, and so far it kicks ass. Are there any other sites as good as this one?
>>
>>1258875
Huh, well okay then. For some reason I was figuring they used active devices to somehow set the output impedance for... some reason. I mean resistors ideally have a flat frequency response and buffers have so low an output impedance that the resistor will always dominate so I guess it makes sense. They must use more exotic resistors though to minimize thermal noise though.
>>
>>1258922
What I like to do is just scribble down the stuff I need on the parts supply list at work. Get a box of stuff for free the following Tuesday. Everyone does it, hell, my boss told me about it during the introduction trip. "If you need anything, just get the supplier ID and put it on the acquisitions form site. You did rc stuff, right? You need servos or stuff, just put in a request."
>>
>>1258931
I work at a clinical lab. The only free stuff I could get like that is laboratory gloves, containers, and hydrochloric acid. It would be nice, though.
>>
Anybody know of a good set of plans for a voice changer? I'm looking for more of a deep, sinister "deliver the money to locker 31 at the south street subway station or the hostages die, you have two hours" style one instead of a robot/dailek one, which is what most commercial ones do.
>>
>>1258873
It's generally easier to get a designated output impedance with a common-collector design.

>>1258931
clearly I need to get a better job

>>1258955
That's more of a digital signal processing project, once you get the microphone and speaker connected up. Might look at a pitch shifter guitar effect for inspiration. Do you need to disguise a single sample, a phone call, or your voice in person?
>>
>>1258958
I was thinking of using it with a CB radio so probably something similar to what a phone would use.
>>
>>1258966
Hmm, well, if you're not super concerned about making from scratch, one of the Orange Pi boards with a microphone and a line-out would probably do this sort of thing pretty well. Wiring it to the radio is going to vary a lot depending on model, if you don't just hook up a small speaker.
As for software, there are a few pitch shifters and suchlike available as plugins, and command-line tools to patch them to the system interfaces and to each other. You could do all sorts of audio manipulations that way.
>>
>>1258973
Yeah I think the way to go would be to set up a switch on the microphone to select between the regular microphone or the modulated microphone, otherwise just use a speaker and hold it over the microphone.

I'd imagine that if I'm doing it with a microcontroller I could also do voice inversion as a crude form of encryption too.

Thanks for the info, I've only done a few small projects before so I think this will be a fun one.
>>
>>1258980
The STM32 would make a great micro for this sort of thing. You'd have more than enough CPU, a good amount of RAM, decent ADC, and some PWM, all with bretty gud battery life if you should want to take your show on the road. You could do a bit more than spectral inversion, think frequency-domain permutation with the mapping table shuffled according to a key.
It does sound fun. Come back and let us know how it worked out!
>>
Can someone explain what the hell input and output impedance means? what effect does it have on a circuit?
>>
Can I use persulfate etchant mixed with sugar or something to make an explosive? Because I happen to have a few hundred grams of it lying about, and I'm not using it quickly. I wonder if I should melt them together or crush them into a powder or dissolve them together and recrystallise...
>>
>>1258986
in very lay terms, high impedance input will draw very little current from whatever you are supplying the input with.
a high impedance output is a "weak" output that is unable to drive a significant load, and the output signal probably needs to be amplified depending on what it's used for.
>>
>>1256922
Principles of electric circuits 9th edition by floyd
>>
>>1258983

yeah, the guy can spend a month doing it that way, or he could get one of these things; they're like $8 new, $1 used.

alternately, he can try an android/iphone app. there may be one that outputs out the earphone jack.
>>
>>1258927
>exotic resistors
No, standard metal film. Pth=kB·T·∆f. For a given bandwidth ∆f, the only way to reduce thermal noise is cooling. Thermal noise is far below the additional noise generated by any electronic circuit.
>>
File: cd4093b.png (15 KB, 275x245)
15 KB
15 KB PNG
I'm using pic related to drive a little piezo as alarm, the problem is it doesn't want to be toggled through the control signal, even if I throw it directly in ground its still going. its on a cd4093b ic and my current solution is just turning off power to the ic to toggle it.

am I misunderstanding something in the circuit or ic workings here or should that work with toggling control signal?
>>
>>1259275

the schematic is good. either the chip is bad, or more likely, you made a mistake. try a different gate, then a diff chip to eliminate first possibility.
>>
File: toggle.png (8 KB, 524x66)
8 KB
8 KB PNG
>>1259275
>>
I want to make a home-intruder alarm. I want it to have a PSU and a battery, so that it will run even while the power is out.

How do I make it so that it charges the battery at the same time the microcontroller is powered, and it can detect a power outage and do something about it (i.e: just logging the event in a SD card)?

I have several op-amps, mosfets, a PIC16F676, a few arduinos, a PSU and a car battery.
>>
>>1259388
What have you got so far? Show schematic or block diagram.
If "nothing", come back when you've got a design.
>>
>>1258671
For that thick you'd want spade terminals.
>>
>>1259388
Power supply goes through diode to battery, set up the micro to read the voltage on the power supply side, if it's zero, you're running on battery. You'd want a voltage divider to go down to either ADC suitable levels if you want to actually measure, or just suitable levels for logic 0/1. Measuring battery side may also be useful.
>>
File: IMG_20171012_210308.jpg (737 KB, 4160x3120)
737 KB
737 KB JPG
>>1259391

Pic related is what I use to interface the arduino with the 12v siren and the photodiode. There are also some light switches I need to toggle in a certain order to untrigger the alarm.

I know I can charge the battery by connecting it straight to the PSU, and I can power the alarm system with the battery.

What I want to know is how to design a circuit that will alternately charge the battery and the circuit, like some commercial IC's do.
>>
I've read that the L298 wastes a lot of power. Would I see any notable benefit from making my own H-bridge out of MOSFETs?
>>
>>1259400
Have it set up to charge the battery while the arduino runs off that. Your power supply will charge it and power the arduino at the same time. You don't need to do any switching.
>>
>>1259400
>alternately
You wouldn't really need to. Just power it all with a non-timed charger, just like cars do. A constant voltage charger set to 13.8V and maybe 3A is probably fine since you're unlikely to supply enough current to exceed a car battery's charge rating. What the battery doesn't absorb will go to your circuit.
As far as monitoring, why not just connect a mains-voltage relay to the mains?
>>
>>1259401
MOSFETs will have less voltage across them while on, so you'll get a higher output voltage and less power wasted as heat, so may also allow eliminating a heatsink if you were going to be pushing the limits of what an L298 could do.
>>
>>1259407
Right now I have two hours worth of battery time, and the heatsink does get pretty warm. The constant drain is about one amp, but it gets higher for a few seconds periodically. If I were to swap to MOSFETs, could I expect at least ten minutes more battery time?
>>
>>1259408
at 1A your L298 is dropping 2.6V nominal. at a 9V supply for example, almost 30% of your power is being wasted by the L298.
>>
>>1259410
>>1259412
Holy shit. Yeah, definitely switching to MOSFETs then. I don't think I'm losing 1A in the L298, I mean that the power the L298 and motors consume together (measuring amps at the negative lead from the battery to the L298) is 0.98A on average, with peaks and drops. My battery is a 3S LiPo pack, so the supply ranges from 12.6V to 9V, but if I can save around 30% power by scrapping my H-bridge I'll be more than pleased. I was hoping for maybe 10% more power, 30% is mad.
>>
>>1259420
yeah it's the wrong way to think of it but you're losing 0.3A equivalent. the LM298 has 1A passing through it but only dissipates some of that.
>>
>>1259420
sorry, got your post mixed up with the alarm system post.
Consider the DRV8828 motor driver which should be a functional replacement for the L298, but with 0.65ohm MOSFETs instead of BJTs.
>>
>>1259422
https://www.ebay.com/sch/i.html?_from=R40&_trksid=p2323847.m570.l1313.TR0.TRC0.H0.XDRV8828.TRS0&_nkw=DRV8828&_sacat=0
A bit outside my budget, I'm afraid. Did you mean the DRV8825, that one's more affordable. Still wouldn't work I'm afraid, I use brushed DC motors and couldn't easily switch to steppers.
>>
>>1259425
Don't buy the dev board, just buy the chip, they're $3.56 for one https://www.digikey.com/product-detail/en/texas-instruments/DRV8828PWPR/296-29725-1-ND/2764602, if you want it to be more breadboard/cheap through hole board friendly, get a breakout board like this https://www.digikey.com/products/en?keywords=tssop28%20dip
>>
>>1259425
the DRV8828 is only $3 on digikey but it's a surface mount part. your problem is that ebay only has overpriced evaluation boards for that chip. one alternative is the TLE5206 for $5 .i used it on a project with 2A motors with only a slab of aluminum as a heatsink, and you'll be dissipating 1/4th the power so you won't need one at all. don't buy obscure chips like these on ebay though. use digikey or mouser.

honestly though if you're willing to go to a bit of hassle you can easily make your own H bridge with two IRF530s and two IRF9540s, or whatever else is cheap on ebay. just buy a bunch of them and have them around to put in any project.
>>
>>1259425
Also, dumb bipolar stepper motor drivers can work fine on regular DC motors, but smart ones that do the stepping for you like that DRV8825 you mentioned can't.
>>
>>1258239
>Just not all that generically useful.
The multi million dollar market for them disagrees with you.

FPGAs will soon be integrated into all your processors.
>>
>>1259428
What I originally wanted to do was just cobble together something like http://www.instructables.com/id/DIY-MOSFET-Motor-Controller/
Pretty simple thing really. Set high and low to pick direction, and supply pwm to set the speed.
>>
>>1259441
the TLE5206 will do essentially that for you, but direction is chosen by toggling either one of two pins. having two pins means you can also stop the motor by setting both to 1 or 0. you can also hook the EF pin up to a led+resistor if you want to visually indicate when you've fucked up.
>>
File: file.png (56 KB, 1412x781)
56 KB
56 KB PNG
>>1259446
also i won't guarantee that i didn't fuck something up in my pic but you can make your own H bridge out of discrete parts that will function essentially the same as the TLE5206. it's free if you have shit in your parts bin. zero idle current and inputs are safe to leave floating.
>>
>>1259433
>mutli-million dollar market
>in a world where there are more ARM processors than humans
sounds boutique to me
>FPGAs will soon be integrated into all your processors.
Doing what, exactly, and how often? Amdahl's Law: the performance benefit of a hardware speedup is in proportion to how much it is used. Can you just imagine the context switching time on that array? Gee. I can almost guarantee that, whatever it's finally there for, it's not going to outperform a GPU on any significant task.
I've seen the Xilinx ZYNQ, which is basically the reverse of that. Hard cores still perform better.
>>
>>1259477
not the other poster.
fpga market was estimated at ~6.36B in 2015, it's not merely multi-million.
Doing what? The stuff that data centers use fpgas for *today*. Stuff where they didn't use a GPU because it wasn't the right choice.
It won't be the heavy vector floating point math that GPUs are good for, it'll be more packet processing, big-data acceleration, HFT, things like that. For examples, google fpga database accelaration. Or google amazon FPGA instance.
It won't make sense to "context switch," there will be sections of synthesized logic that remain in place for long-lived processes to use.
>>
>>1259502

thx. now my computer is infected with a virus, you SOB.
>>
>>1259433
>The multi million dollar market for them disagrees with you.
Reality disagrees with you. You don't even need to take my word, just open a random electronic device (hard mode: something relatively cheap). You can consider yourself rather lucky if you find any FPGAs or even CPLDs.

>will soon
You aren't the first to claim that, but FPGAs are inherently expensive and power-hungry way to implement logic, so I'm not exactly convinced.
Either way, your "soon" isn't going to be so soon that it is relevant to that anon pondering whether he should invest time now.
>>
>>1256544
Can I control a 120v / 12amp appliance with a regular 12v switch by using a relay? That's how relays work right? I'm trying to make a secondary switch for my dust collector with just a regular switch
>>
>>1259490
Huh, other anon was claiming it'd be in "my" processors. Data center usage makes a fair bit of sense, and an FPGA-driven IPv4/v6 router has been on my bucket list for a while now.

>>1259568
Surveillance hard disk recorders seem to contain them often enough. I cracked open an ADSL modem from long ago and found ten thousand gates or so of FPGA. In the 80s and 90s it was hard to find a significant piece of equipment or a home computer that didn't have at least a PLA/PAL/GAL in it, except maybe IBM.
Volume devices don't get much advantage from programmable logic these days when microcontrollers of all shapes and sizes can do just about everything you'd need them to.

>>1259614
Indeed, that's how relays work. Contacts and coil each have their own ratings, so as long as both are satisfied, you're golden.
>>
You could use a hall effect sensor to measure DC current so long as you calibrated it and kept the wires in the same place and all that, right? Just thinking it could be better for a high-power variable load since you could isolate it in case you gave it power spikes. Run the transistors off opto-isolators and you'd be golden, right? Though I guess the voltage measuring ADC would need to be on the hot side of the board too, hopefully one that fits into a socket and has parallel outputs to go straight into an array of opto-isolators.

What sort of transistor do you use for a variable load anyways? A BJT or some kind of FET?
>>
What sort of solar panel type would be better for the cloudy, cold and wet UK climate? I know there are different types but google searching just gives me a load of ads disguised as articles. I don't live near the ocean so sea salt is not an issue.

I hear that certain types work better in certain climates.
>>
File: 1498678780302.png (463 KB, 755x1069)
463 KB
463 KB PNG
I dunno if here or somewhere on /g/ would be the best place to ask, but I'm trying to make an ADB to USB converter, but I cannot find the full Apple Desktop Bus Specification document.

All the apple documents keep referencing it, but saying it comes as a part of a license to use the port, does anyone know where to find these more obscure things, or do I just need to go with what hobbyists have done?
>>
File: currentsource.png (10 KB, 220x311)
10 KB
10 KB PNG
I see a lot of simple current sources with this topology but I don't understand how it would be used for currents of an amp or more. Wouldn't R2 need to have a very low resistance and thus need to be very large to dissipate the heat?

Trying to drive about 0.7-1A through an LED as the load, so the sense resistor would get very hot. Is there an alternative way of doing this or would I be better off using an emitter follower and limiting the current into the base to control the current through the load?
>>
>>1259657
>You could use a hall effect sensor to measure DC current so long as you calibrated it and kept the wires in the same place and all that, right?
Some DC current sensors ICs work on this principle. Allegro company comes to mind. The two heavy current sense pins are wired in series to the load to measure current, the other pins are for reading the current sensed.

>Just thinking it could be better for a high-power variable load since you could isolate it in case you gave it power spikes.
http://www.allegromicro.com/en/Products/Current-Sensor-ICs.aspx
4.7kV galvanic isolation

>Run the transistors off opto-isolators and you'd be golden, right?
You might trigger your gate drive circuitry with opto-isolators but you wouldn't necessarily run the transistors right off optoisolators. One reason is that they have low output current and you want high current gate drivers to efficiently switch mosfets on and off, for example.
So, what transistors? BJT or MOSFET? Since you don't know you should look up why you might use one over the other. Hint: MOSFETs are the modern, popular choice for switching heavy loads.
Is your design switched or linear? Maybe I should assume "switched" but if you don't know, you need to think some more about it.
You're clearly building a power supply, what's your application?

>>1259710
>Is there an alternative way of doing this
Yes, forget the circuit you posted. Google switch mode LED driver, Linear Technologies makes tons of them.
>>
File: bridge.png (25 KB, 419x274)
25 KB
25 KB PNG
>>1259454
I don't see what the lower two diodes would be good for. For a = b = L (or open), all transistors are OFF and the motor floats (or coasts). For a = b = H the low side is OFF, the high side is ON and the motor is shorted (fast stop). I changed the base resistors to 10K because running transistors near B=1 (oversaturation) makes them slower and unnecessarily loads the input.
>>
File: adb.png (35 KB, 738x257)
35 KB
35 KB PNG
>>1259706
www.google.com/search?q=Apple+ADB+documentation
Do you have the AN591 data sheet?
>>
File: ldocc.png (3 KB, 200x188)
3 KB
3 KB PNG
>>1259710
It all depends on source voltage vs. load voltage. For a large difference switch mode is the way, for a small difference this idea may be useful.
>>
>>1259687
What you want to do is get a shit ton of plywood panels, and spray them all with chrome finish. Then you mount them up in such a way that they reflect sunlight at a central point. On that central point, you place a kettle. Ta-dah, solar powered tea.
>>
>>1259723
Not a power supply, a variable load as I already stated. I figure with BJTs you could more easily control the amount of current you're sinking by using the transistor itself as the load. Though I'm not sure how easy it is to control a FET's current, MOSFETs especially.
>>
File: IMG_20171013_231112013.jpg (1.23 MB, 2592x1456)
1.23 MB
1.23 MB JPG
Hey /ohm/, I recently purchased a Philips PM 3260 oscilloscope for a really low price but unfortunately the second channel doesn't even display a trace. After trouble shooting and opening it up it became really obvious that I wasn't the first person in here. I know already that I'll need to replace a few caps, but I was wondering if the safety cap on the power supply board needs replacing because it has hairline cracks. I was wondering if I should bother replacing this or if I could get away with out it?

