Author Topic: USB isolation problem?  (Read 2155 times)

0 Members and 1 Guest are viewing this topic.

Offline MapleLeafTopic starter

  • Contributor
  • Posts: 25
  • Country: 00
  • Why don't wands have Si in them? Real Magic!
USB isolation problem?
« on: November 01, 2015, 12:21:45 am »
Just got some smoke out of a ST Discovery board! It was hooked up to my PC via  USB and  my Rigol ds2017 scope. With the O-scope and board not hooked up, I measured the USB ground and O-scope ground and got 68 mV. The Discovery board still works but a diode by the USB port is a little burnt.  For some reason my PC  ground is different from the rest of my bench. Both are on the same circuit coming into the house. Ground loops? Cheap PC power supply for USB?

So i need a USB port isolator before i burn up something really important.
It needs to be self powered. I think USB 1.x is good enough for the stuff i play with.
Any suggestions on what is good or bad for USB isolators?

Anybody else have this problem?

 
 

Offline Andy Watson

  • Super Contributor
  • ***
  • Posts: 2075
Re: USB isolation problem?
« Reply #1 on: November 01, 2015, 12:28:43 am »
In this situation that you have described, 68mV should not cause any smoke! Are you sure that you didn't short out the +USB voltage?
 

Offline MapleLeafTopic starter

  • Contributor
  • Posts: 25
  • Country: 00
  • Why don't wands have Si in them? Real Magic!
Re: USB isolation problem?
« Reply #2 on: November 01, 2015, 02:10:11 am »
Nope. The ST discover board was being powered by the PC's USB port. No open USB cables. Only the scope was hooked up when the USB cable was plugged in and the smoke came out.
 

Offline mij59

  • Frequent Contributor
  • **
  • Posts: 693
  • Country: nl
Re: USB isolation problem?
« Reply #3 on: November 01, 2015, 08:12:58 am »
Ground loops are unlikely, you could check by connecting an ampmeter between the grounds.
Looks to me you connected the ground lead of the scope to the positive terminal of the usb ST Discovery supply.
A while ago I bought a usb isolator based on the ADUM4160 on eBay, [url]http://www.ebay.nl/itm/USB-protection-ADUM4160-Evaluation-Board-USB-Isolator-/161078762151?hash=item25810adaa7:g:OtkAAMXQ4uJR-5Wj]http://www.ebay.nl/itm/USB-protection-ADUM4160-Evaluation-Board-USB-Isolator-/161078762151?hash=item25810adaa7:g:OtkAAMXQ4uJR-5Wj] [url]http://www.ebay.nl/itm/USB-protection-ADUM4160-Evaluation-Board-USB-Isolator-/161078762151?hash=item25810adaa7:g:OtkAAMXQ4uJR-5Wj[/url]
« Last Edit: November 01, 2015, 01:04:34 pm by mij59 »
 

Online Gyro

  • Super Contributor
  • ***
  • Posts: 9430
  • Country: gb
Re: USB isolation problem?
« Reply #4 on: November 01, 2015, 09:49:09 am »
Are you measuring that 68mV on the DC or AC range? If on DC then it might (just possibly) be an artifact of a more significant AC voltage - worth a quick check.

USB 1.1 isolators are getting relatively cheap now, I had to make my own when I needed one. You need to check is that they can output enough power, a lot are limited to 1W (200mA), assuming the ST Discovery board doesn't have a mains adapter.
Best Regards, Chris
 

Offline Jeroen3

  • Super Contributor
  • ***
  • Posts: 4069
  • Country: nl
  • Embedded Engineer
    • jeroen3.nl
Re: USB isolation problem?
« Reply #5 on: November 01, 2015, 01:28:34 pm »
Pulling the maximum of 2ma leakage of a laptop supply through one of the data pins as 0.5*mains is enough to kill your USB port.
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3233
  • Country: gb
Re: USB isolation problem?
« Reply #6 on: November 02, 2015, 07:06:55 am »
Pulling the maximum of 2ma leakage of a laptop supply through one of the data pins as 0.5*mains is enough to kill your USB port.

2mA is unlikely to smoke a typical diode though.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf