07-14-2015 10:44 AM
We are using Virtex 6 ML605 board for one of the projects. The FPGA board and the Test PC communicate through ethernet. Initially, The board was working fine for a while before the ethernet communication did not happen. I am not very sure on what action led to this failure.
I ruled out any defects in the ethernet cable, test PC's ethernet port, DIP switch settings in the FPGA board. Now, when I turn ON the FPGA board, DS25 (12 V LED) turns ON. The FPGA init LED (DS 31) momentarily becomes red. The FPGA done (DS13) LED turns green in few seconds. The fan is working.
I tried to re-program the BPI Flash to see if it resolves the problem but I was not able to. When I initialize the chain in IMPACT tool , "Cable not detect" message is displayed.
I was looking through some documentation available in the web. Based on them, I understood that there are power status LEDs that turn ON during power up. But most of the power status LEDs do not turn except for DS25 and DS13. I probed the power supplies in the FPGA board and the following voltage rails seem to work fine: VCC5, VCCINT , VCC2V5 , VCCAUX and VCC1V8. The following voltage rails did not produce the intended voltage : VCC1V5 and VCC3V3. I did not measure MGT_AVCC and MGT_AVTT. But the fact that DS27 and DS28 did not turn on during power up is indicative that thre is a fault in these two power rails (MGT_AVCC and MGT_AVTT)
Could the problem related to the onboard power regulators impact the Ethernet communication and programming the BPI Flash. Please advise on the further course of action.
07-14-2015 01:04 PM
Having power supplies not working is never a good sign. As to what is powered by what, you could examine the schematics on line. In general, even one bad power supply would be considered a failure that precludes use of the board.
Is it possible damage was done by mis-handling? Electro-static discharge is the common reason why devices fail (not just Xilinx).
UG116 lists the ESD tolerate (no damage) numbers. These values are tested every year to make sure we are within specification.
07-14-2015 01:21 PM
Thank you Austin. Mis-handling of the board could be one of the probable reasons. But, I am not entirely not sure. It would be beneficial if you could advise on the future course of action. Would you recommend us to debug the board based on some checklist? I was going through other forums and there was a suggestion to reset the power controller for a similar issue. You had replied in another forum that webcase has been stopped for smaller companies. I can provide more details to you over IM if that would help to give access to webcase.
07-14-2015 01:36 PM
There is an answer record for debugging board problems, you may wish to follow that:
The above is NOT for your board, but the approach should be similar for checking the power supplies.
Or, you may go back to your distributor, and request they help you (which may be expensive, as no one works for free).
07-15-2015 04:48 AM
07-15-2015 07:50 AM
Hi Austin, Sorry to come back. One of my colleagues had checked with the digikey distributor and they told to check with manufacturer. This was before I got in touch with you. Probably, I or my colleague will get in touch with digikey again and communicate to them your response.
Before that, I just wanted to check one thing. The "Returns" tab in the service portal is still active for me and allows me to raise a service request. Can I raise a return request in it or still should I go through my distributor.
Sorry for any confusion.
07-16-2015 01:21 AM
It looks this thread is about an ML605 board so you can refer to AR40398 for debugging steps
Wrt return request, if the board was purchased via distribution, then the disti should open a request on Xilinx portal.
07-20-2015 07:28 AM