Pic related, but the crack is barely visible.
>>
>>1259988
MOSFET drain-source channels are resistive in nature. For a simple current sink, you might be able to put together an op-amp circuit that drives the gate to an appropriate level based on the drop across a source resistor and a linear control voltage. For a simulated resistor, you would probably need to measure the load input voltage and multiply the control voltage by that.
>>
>>1260018
Well since I'd want it to be constant-current or constant voltage or maybe even constant power as well as constant resistance, I'd probably want to have all the feedback go through something digital, right?
>>
>>1260050
I'm assuming the load needs only handle one polarity at the load terminals. If you need to handle both polarities, the project gets a lot more complex.
Anyway, if you want timely, smooth, and precise response to transients, you probably don't want digital. Analog computers are just what you need here.
Taste
ti com/ww/en/bobpease/assets/AN-31.pdf
Also, a transconductance-type op amp is useful for multiplication and other computational tasks. OTAs can be a bit hard to wrap your head around at first, but the equations work out really nicely in some respects when you can think in terms of current. See also the LM13700 datasheet from TI.
>>
File: 1487219157546.jpg (231 KB, 755x1057)
231 KB
231 KB JPG
>>1260052
shit forgot pic
>>
>>1259996
Considering the place, it's probably just a noise suppression capacitor. Unless it is shorted, let it be; you have bigger problems to find.
Which probably are in the second Y amplifier, since the other channel is working.
>>
>>1259988
The difference between the basic BJT and MOSFET constant current sinks is rather small. MOSFET's benefit is that there's no base current error (emitter current = collector + base current), but they are somewhat more prone to oscillations and many older MOSFET datasheets don't show the safe operating area (as they're meant for switching). So, if you want some accuracy, go with MOSFET, but pay attention to SOA and stability.
>>
>>1259996
It's unlikely to cause any functional issues other than going bang and releasing the magic smoke when it eventually shorts.
>>
File: lZTK8[1].png (11 KB, 600x371)
11 KB
11 KB PNG
>>1259988
MOSFET based constant current is pretty easy, use an opamp for feedback.
>>
File: 1484730845855.png (28 KB, 639x788)
28 KB
28 KB PNG
>>1259710
If you will be fine with fixed 350/700/1050mA get cheap chink AMC7135
Or use something like >>1260075 or picrelated
>>
File: 1502969100904.jpg (11 KB, 250x250)
11 KB
11 KB JPG
>>1260075
>741
>putting a capacitive load on output of opamp
>no frequency compensation
disgusting
>>
>>1260144
yeah a 2MHz pole is really going to make a 1.5MHz bandwidth amp oscillate
>>
File: block diagram-min.jpg (1.22 MB, 9921x7015)
1.22 MB
1.22 MB JPG
>>1260062
So you're saying that I should be looking somewhere here? Chanel 1 seems to be working fine.
>>
File: block diagram detail-min.jpg (715 KB, 9272x4032)
715 KB
715 KB JPG
>>1260158
Shit, didn't mean to post such a huge picture
>>
>>1259988
>Not a power supply, a variable load as I already stated.
Sorry, I thought you were being deliberately vague by saying "variable load" and were asking ancillary questions of how to use transistors to switch power to it.
Anyway you can get big ass mosfets that can dissipate absurdly high amounts of power if you want that to be the load itself.
Or another way you can do it is digitally, so you can switch 8 binary weighted resistors with 8 mosfets, which gives you 256 possible load resistances. Just an idea.
>>
Has anyone ever used one of these before? I have to put together a trial skid at work and was thinking about trying one. Theres maybe a few dozen discrete IO and a few 4-20ma sensors and valves. Trial is set to run for 120 days.

https://www.automationdirect.com/adc/Overview/Catalog/Programmable_Controllers/CLICK_Series_PLCs_(Stackable_Micro_Brick)

Or a more general question what should I be paying particular attention to when looking at PLC specs?
>>
>>1260158
If channel 1 works, then the red block works, the delay line works and the channel 1 amplifiers work.
Either something in the channel 2 attenuators or amplifiers is wrong, or the channel selector doesn't select it.

Assuming you don't have another scope, you can feed the same signal to both channels and use the working channel to check whether the channel inputs to the channel selector are identical. If not, go back towards the inputs to find the faulty block. Or if the inputs are identical, check the channel selector block, channel selection logic and its control inputs. Also re-check the switches.
>>
File: IMG_20171015_002006.jpg (1.47 MB, 1872x3328)
1.47 MB
1.47 MB JPG
Hi ohm, I posted a thread about some random chink pcb making chemicals I bought off a local shop about 2 weeks ago. Some anons helped me figure out what the chemicals were and how to use them.

This time I thought I'd post in here instead of making a new thread. I just developed this board and it came out horrible. Does anyone here have knowledge of this method and can identify what I've done wrong here? I'm thinking I might not have exposed it to UV long enough for all the tracks to harden.

The process I followed was simple enough:
>Paint the board, let it dry
>Print out negative of design on two transparencies
>Place everything under a sheet of glass and expose using a 40watt UV lamp
>Develop with sodium carbonate.
>>
>>1260235
Developer does that if you let the PCB soak too long in it or it is too strong to begin with.
I've never used negative-acting photoresist, but using shit quality films is one of the easiest ways to get that result on positive resist. Both too long and too short exposure produces similar results.
>>
>>1260243
I used about a liter of water in a plastic bowl and about 2 tablespoons of sodium carbonate. Let it sit for 2 minutes then softly brushed over the board with my gloved fingers for about 5 minutes until I got this.

I don't think the developer was the problem here, I noticed tracks breaking even before most of the photoresist was gone.

Speaking of the photoresist sheet. This isn't a sheet, it was the liquid photoresist that you get in a bottle and paint on yourself. If I wanted to use a sheet instead, which would you recommend? It has to be deliverable to the other side of the world from where you are so its gotta come from china or somewhere close probably.

Also, don't you need a laminator to make the sheet stick? Or do they stick to the boards without anything? Sorry, I'm asking a ton of questions, I've never tried these methods before, I've only ever done toner transfer.
>>
>>1260248
How is this process compared to toner transfer? I've been wanting to try it out, specifically for two layer boards, but I don't know if it's worth the investment, particularly in how long it takes to make a board.
>>
>>1260257
Same story here anon, I've been making toner transfer boards for about 5 years now and have gotten quite good at them. I found a bunch of cheap chinese supplies for the photoresist method and I thought might as well give it a try. From what I've seen other people online, its an amazing method when you get the hang of it, I've only tried my hand at making two boards yet and both of them turned out shit.

I'm thinking of taking apart an old printer to turn it into a flatbed PCB printer like other people have done online. That might be the best method of all, with the quickest prints and the best quality but it takes a ton of time and effort to get it up and running. I'll probably work on it during the December holidays if I have enough money at the time.
>>
Has anyone here bought ws2811 strings from china? They are always in multiples of 50. Does e.g. 100 come as one set of 100 or two sets of 50? Cant find any sellers that go into detail and I'm certain that I wont be satisfied by the answer if I ask the seller directly.
>>
>>1259477
>outperform a GPU
They do that on the regular. FPGAs currently and will continue to beat GPUs in performance per watt in neural network implementations.

The Zynq is a hardcore. The MPSoC is an A53. Sure there will always be your run of the mill shitty ARMs on everything but if you need extra performance in a similar package why not have a Zynq instead?

>>1259568
You must not open up many things more complicated than a toaster then.
>Intel buys Altera, US will not sell Lattice to China
>t..they're not coming soon.
>>
I have 2 capacitors, one of them is 5uF the other one is 20uF. the 5uF one is charged to 480 volt and the 20uF one is charged to 120 volt. what will be the voltage if I connect these capasitors together (Plus to plus and minus to minus). I already tried this out in ltspice and I get an awnser of 298.43 Volt but this just isn't consistent with my calculations. how the fuck do I calculate this. I get that they both have 2.4E-3 coulomb but I don't know how to go further from here.
>>
>>1260331
Conservation of energy. E = 1/2CV^2
>>
>>1260235
are you making a drone controller?
>>
>>1260332
ok so,
E = 0.5*20E-6*120^2 = 0.144J
E = 0.5*5E-6*480^2 = 0.576J

so a total of 0.144 + 0.576 = 0.72 Joule
this gives 0.75 = 0,5*25E-6*V^2
60000 = V^2
V = sqrt(60000) = 244.95

244.95 ≠ the simulated 298.43
am I retarded or is ltspice retarded?
>>
>>1260331
Try putting a resistor in your circuit, even a very small one.
I've def had ltspice go a little haywire before where I was doing something unconventional, like a switched flying capacitor circuit.
>>
>>1260172
it's an imageboard, anon

>>1260235
I've heard of that stuff, actually. I've been thinking about it, but it seems to be a really touchy process for home use. I wonder if it's just a clever repurposing of LPI soldermask as an etch resist,
Anyway, did you run a calibration plate for exposure time or any other parameters?

