Non-standard plug on a board with a low-speed device

[alberich]

I wish a Happy New Year to all members ...

A few weeks ago, I started a topic about detachable cables on low-speed devices. So I developped my application boards : 2 items equipped with an IOWarrior24, never connected simultaneously (and not to be sold) ... using a B-plug and a 2.0 cable ; they work without any problems, the cable remains attached to an A-connection on the motherboard and I connect and disconnect my boards via the B-plug ; the HID-component by R. Marquardt works fine with Delphi6, in order to detect connected (or not) state and disconnecting while the prog is running.

But the problem is that the B-plug seems (to me) to be not very reliable (and durable) if it is too often plugged in and out, in German we would say it moves "like a cow's tail" when plugged in (sorry for the bad joke ! ).

So I want to ask if anybody has some experience using a "non-standard" (ie not B-type) connection system, which would be more reliable in durability, perhaps (mini)DIN connectors ? It would not be very difficult to cut the B-plug from the cable and solder the new one in place, and the layouts would only need minor modifications (I am using a CAD program, so it is easy to make a new PCB, and all DIL components are mounted on sockets).

Thanks for any answer.

[Jan Axelson]

If I recall correctly, it's a low-speed device so it shouldn't be using a series-B connector anyway. I have no specific recommendations, but the USB specs call cables with vendor-specific connectors "captive cables" if that helps in a search. I think a DIN or mini-DIN would work.

[alberich]

Thanks for Your answer. There was a "long" discussion, about using or not B-connectors on HOBBYIST developments ... I think that the best solution (solving most problems) would in fact be a 4-pin mini-DIN connection system, with the cable's shield connected to the male plug's shield, and a female connector on the PCB. 4 pins allow the connection of the power-supply pair and the differential-signal pair, and I think (?) that the shield should NOT be connected to the board's GND ... or should it ? In the present version with B-type connectors it is NOT connected and that works fine, without any "noise" (despite of the fact that there is a RF 27MHz transmitter near the interface board, which is controlled via opto-couplers by the IOWarrior).

[Jan Axelson]

The USB 2.0 spec says:

6.8 USB Grounding
The shield must be terminated to the connector plug for completed assemblies. The shield and chassis are
bonded together. The user selected grounding scheme for USB devices, and cables must be consistent with
accepted industry practices and regulatory agency standards for safety and EMI/ESD/RFI.

[alberich]

I was ready to order the parts to build 2 new interface boards with 4 pin miniDIN connectors, but at last minute I remembered that they would not ensure a clean hot-plugging-in/swapping, because with non-USB-compatible connecting systems the problem is that the power terminals will not be connected before the signal connections are performed. So the A-plug cannot stay connected to the motherboard (computer running ! ) while the interface is changed on the miniDIN side !
Modifying the length of the pins at the male miniDIN plug would result, as I think, in loss of reliability. And otherwise perhaps there will be some hardware damage, or at least a "dirty" reset of the board.
So I think the best (or "less bad"   ) solution is to keep the B-connectors ... and to connect the shield to Power-GND on the board !
What is better : a system in conformity with all the specs, but which does not work properly, or a system in contradiction with some specs (but as I said it is a hobbyist system, not a system to be sold), but working ?
In addition, I wonder what will "happen" to the motherboard's connectors if the A-plugs are "too often" plugged in and out ? Warning to all USB-memory-sticks users !

Alberich