1. Home
  2. Field-Programmable Gate Arrays
  3. Virtex 4 could not work correct,is it damaged?

Virtex 4 could not work correct,is it damaged?

One chip of Virtex 4,SX35F668,is used in my PCB board,When the board is first assembled,the FPGA could performance very well,but about a month later,the FPGA could not work correctly even download with the same bitstream file once worked OK. I check the operate current of the total board,the current is about 0.5A lower than normal(5V input voltage),After the review every detail parameters of each ICs on the board,the static resistor of 2.5V is much higher than before(approximate 800ohm now but approximate 10ohm before),the 2.5V is only provided to FPGA`s Vccaux and two banks`s Vcco.because the Vccaux is provided to DCM,so I write a Program to test the 8 DCMs in Virtex 4 SX35,the test result is that:every time I turn on the power and load the same bitstream file to FPGA through JTAG,the output of the

8 DCMs is not the same!Sometimes,2 or 3 DCMs`s output is not correct,and the DCM lock signal keep low,the worst condition is 5 of 8 DCMs`s output is not correct, and the correct DCM and incorrect DCM is not fixed during each time.

could anyone give me some advice to solve the problem?

Reply to
Borry.Wang
Loading thread data ...

That's a tough call, as Xilinx doesn't provide an internal (no IO) self test bitstream for their FPGAs making in-field, on-board, testing of the parts impossible. Not specificed for OEM to test parts before placing on a board either. It's a total nightmare to qualify these parts, on or off board, in a lot of respects.

John

snipped-for-privacy@gmail.com wrote:

Reply to
fpga_toys

Borry,

Not enough information in your posting.

Is this one device? Out of how many? If this is one of only one devices, maybe it was destroyed by an ESD zap? Perhaps it sustained damage when the power supplies were applied incorrectly?

Resistance of a power rail is useless. At what voltage? How measured?

Better to know is what is not working. You mentioned the DCMs, how is the output not correct? If a "bad" DCM sometimes works, then how can this be the device is bad? Have you checked you input clocks for jitter and if they are within the ranges specified for the DCM?

I would submit a webcase, online, with every single detail, if you wish to get an answer.

Austin

snipped-for-privacy@gmail.com wrote:

Reply to
Austin Lesea

snipped-for-privacy@gmail.com schrieb:

we have JTAG based frequency meter - clock analyzer

formatting link

see second screenshot, it shows the 4 DCMs of Virtex 4 FX12 running in factory test mode, eg self oscillating. You should be able to use this tool to check your DCMs, the software uses Xilinx cable IV (in cable III compatible mode) and runs under WinXP.

If you are interested to check out the DCMs with our tool let me know, I will create the SX35 bitstream for the DCM factory test setup for you.

Antti

Reply to
Antti

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.