>>1260297
Fair enough. I don't think they're going to hold up well against hard-core neural processing units, Thompson 1996 notwithstanding.
>need extra performance in a similar package why not have a Zynq instead
Zynq is a berry gud idea and, if my board-on-shelf budget were a bit higher, I'd love to screw around with one. I bet they're great for stuff like smart PCIe peripherals that need to run layer 4 and 5 or such.
>>
>>1260333
Yes, well kinda. It's not a controller but an extension for an existing controller which will allow me to have shit more organized on my quad rather than jumper wires flying everywhere, everything gets connected to this board and a controller board sits on top. This was mostly just something I cooked up in an hour to test the solder resist method. I think I'm going to be giving up on it for now and shill out to get some boards manufactured.
>>
>>1260344
well, I just tried it and I get an awnser which is even further off from my calculations (192V) could it have something to do with my test circuit, I am using these switches because I couldn't find any other way to simulate this in ltspice.
>>
>>1260345
>calibration plate for exposure time
Shit, why didn't I think of that. I'm stupid. I guess I know what I'm doing tomorrow.
>>
>>1256667
i've used multisim for class, pretty good. have a prated version running in a VM for when i need it.
>>
File: images.jpg (18 KB, 347x424)
18 KB
18 KB JPG
Hello /ohm/. I completed my electronics engineering degree in June this year and I've been basically jobless till now. I got some work pushing paper and excel spreadsheets at a real estate firm. Anyway, my question, since I have some free time and a little bit of money now, I want to get some certifications or accreditations to hopefully get a career in electronics. What would people recommend be the best? Most people are recommending me to get into networking (Cisco certs and what not) but I really love base electronics. Are third party certs for stuff like altium worth it? I'm not sure what I should be doing right now. Any advice would be appreciated.
>>
>>1260372
as an engineer in a different field the only certification i've ever seen a single person actually get is their professional engineering license.
>>
>>1260223
I'm a PLC babby but I have used a CLICK PLC.
The windows Click software is pretty featureful and easy to use, I ran and tested some simple ladder logic circuits with literally zero training. Was able to see the I/O values live as the program was running. I was happy with it.
They are cheap, the software is free, you can stack on more I/O, they have an okay warranty so it's worth a try.

I have an analog click and one of the analog inputs is burned out, and it's quite possible it was my fault (overcurrent accident?). It did leave me with the impression that its analog inputs may not be very ruggedized.

>Or a more general question what should I be paying particular attention to when looking at PLC specs?
"Does this PLC have the I/O capabilities and CPU power to do the job I need?"
So look at max I/O points, analog I/O, scan cycle time, max program size, motor control if desired, communication methods (rs232, ethernet, or others)
Look into how well your PLC can communicate with a human-machine interface device, like a touch screen.
For interest's sake you can casually browse the more expensive, high end ones like ABB, Allen Bradley, SIEMENS brands.
>>
>>1260349
What about simulating with mosfets?
>>
>>1258646
thicker wire means lower resistance, thats about it unless you are doing things with high frequencies
>>
>>1258683
skin effect on mains AC is about 2.5mm and typically you want ~5 skin depths to capture most of the current. unless he's using wire larger than 4/0 its not really an issue
>>
>>1260373
Already have my professional engineering license but I've gotten 2 interviews in the past 4 months, one of them never replied back and the other wanted me to work in tech support on less than minimum wage. I don't know what I'm doing wrong desu, I just thought getting some certifications might help pad up the CV a bit and land a few more interviews.
>>
>>1260347
Have you considered pre-sensitized boards?

>>1260349
In SPICE it's apparently helpful if every node has a dc path to ground. Maybe if you make the caps bigger and add some 10 gigohm resistors?

>>1260372
>altium
PCB design is for scrubs.
Given the higher-frequency nature of electronics today, something as simple as a ham ticket might be a worthwhile endorsement across the board.
Aside from that, I'd think applications in the electronics sense of the term, not so much the computing sense. What specialties are in demand in your area? Embedded programming? IC design? Power electronics?
>tech support
Do you mean applications engineering? Frankly, I think I'd enjoy that job.
>t.non-degree holder
>>
>>1260385
just keep looking anon, it also helps if you know someone in the industry. if you have you PE you could always go work for a civil engineering and do something like intelligent transportation systems.
>>
>>1260385
i've never once got an interview from applying online. my past two jobs and my last interview otherwise were all found either from connections or from calling and inquiring.

if you have a real degree from an accredited institution and don't have any mitigating factors like a criminal record, no citizenship, or severe autism then i can't imagine you won't get an offer in the upper half of five figures eventually. don't ever stop looking until then.
>>
>>1260387
>pre-sensitized boards
Those things are expensive, like 5 bucks a board I think and if you mess up the exposure, it goes down the drain. On the other hand I could order 10 boards from china for 15 bucks delivered and they will have a much better finish than anything I can get at home. The only up sides to toner transfer or photo resist is that you get it done in a day or two whereas it takes 10-15 days to get boards from china and you might get them done for much cheaper. (Less than a dollar per board I'm getting for toner transfer)

Pre sensitized boards are just not worth the money or effort
>>
>>1260349
192V is correct. Don't forget that in this case charge is conserved while energy is not.
>>
File: jfetpreampsimplified.gif (30 KB, 1503x1050)
30 KB
30 KB GIF
I'm a beginner looking to build the pre-amp pictured after it was suggested in the previous thread.
http://www.albertkreuzer.com/preamp.htm
I'm planning to run it from 2 9V batteries in series. Would it be able to drive headphones (32 ohm superlux hd681). What kind of capacitors should I use? (electrolytic, ceramic etc) How would you recommend I build it? I've never done a proper circuit like this so don't know whether to use a breadboard or perfboard or whatever else so any advice on that or anything else would be appreciated.

