General > General Technical Chat
USB Diagnostic Board: What Connectors To Include?
EPAIII:
I am making a USB Diagnostic PCB. It will be an in-line device where I can scope the signals, interrupt the signals, connect the lines on either side to Hi or Gnd, and perhaps more. I plan to have two of the various types of USB connectors, one on each side of the board with switches between and test points on both sides so I can scope either end of the cable to see what both connected devices are doing.
I do not plan to have more than five pins connected: just using the basic four and I am not completely sure what to do with the fifth yet. There is a lot more reading to do. And I am not going to worry about speeds. This is mostly for the earlier, slower versions of USB. But it may get pressed into higher speed circuits. As it is a diagnostic tool, I am not sure on that.
What I want some input on is the types of USB connectors to include. So far I have started an order that includes types: A, B, B-mini, B-micro, and C. The type C is presently going to be an Arifruit 5180 Simple USB C Socket b/o board which only brings those five basic pins; none of the others.
I do definitely want to include that type C along with the A and B types. I am not so sure about the A and B mini and micro versions, some of which are listed as "deprecated".
When I said "in line device" above I meant that it would be placed between two USB devices where a single USB cable with the same or different connectors would normally be.
I intend to look through all my devices for guidance on which to include, but any ideas or suggestions would be appreciated.
Ian.M:
This isn't going to go well. Even at USB 1.1 'Full-Speed' (12 Mbps), signal integrity is important and D+/D- need to be routed as a controlled impedance differential pair to within a few cm of the host or device USB controller, minimising or preferably avoiding any stubs, length mismatch or extra capacitance. See https://resources.altium.com/p/routing-requirements-usb-20-2-layer-pcb
Given that you want to use this breakout for diagnostic purposes, the idea of introducing a bunch of stubs to provide different connectors in parallel doesn't seem like a good idea. Using USB micro-B connectors on your breakout and *short* adapter cables either side of it is far more likely to do what you want. If I had to do a hack job, I'd cut up some USB cables to get the ends I wanted, fit female DuPont connectors on the cut end, and simply put 0.1" pitch pin headers on the breakout board, taking care to keep D+ and D- adjacent and in between Vbus and Gnd in the header pin order.
ataradov:
Also, with Type-C it is electrically impossible to make an in-line cable tap with connectors. You will end up with a board that works only with certain orientation of cables. It may not be an issue for a test device though, you will just have to keep that in mind when using it.
EPAIII:
And that is precisely why I took the trouble to state that it is to be used as a test or diagnostic device only.
--- Quote from: ataradov on June 27, 2023, 03:32:30 pm ---Also, with Type-C it is electrically impossible to make an in-line cable tap with connectors. You will end up with a board that works only with certain orientation of cables. It may not be an issue for a test device though, you will just have to keep that in mind when using it.
--- End quote ---
EPAIII:
OK, good sanity check. I just gotta start going to bed before 3AM. Thanks a million!
You mention two types of connectors for my board; female DuPont connectors and USB micro-B. Assuming I could use either, which would stand the best chance of preserving signal integrity?
And I have never hacked into a USB cable. What are the chances that the conductors will work in what you call a "DuPont" style, header connector that was intended for ID attachment to standard ribbon cable?
I DO see that the kit you suggest uses crimp pins, not iD style. That kit is inexpensive and I think I have tools to crimp the pins so I may get it in any case.
--- Quote from: Ian.M on June 27, 2023, 11:27:50 am ---This isn't going to go well. Even at USB 1.1 'Full-Speed' (12 Mbps), signal integrity is important and D+/D- need to be routed as a controlled impedance differential pair to within a few cm of the host or device USB controller, minimising or preferably avoiding any stubs, length mismatch or extra capacitance. See https://resources.altium.com/p/routing-requirements-usb-20-2-layer-pcb
Given that you want to use this breakout for diagnostic purposes, the idea of introducing a bunch of stubs to provide different connectors in parallel doesn't seem like a good idea. Using USB micro-B connectors on your breakout and *short* adapter cables either side of it is far more likely to do what you want. If I had to do a hack job, I'd cut up some USB cables to get the ends I wanted, fit female DuPont connectors on the cut end, and simply put 0.1" pitch pin headers on the breakout board, taking care to keep D+ and D- adjacent and in between Vbus and Gnd in the header pin order.
--- End quote ---
Navigation
[0] Message Index
[#] Next page
Go to full version