RE: [Discuss-gnuradio] USRP1 USB I/F failure & repair

[Paul Mathews]

Thanks for the clarification about C73, C74. The Raychem datasheet doesn't
say so directly, I'm guessing that these parts are some form of varistor. In
my experience, varistors are less reliable than silicon clamping devices,
both in terms of their dynamic clamping characteristics and their ultimate
failure characteristics.

http://www.conformity.com/artman/publish/printer_211.shtml

http://www.nxp.com/acrobat_download/applicationnotes/AN10753_1.pdf

I work in a high humidity environment, and I'd be surprised if ESD was the
cause of failure of these parts, anyway. I suspect that flexing of the
circuit board during USB plugging/unplugging caused cracking. The large
solder fillets on the parts may have contributed to the problem. 

Paul Mathews




-----Original Message-----
From: Matt Ettus [mailto:address@hidden 
Sent: Monday, May 18, 2009 4:01 PM
To: address@hidden
Cc: 'GNURadio Discussion List'
Subject: Re: [Discuss-gnuradio] USRP1 USB I/F failure & repair


Paul Mathews wrote:
> About a month ago, our USRP1 Rev. 4.5 (Ser. #2728) became erratic in 
> it USB connection and communications. A short while after, USB failed 
> altogether. I put scope probes on the USB + and - data lines and 
> observed that one of the signals had a much different sort of signal 
> than the other. Following advice from Matt Ettus given to others who 
> had experienced USB failure, I ordered a replacement USB chip along 
> with a Chipquik SMD removal kit. Replacement of the chip had no 
> effect, so I was forced to look more carefully for a cause. It turns 
> out that one of the 0603 capacitors installed near the USB receptacle 
> (C73, C74) was shorted, and removal of these 2 capacitors restored USB 
> function. I'm posting this description of my experiences for the 
> benefit of others on the list.
> 
> I suspect that C73 and C74 (which I can't find in various schematics 
> and BOMs for USRP1) were added in the hope that they would provide a 
> path for electrostatic discharges, which is ironic. In my experience, 
> low-capacitance avalanche breakdown devices are more suitable for this 
> purpose.


Paul,

Thanks for your analysis of the situation.  C73 and C74 are not actually 
capacitors, they are, as you say, electrostatic discharge protectors, 
part number PESD0402-060.  I have not seen any fail in the manor you 
mention, but I suppose enough static could cause that.

Has anyone else seen this failure mode?

Matt