USB cables - which is best for noise rejection?

But you generalized, AlwayWrong. You're wrong and insist on proving it...

...always.

it may not be sufficient if the USB is opto coupled on the board because ground of computer and pinball may still be connected.

We had this on a very high cost very precise power supply where the USB was opto coupled but the USB socket connected to case of power supply.

To measure nA/pA with this setup, we needed one of these USB isolators which separate the shields of the two USB cables.

Pinball games do have shock sources - flippers and kickers, etc. - so I am a bit concerned about the USB plug to socket connection. I will look into those Permatex fillets.

We do have the cables secured with tie straps of course.

Thanks!

John :-#)#

Why do you suspect the USB? Those cables have twisted pair data differentially sent/received, and an overall shield, and except for the ground loop, shouldn't be any kind of problem. Ground loops oughtn't affect USB signals. The random lockup or freeze is just as likely a memory leak or noisy power converter.

If the cables test as low resistance, and you've already tried contact-cleaner (stabilant-22 or cramolin for instance) solutions, it's likely the problem is elsewhere.

Actually the test was simple. With the original (assumed to be factory) USB cables installed the game would freak out after a short time - mostly the flippers would freeze in either a raised or lowered position. The flippers are PWM and there was no problem with that aspect - coils were not burning out as if the PWM failed - rather the computer seemed to lose track of what was turned on/off in flipper land. The computer was still recording all the other switches and running the lights and most of the other coils were behaving properly.

No schematics, and the company went t*ts-up so no factory support.

New, shielded cables and the game was stable for days. Other folks had similar problems and new USB cables that were shielded got them running too.

There are other electrical design issues (IMHO) with the game but this one was somewhat easy to solve. Still it has the odd communication issues and it might be USB ports aren't isolated or something else. Electrical noise is relatively high in these games, lots of coils and LED busy doing their things.

I am looking into some other good suggestions that have been offered.

Thank you for taking the time to look at this!

John :-#)#

If this thing is in a pinball game then I would suspect connectors could be getting banged around, Shielding, power supply noise, or software bugs. For whatever it's worth :-/ Steve

I wouldn't expect any of this stuff to be isolated in the nA range. I use it for audio stuff but mainly to make sure stuff keeps its smoke on the inside. Anything with a PC (particularly a laptop) in it can be a grounding nightmare. Isolated hubs make things a lot simpler.

John Robertson wrote in news:irudnXccYpGaDZjBnZ2dnUU7- snipped-for-privacy@giganews.com:

Do all of the flipper solenoids have snubber diodes across their leads?

This could be an EMP pulse issue.

The flippers AND the bumpers produce an EMP pulse each time DC is sent to those solenoids. It is actually the collapsing flux from the DC energization that does it. So ANY actuator solenoids in that machine needs to get snubber diodes if they do not have them.

USB isn't all differential, some of the signaling is single ended

On a sunny day (Tue, 4 Dec 2018 05:13:12 -0800 (PST)) it happened Lasse Langwadt Christensen wrote in :

Today I had a problem again with USB connectors. Somebody mentioned flipper boxes moving, now my advice it to replace the USB connectors with for example high quality D connectors, or simply solder the wires. USB connectors are a disaster. I inspected some with a magnifying glass, you see burned and eroded track traces, all sorts of dirt and goo. Reason I try to do everything with ethernet, galvanic isolation comes free, never a problem.

Fancy a good laugh? Try this.

Feel like a good laugh? Give this a try.