Anyone ever designed their own PCB?Trying to replace the front IO Panel that supports a USB 3.0Anyone know the name for Female 19 pin header?Want to place it on the board so I can just use off the shelf USB 3.0 19 pin header extension cable.
>>2962509Recereated the HD Audio for now.Now need to figure out the USB3.0 connectors. And maybe replace the Firewire with USB C
Why are you trying to put a modern motherboard inside a case from 2006?I don't think slapping modern connectors on is going to magically give you modern performance from unshielded untwisted ribbon cables.
>>2962516Because it looks cool, and I'm reusing older motherboard.Intel 8th gen.It's going to sit under the TV as an emulator box and TV watching. The case supports IR remote and there's also VFD display! I want to see it in action
>>2962516VFD display itself
>>2962509if you're new to pcb design anything that involves USB3, PCIe, HDMI etc. is probably not a good idea
>>2962537>if you're new to pcb design anything that involves USB3But then I'll never learn...Found a part that seems similar, should fit but has no notch. ZX-PM2.54-2-10PY
OP, what are you using to calculate the diff & single ended impedance of the USB data lines?
>>2962509There's nothing special about the PCB. You could just mount all the shit to a piece of protoboard that's cut to shape. Get protoboard with the correct pitch so you don't have to drill out holes to fit the jacks and pin headers.
>>2962574They don't want you to know this but you can actually ignore line impedance and reflecting waves. The Smith chart is a tool made by big antenna to sell more antennae. Who is Smith anyway? Nobody knows. Electricity isn't a wave; I'd know, I'm a surfer. You actually don't need to know math to make a circuit. Just dabble 'till it works. That's what everyone does anyway. Math is for nerds. Don't be a nerd
>>2962586The PCB connects to the motherboard with shielded USB cables in the original PC. A couple of strips of copper foil tape for the ground plane is enough and if you use equal lengths of magnet wire flat against the protoboard it won't be an issue for OP's use case.
>>2962624Yea
>>2962551Thanks anon, you just made me go down the rabbit hole.Apparently my traces will need to be on 4 layer board.2 layer board needs wider traces.Used this calculator and PCBway formulas.https://www.digikey.ca/en/resources/conversion-calculators/conversion-calculator-pcb-trace-impedancehttps://www.pcbway.com/multi-layer-laminated-structure.html
>>2962574I feel like the protoboard won't be able to fit all components in the size I need.I only have 140mm x 18 mm to work with.
>>2962509Fuck no. Fuck that. Absolutely the fuck not is it a beginner PCB design project. USB 3.0 is enough bandwidth to be a bitch about trace routing signal integrity shit.Your best bet is ripping out the USB 2.0 components and bolting/gluing/bolting+gluing a premade molded cable to the PCB in their place, something like this.It's what most off the shelf cases do for USB 3.0 front panel as well.
>>2962731But anon, the whole signal integrity made me go down the rabbit hole of learning.I think what I have should be good for 4 layer PCB2 ground layers in the middleI just now need to check it over and do all calculations. Hopefully i wired it up correctly.
Should be good now.Just need to figure out what to do with Firewire slot.
>>2962769https://www.ebay.com/itm/184850995329
>>2962509I absolutely hate this shit Makes my skin crawls feels so fucking gookishI can't wait to finish this shit I'm working on and throw away my soldering iron Yuck
>>2962769Going to leave the Firewire location blank so I can mount USB C cable there.
Dude just install kicad and figure it outIt has built in footprints for most connectors by name so you can drop one down and get a part number to go order.It does differential signal trace length matching as well which you will need.You could knock out a board in like 3 hoursYou do not need 4 layer, skill issue and a half.
>>2963191Yeah I'm going to pull the trigger and order the PCB on the weekend.Hopefully I wired it up correctly
>>2963882wrong pic
>>2962530That is a very cool case, if you get the remote and VFD working then that would be hella sick.
>>2962550The reason he says that is that high speed digital signals turn everything into weird antennas and it's almost impossible for someone inexperienced to predict how it'll act, meaning it probably won't work. There are additional design rules for USB3 signalling and USB-C connectors like impedance matching, trace length matching, stub antennas, and so on. Have you looked at the implementation notes regarding ground plane (de)coupling? USB2 is comparatively much easier as even shitty designs will probably work.
>>2962769>Should be good now.Prepare to be surprised.https://www.renesas.com/en/document/gde/usb-30-board-design-guideline-rev-10-20121203
>>2964480Thanks.Surprisingly My traces are 0.5mm or more from each other and since it's 4 layer board, there's ground planes in between the traces on the other side.Ordering the PCB today, with PCB assembly included. Will let you guys know if it works or I fucked up.>>2964329Yeah I got it for free, It's missing the cage for the DVD drive, so I'll need to think this one through, how to print it out on a 3d printer. (Pic related)
>>2962530love these cases lol. got a few on ebay for similar idea of upgrading them never got round to it. just use them to put hdmi out to digital signage instead of smart tv. look forward to see where this thread goes.
Ordered
>>2964518It's a very short adapter wire essentially so your chances are better than if you were designing a dense PCB but that spiderweb of traces with hard 90° bends all over the place and differential pairs not being routed in pairs half of the way and sitting right against other signal lines is not going to do you any favors. We don't make soft bends because they look pretty. We do it because once you get into the GHz range, the edges of traces become both emitters and reflectors and it starts to get properly weird.
>>2964708>than if you were designing a dense PCB but that spiderweb of traces with hard 90° bends all over the place and differential pairs not being routed in pairs half of the way and sitting right against other signal lines is not going to do you any favors.Interesting, yeah if I could I would have made the PCB larger for all of those traces. Having to fit it on 18mm pcb constrained a lot of my design choices.If this is successful. I might try my hand on doing some USB routing board from scratch. Just to learn
>>2964712You don't need a bigger PCB. It's perfectly simple to route those traces sensibly on the one you have. You just need to have a general understanding of how signal lines interact with one another and to follow the design guidelines.
