I have two boards with Cyclone III FPGAs that appear to program and verify properly but it doesn't boot. It has activity LEDs that turn on, showing that it is doing something but not operating as it should. It also has a buzzer that should beep on boot and an ethernet connection but neither are working.
What could be happening where an FPGA could be programmed properly but not load from the memory and boot?
Thank you for contacting Intel community.
- What Quartus version that was used?
- What type of configuration was used? Active seria, Passive serial, etc..
- Have you refer to Cyclone III device handbook for further information?
I used Quartus 12.1 to program with an Altera USB Blaster.
I forgot to mention that this problem is unique to these two boards.
I don't think programming setup it wrong because we have programmed many of these boards without issue. The two boards that I am looking at are outliers that I would like to get working.
Our customer did the design and firmware. We produce, program and test these boards.
Are there any signals that the FPGA needs to load and run the image from flash that isn't needed while programming?
I am mostly wondering about possible hardware issues with bad parts and bad connections.
The USB blaster appears to be fine. I can successfully program other boards without problems, including other boards exactly like these two.
I'm not setup on other computers yet but I could try that. Since I am able to program other boards on that programming computer, I don't feel like it is a PC problem either.
These boards did not work previously. These were assembled and failed to boot during test. After they failed, we have inspected the boards for irregularities, reflowed the BGA package FPGA, and replaced the FPGA. Nothing to this point has worked.
I'm sure it is possible but sketchy at best.
We don't have the capability to re-ball these BGA parts. So even if we had a known working board and wanted to borrow the FPGA, we wouldn't be able to re-mount the donor FPGA. Not to mention, I would worry about damage of a good working board and a good working FPGA.
The reason why we ask to test the good FPGA on the problem board is to check whether the board is the problem or the FPGA is the problem. If the good FPGA works on the problem board, it means the board is fine but the current FPGA is the problem, if the good FPGA is not working on the problem board, it means the board is the problem.
Since you mention that these boards did not work previously and they were assembled and failed to boot during test. You have also try inspected the boards for irregularities, reflowed the BGA package FPGA, and replaced the FPGA but nothing works. Hence testing the good FPGA on the problem board can be our last effort.
We do not receive any response from you to the previous reply that I have provided. This thread will be transitioned to community support. If you have a new question, feel free to open a new thread to get the support from Intel experts. Otherwise, the community users will continue to help you on this thread. Thank you.
I have no response because you haven't added anything to this thread. I was hoping that you would have a more in-depth insight to a possible problem outside of basic troubleshooting steps.
Yes, it is possible for someone to take an FPGA off of a good working board and put it on a non working board. However, there is a lot of room for error in there. During removal, we could wreck the FPGA. During the placement of the known working one, we could wreck the FPGA. Even if we didn't wreck the FPGA, we could still have issues with BGA balls not soldered. The many possible errors during the removal and replacement make this troubleshooting step very questionable and if the board still didn't work, I would still be questioning the FPGA.
Are there particular signals that make it possible for Quartus to say it programmed successfully but will not run?