I have also started to look into poweramp options and found 2 ICs which seem to fit my requirements (practice amp), LM2876 and LM3886. The 2876 seems more straightforward since it is a mono chip but seems harder to find. The 3886 needs to be bridged which I'm not too familiar with though I have found ready made schematics (https://www.digikey.com/reference-designs/en/audio-amplifiers/106)
Any ideas on how to power this would also be welcomed. Thanks
>>
File: file.png (11 KB, 225x282)
11 KB
11 KB PNG
>>1260433
polarized capacitors are drawn differently in that schematic. you'll use electrolytics there and film or ceramic elsewhere.
>>
File: 47_headphone_amp.jpg (78 KB, 1162x542)
78 KB
78 KB JPG
>>1260433
No, the TL072 can't drive a headphone but 2x NE5532 can, also from +/- 9V (pic).
There's a good article about the LM3886 at https://www.neurochrome.com/taming-the-lm3886-chip-amplifier/
>>
>>1256918
TL 431 is not a transistor.
It's an adjustable shunt regulator.
Info:
http://www.righto.com/2014/05/reverse-engineering-tl431-most-common.html
righto.com is a GREAT blog BTW.
>>
>>1260235
You can get liquid photoresist? Is it better than film?
>>
>>1260508
no, having used both the film is much easier to use with a laminator
>>
>>1260414
I was thinking that would be the case I remembered an assignment where we had to calculate how much energy we lost while doing something similar to this
>>
do you guys solder wires often or is it just once in a blue moon?
>>
>>1260560
It depends on which blue moon. We give every blue moon a number and we only solder when this number is devidable by 4 or by 100, but not if it is devidable by both, allthough it is devidable by 100 but not by 400 then it doesn't count.
>>
>>1260414
So, (20E-6*120+5E-6*480)/25E-6 which is indeed 192, tnx
>>
scope ground: if i have several probes attached to a circuit should I connect all the ground clips or is one ground clip sufficient? if I connect more than one ground clip does it help or does it possibly pick up noise due to the loop(s)?
>>
>>1260577

one single ground, please. otherwise, you're like the guy with 3 watches: never really knows what time it is.
>>
>>1260577
those clips are in the end just connected to the same piece of metal in your scope
>>
>>1260589
This applies to slow signals only. Anything even slightly fast requires you to ground each probe properly.
>>
File: inelastic_s.png (14 KB, 561x457)
14 KB
14 KB PNG
>>1260555
It can be modelled as an inelastic collision when you replace momentum with charge, mass with capacitance and velocity with voltage.
>>
>>1260610
>Anything even slightly fast requires you to ground each probe properly.

I'm relatively new to electronics. What is "slightly fast" in terms of frequency?

My scope has the tiny ground thing that you clip on to the probe after you remove the ordinary probe clip so that the ground connection is very close to the probe point. At what sort of frequency would you use that, and what are we trying to do here?
>>
>>1260634

just try it and see. probe a 1Mhz pulse. does it look ''ringy'' or have curved sides? does adding a local ground make it nicer? probably not. try 10Mhz.
>>
>>1260634
It's not so much the frequency as such as the slew rate (rise/fall time), what you see when you calibrate your probe, only faster (if your scope can handle it).
>>
>>1260517
What about with a clothes iron?
>>
File: probes.png (42 KB, 1394x674)
42 KB
42 KB PNG
>>1260634
Pic related: 1MHz canned oscillator, measured with two channels. The other channel's probe was grounded, the other wasn't.
Even the better trace shows ringing due to non-perfect grounding.
>>
File: Capture.png (191 KB, 691x573)
191 KB
191 KB PNG
>>1260658
I've been interested in that too. I found this document which is a very detailed manual for using the photoresist sheet method to make PCBs. It says you can use a clothes iron as long as you slightly wet the board before hand so that the film doesn't burn.

https://sparks.gogo.co.nz/dry-film-tips.pdf
>>
File: 20171015_131541.jpg (1.76 MB, 3264x1836)
1.76 MB
1.76 MB JPG
Perhaps someone could help me. I'm looking for a 33uf 500v axial electrolytic cap in the US. I can only find them on European websites. I'm trying to return an amp I bought to stock, but if I can't I'll just throw a 30uf or 47uf in there. This is what's in the amp currently, from a previous owner.
>>
File: what a mess.jpg (31 KB, 368x476)
31 KB
31 KB JPG
>>1260707
33uF is E6, ebay has, not cheap. 47uF is E3, preferred, use that.
>>
>>1260667
Home-cooking PCBs has been a simple, cheap and reliable process for decades. Why has it become such a tortuous ritual that borders on self-sabotage?
>>
File: 1494882682720.png (158 KB, 1275x778)
158 KB
158 KB PNG
>>1260727
does it have to be axial?
>>
File: file.png (21 KB, 571x710)
21 KB
21 KB PNG
this is an AND gate used to charge a FET gate where the source could be between 0v and 30v below logic ground. the slow charge and discharge times are a benefit here. pretend the diode isn't there. why is it that when i cut off the top transistor while the bottom is still pulled to ground, current backflows through the bottom transistor out of the gate until it reaches 0v? aren't collectors supposed to be very low leakage?
>>
>>1260727
I'll just use a 47uf then. It's a common mod in guitar amps to raise the value of this cap anyways. I saw the ones on ebay but I don't necessarily trust ebay for something so specific. That's not the only hack job in this amp(I'm not even sure what the resistors are trying to accomplish, those are 2 70uf ruby's I'm sure the owner intended to replace the 33uf with) , but I got it for $200 when one in good condition goes for $450, and I like projects. Output transformer seems good though, was able to coax some oscillations earlier. Getting the right volatges in most places. It's not making sound but the expensive parts are good.

>>1260752
I'd much rather have axial because that's what the pcb was drilled for. Doesn't really matter, it's not like it will move or I'll place it in a way that the leads could come in contact with eachother. It's just autism on my part. I'd prefer f&t because aside from 2 phillips and some small nichion caps that's what fills the rest of the amp. More autism.
>>
File: IMG_20171015_153410.jpg (282 KB, 1280x720)
282 KB
282 KB JPG
I have a Minnowboard, pic related. I tried hooking it up up a TV before, didn't get anything to display. Don't have a microhdmi right now, so I'm trying to read the flash over spi.

