WARP Project Forums - Wireless Open-Access Research Platform

murphpo wrote:

Which LT1963 is it? I'm guessing U7, the 2.5v LDO in the middle of the board (near the user LEDs).

Does it get considerably hotter than the same part on other boards? Check for debris on both sides of the board; a short across one of the 2.5v bypass caps could overload the regulator.

What was the failure when loading a program from the CF card- red error LED near the CF card or green LED but no FPGA activity?

What's the FPGA board serial number (from the sticker in the bottom-left, starting with WFV2P-)?

Thanks so much for your quick response.

Yes, the high temperature is really abnormal, because I have other boards which are also running, and after a comparison, the degree of hotness is unacceptable.

And the abnormal LT1963A is U7, the 2.5V LDO in the middle of the board.

I have only red error LEC near the CF card, and the fan on the board works ok.

The serial number is WFV2P-1.2-00031

murphpo wrote:

U7 supplies VCC_AUX, which powers a lot of the FPGA's clocking logic. If this supply isn't 2.5v, the FPGA will definitely not work.

Since the board was sold by Rice 2+ years ago, it isn't covered by any warranty. That leaves two choices-

1) You can arrange for testing and repair through Mango Communications (email support@mangocomm.com).

2) You can try some more debugging and repair locally; a few things to try:
  -Double-check for any debris on both sides of the board
  -Using a multimeter (or oscilloscope), measure the output of each voltage suppy. There are red testpoints for each supply on the board. The terminal in the bottom-right corner (J39) is ground for all voltages. See the FPGA board users guide for the details.
  -Check the resistance across the two tantalum capacitors (C18, C19) near U7. They're the yellow (maybe black) caps in the "2.5v supply" region in this image. They should both show at least 1kohm across their terminals.
  -If the caps aren't shorted, and the other voltage regulators are working, you can try replacing U7. Digikey sells the LT1963AET (adj. TO-220) in small quantities.

Thank so much again for your response.

I will try to use multimeter to debug myself first, and hope I can figure out what happened.

By the way, could you please give me any hint about what may be the reason that caused the hot chip issue on my board, and I can avoid such errors in the future.

murphpo wrote:

U7 supplies VCC_AUX, which powers a lot of the FPGA's clocking logic. If this supply isn't 2.5v, the FPGA will definitely not work.

Since the board was sold by Rice 2+ years ago, it isn't covered by any warranty. That leaves two choices-

1) You can arrange for testing and repair through Mango Communications (email support@mangocomm.com).

2) You can try some more debugging and repair locally; a few things to try:
  -Double-check for any debris on both sides of the board
  -Using a multimeter (or oscilloscope), measure the output of each voltage suppy. There are red testpoints for each supply on the board. The terminal in the bottom-right corner (J39) is ground for all voltages. See the FPGA board users guide for the details.
  -Check the resistance across the two tantalum capacitors (C18, C19) near U7. They're the yellow (maybe black) caps in the "2.5v supply" region in this image. They should both show at least 1kohm across their terminals.
  -If the caps aren't shorted, and the other voltage regulators are working, you can try replacing U7. Digikey sells the LT1963AET (adj. TO-220) in small quantities.

thanks so much

for the redpoints test, do I need to turn on the power?
because as long as the power is turned on, the U7 becomes very very hot;

Also, we have test C19 and C18, C19 shows about 1Kohm, but C18 displays 0 ohm.

Thanks again for all your suggestions

Our testing result is:
C19: 0.8k ohm,  this capacitor works as a Cin and has a 3.3V at power on.
C18: 0.4 - 0.5 ohm, this capacitor works as a Cout and should be shorted. Because the voltage of C18 is 70mV at power on.

Then the voltage of testpoint tp4 is 70mV when power is on. The TP4 should be 2.5 V when power is on.
And we also test the TP5 = 1.5v , TP6 = 3.3v , TP7 = 5v. These three points seems working fine because of reference voltage from the board.

D14 seems not good. Because of the votlage over D14 is only 70mV when power is on. Diode conduction voltage should not be so small.

Other resistances are ok by now.

Do you have any suggestions for our local dignosis? We really appreciate that. Thank you so much

They're in the repository: http://warp.rice.edu/trac/browser/Hardware/FPGA_Board/Rev1.2/Schematics_FPGABoard_1.2.pdf

Also, I forgot to mention the short could also be due to foreign material on the board, bridging the pins of any of the caps between 2.5v and GND.

It's impossible to measure for the short, since all of the corresponding pins of all the caps are connected together (all should measure 0 ohms, even if only one component has failed). I'd visually inspect each part first.

The two terminals on C18 are connected to 2.5v and GND. Many other pins on the board are connected to the same nets (many caps, the U7 regulator, the FPGA VCC_AUX pins, etc.). There is a short between those two nets, but you can't determine its location with just resistance measurements.

If it's a failed cap (which is only one possibility), you may be able to see it (its case could be cracked or burned). Failing that, it would require removing them one-by-one. If it comes to that, I'd start with U7, then the big caps, then the small caps.