So far I haven't been able to even get the flash chip to reply to the READ id command. Anyone feel like helping me debug it?
>>
File: NextGen_area_en.png (36 KB, 426x305)
36 KB
36 KB PNG
Has anyone of you been playing with the new-generation non-Silicon transistors made out of Gallium-Nitride (GaN) and Silicon-Carbide (SiC)?
>>
>>1260765
too expensive and the gates are obnoxious to drive. the implications for PFC and flybacks are great though, given a decade to age in and for some fitting gate driver ics to come out.
>>
File: equiv.png (2 KB, 211x188)
2 KB
2 KB PNG
>>1260760
Strange question. What happens to a charged 10n capacitor? It gets discharged.
>>
>>1260814
yeah i was retarded and looking at an npn equivalent circuit instead of pnp
>>
>>1260736
Because everything is SMD now anon and you need tiny little traces and multilayer boards if you want to get anything done. These are just ways to get a better result out of your PCBs
>>
>>1260350
https://github.com/sleemanj/pcb-test-pattern

>>1260667
One of the references is actually really worth reading. Adam Seychell, the guy who popularized the use of air-regenerated CuCl etchant, on the use of a home-built squeegee board as an applicator for dry film resist:
http://www.voodooengineering.com/images/stories/Files/Wet_Lamination_of_Photoresist_for_Hobbyists.pdf
Apparently a boiling kettle wash provides all the heat you need.

>>1260736
Because fine-pitch SOICs, and without them your choice of components and therefore your capabilities are greatly restricted. Also, not having to own a drill press.

>>1260761
Possibly voltage equalization between the caps to a finer tolerance than the caps alone provide.
>>
>>1260876
You can always make a super janky drill press with a dremel and a drawer slide.
Also, SOICs are still big enough where toner transfer is more than enough to make good boards with them.
>>
File: 33uf 500v.png (271 KB, 986x629)
271 KB
271 KB PNG
>>1260707
http://www.ebay.com/itm/33uF-500v-Electrolytic-Axial-Lead-Capacitor-2-Pieces-/282595634889
>>
>>1260877
Fine pitch, like 0.4mm. 8/8 spacing with toner transfer isn't realistic afaict.
>>
>>1260274
I almost bought some of these, but the answer is it doesn't really matter because you can just string them together. The connection is so easy that you can't fuck it up.
>>
>>1260232
I had a feeling that the issue is something that is in Channel 2's attenuator, because I read the tektronix troubleshooting guide and followed the steps to that conclusion, but someone told me to check some other area instead (wasn't really sure since I never troubleshooted an oscilloscope). Now that I opened it up again, I noticed that someone was messing around with the attenuator (there was a screw missing, there was a chip on a ceramic capacitor that was near one of the legs indicating that it was bent in and out of position). I now just need to get to the board , but it is mounted with screws that are really hard to get out due to their placement in areas where it is hard to fit a screwdriver in. It looks like I might have to put off fixing this scope for a little bit, since I have more important things to get to.
>>
>>1260860
This is why I just solder my components together directly to each other, and then coat it all in a few litres of glue with a glue gun.
>>
>>1260893
Please don't encourage them.
>>
>>1260893
Maker spotted
>>
>>1260933
>>1260935
Honestly, it isn't that bad. I use perfboard and cut apart jumper wires to connect stuff. Then coat it in glue. If you're not planning to change it up, it really works rather well. Warmer components sink their heat into the glue, and it works acceptably as a radiator. It's also waterproof and prevents short circuits when you have a bunch of circuits together.
>>
>not decapping your ics and wirebonding them together, then suspending them in a high vacuum chamber
>>
File: 1490760429626.jpg (37 KB, 550x363)
37 KB
37 KB JPG
>>1260944
delet
>>
File: sad pete.jpg (41 KB, 480x270)
41 KB
41 KB JPG
>>1260944
>it wasn't bait
>>
File: IMG_20171016_161647.jpg (1.79 MB, 1872x3328)
1.79 MB
1.79 MB JPG
>>1260876
Looks like about 4-6 minutes is the sweet spot for UV exposure. I also took the developer up a notch by adding almost twice the amount of Sodium Carbonate and by heating the water before hand.

I'm thinking I'll try stripping the board I made earlier with some Sodium Hydroxide and re-painting the board and trying again.

Would sodium hydroxide strip the copper layer too? I'm guessing it might oxidize it heavily and leave the board unusable. Otherwise I'm gonna have to get more copper clad boards.
>>
File: IMG_20171016_161700.jpg (1.71 MB, 1872x3328)
1.71 MB
1.71 MB JPG
>>1261101
Closeup
>>
just bought labview home + a beaglebone, i don't even know how to do circuits and shit

smdh @ myself
>>
>>1261101
It will oxidize it, yes. Dunno where and how much you're paying for copper clad boards, but don't but chink shit, it's always paper boards and not fiberglass like they should be.
>>
>>1261107
I'm in a shithole third world country anon, I can only get chink boards. That said I get them for about $1 a pop for a 6x6 inch single sided board and they're pretty ok, I've used em for almost 5 years now with the toner transfer method and not had too many problems with them.
>>
>>1261101
Slide 4 looks bretty gud for 12mil or even 10mil, but consider doing a 3.5-5.5 minute series to really dial it in (looks like you'd nail it at 4.5 or 5).
I'm concerned about those bumps on the vertical lines trying to become bridges to the horizontals. You're putting the toner sides down, right? How close is the exposure lamp? If it's too close, the double-thick mask will leave shadows. You can trade off crispness of mask vs. fast exposure time by changing the light distance.
The foot of the 2 in 0.2 is gone on slide 4 yet the centers of the numbers seem to border on underexposure. Maybe the developer's overactive?

>>1261107
It looks like there are a bunch of boards on aliexpress for 10x15x1.5mm for ~a burger and a half that look like legit glass epoxy. Are they bait-and-switching?
>>
>>1261120
>consider doing a 3.5-5.5 minute series
I'll do that tomorrow and update if anything comes out of it. Yes I'm putting the toner sides down and using two prints, one on top of the other. Not too sure why the bridging is happening, thinking my printer is atleast partially to blame here, the print isn't as solid black as I'd like it and the edges might be allowing light through.

The exposure setup is a piece of glass under which I put the board and a lamp which sits about 5 inches above the board. I don't think moving is much farther will have too much affect but I'll try it nonetheless. Thanks for the tips anon.
>>
>>1261107
>paper boards
Never heard of them but they sound awful. Where do people here buy their PCB making supplies if the cheapest ones are significantly bad?
>>
File: Weipu connector.png (157 KB, 427x315)
157 KB
157 KB PNG
hi /ohm/

How do I connect cables to these connectors, should I just solder them? Or should I ferrule the cable ends and clamp them?

I was soldering a few when I thought about clamping, since there will be about 8 amps AC going through the cables.

any suggestions are welcome.
>>
>>1261159
Those are solder cups, so solder.
>>
>>1261136
>paper boards
Look up phenolic resin PCB, you've have likely encountered these in cheap devices with single-sided circuit boards.
>>
File: DSC_0064.jpg (760 KB, 1920x1080)
760 KB
760 KB JPG
How do I solder this?
I'm not experienced. I just took a small tip. Turned my iron to 360degrees Celsius, added flux and went at it with my iron and just fucked it up. Every adjacent pin is soldered together now.

It doesn't seem humanly possible to do this.
>>
>>1261236
You can use a dremel to cut away the solder connecting the pins.
The secret to soldering tiny stuff like that is tiny soldering tips, tiny solder, and steady hands.
>>
I just bought my first soldering station, and I noticed that the solder barely even smells, this thing is so much better in so many ways.
>>
>>1261236
clean it
then moar flux
put solder on the tip and drag it from one side to the other
>>
>>1261165
Ah, that's what those brown PCBs are. They're probably why It's surprisingly easy to delaminate traces on this dot-board.

>>1261236
Put a piece of thin aluminium in between the traces if you have to, a piece off an old heat-sink works for me in a pinch. But in general use more flux and probably less solder. Your wires aren't too thick to make you want to use excess solder, are they?
>>
File: beep2.jpg (136 KB, 750x1000)
136 KB
136 KB JPG
>>1261273
>delaminate traces on this dot-board
Only if your tip is way too hot. I love dotboard because layout for test production.
>>
>>1261292

nice work. I bought some of those but I can't get the solder to bridge. Do you just keep piling it on or is there a certain technique to connect two specific dots?
>>
>>1260860
>tiny little traces
That is true. The smallest I can reliably do is SO (20/50) on FR2/FR4 with a carefully calibrated process, especially exposure time and etching temperature. That's the limit for now. It's definitely not the limit of the (cheap) base material I use: a hair on the board becomes perfectly visible as a copper line.
>multilayer boards
I don't think any hobbyist does multilayer boards because plating vias is an additive process. How do you do your vias on a double-sided board? I drill and rivet which is annoying and doesn't allow effective thermal vias. Fortunately I have a cheap 'spare space' PCB service if I met my wit's end.
>>
>>1261297
Again, your soldering iron is too hot. Also, I use 0.7mm Sn60Pb38Cu2 fluxcore, no lead-free yet.
>>
>>1261302
>Again, your soldering iron is too hot.

could be, since I just recently got a decent soldering station. the default is 750 f (about 400 c). is that too hot? if so, what would be a better temp?
>>
File: lion.jpg (712 KB, 2592x1944)
712 KB
712 KB JPG
Hi /ohm/, retard here.
I have removed this small lithium ion battery pack from a broken portable hotspot. It consists of two ICR18650-30B cells, and also has a small board of some kind attached to it (regulator?).
My question is: How might I go about recharging this without burning my house down?
>>
>>1261303
>is that too hot?
Definitely. Start at 200 and see if it melts, increase as necessary. The 'technique' rest on temperature, timing and addition of solder wire. Experiment: put solder on one dot (wait a few seconds) put solder on next dot (wait) melt solder between the two dots by adding a small amount. Remove tip in the right moment. The flux should never fully evaporate. If it turns dark, tip is too hot.
>>
>>1261313
The board looks like a protection module to me. The typical way to test-charge it would be 9V and a diode + small resistor in series, red (+) black (-). Monitor the voltage and stop at 8.4V.
>>
>>1261315
thanks. i will try that.
>>
>>1261320
Sounds like a plan. Thanks, anon
>>
File: Capture.png (331 KB, 541x377)
331 KB
331 KB PNG
>>1261298
>How do you do your vias on a double-sided board?
Using vias of course.
https://www.aliexpress.com/item/UXCELL-200Pcs-M1-5X3-Copper-Via-Vias-Plated-Through-Hole-Rivets-Hollow-Grommets-Pcb-Circuit-Board/32791992012.html
>>
>>1261320
>Monitor the voltage and stop at 8.4V.
Not him but if I were trying to check whether that board is indeed a protection module would it be a good idea to do so by leaving the battery hooked up to the 9V even once it gets to 8.4V, and see if it stops charging by itself or if I have to intervene?
>>
File: 1501991055953.jpg (182 KB, 1100x1100)
182 KB
182 KB JPG
>>1261273
>>1261292
i prefer these to paper boards, never once delaminated and all holes are through plated. you can get 25 for 8 or so dollars on amazon or cheaper on ebay
>>
>>1261159
Tin the cable
Flux the cups
Stick wire in cup
solder
I have heard of people tinning the cups first but that can make it a bugger to get the wire in.
>>
Ok.. just your basic "silly question" (tm)
https://www.allelectronics.com/item/cbc-25/1-farad-5.5-volt-supercap/1.html
Little 5.5V 1 Farad caps.
Been toying with the idea of using these for tiny solar powered LED lights, or for LED lights.
Any anon out there played with these things?
>>
>>1261236
I'd take some flux and hot air to it.

>>1261292
>through-hole production

>>1261364
>always test your fuses
Protection, anon. You'd take life out of the cells "testing" a protection board. Don't kids look up part numbers anymore?
>>
NEW BREAD HERE

>>1261407
>>1261407
>>1261407

NEW BREAD HERE
>>
>>1261379
Yup, and don't forget to slide a piece of shrink tube on the wire _before_ you start soldering..
>>
File: 0043.jpg (42 KB, 672x279)
42 KB
42 KB JPG
>>1261365
That's certainly a good price. I support my local dealer and pay €8 for 10x 100x50 mm phenolic resin dotboard. Fixed formats are not for me because I want my own size (like stamp-sized..) and I do not like glass dust. A paper dotboard is easily dissected, you slit it slightly along the holes (both sides!) with a cutter and then you can cleanly break it (pic bottom and left). I don't have the delamination problem because I use low temperature solder wire.




Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.