FPGA Failure

[JunL]

Hi

I am Jun Lee from Nanowave Technologies.

This post is mainly for Intel/altera support!

NANOWAVE is having recent anomalies with Intel/altera FPGAs type Cyclone III. The failure modes observed were:

1. Sporadic reset and relaunch of SW code running on NIOS softcore processor.
2. No start-up of softcore processor at all, no JTAG communication.

In both cases the failure was fixed by replacing the Cyclone III (QFN package) component. No other PCB-related issues were found.

We have a suspect that either the devices are not new and might have a history, or that some other problems occurred during handling/shipping, like ESD events.

NANOWAVE kindly asks if Intel/altera would be able to perform a device-level root cause analysis on the defective devices (opening the package and examine the chip itself), and who should be contacted in such a case.

Thanks you…

Manufacturer’s part number: EP3C10E144I7N

Lot# 198191

[wwanalim_intel]

Hi,

Greetings and welcome to Intel's forum.

Please give me some time to check on this issue and will get back to you with the update.

Thank you.

Regards,

Fathulnaim

[JunL]

Hi Fathulnaim,

Thank you for your email.

Please update me when you are available.

Best regards.

Jun

[wwanalim_intel]

Hi,

Sure Jun. In the mean time, is the board can work properly on basic code such as helloworld on the NiosII processor?

[wwanalim_intel]

You can follow the procedure here to try the OCRAM booting method to perform the helloworld in NiosII.

https://www.intel.com/content/www/us/en/docs/programmable/683689/current/processor-booting-from-on-chip-memory-ocram.html

As we don't have example design in the design store for Cyclone III, and have limited resource for this support, we may only give the best support for this issue.

If the testing of helloworld still has the issue, then you can continue for the device-level root cause analysis on the defective devices by contact the local distributor that you purchased the device from.

[wwanalim_intel]

Hi,

Do you have any updates to share about this issue?

[JunL]

Hi,

Please see my reply below:

In the one case with sporadic resetting NIOS softcore processor, we had no time to test other, more simplistic code, as you suggest. But the code is in use on several hundred units and fully tested.

In the other case, the NIOS softcore is not even starting and JTAG is not working.

Again, both FPGAs are nominally new, just purchased and received from distributor and just populated onto the boards. Both worked and passed final testing, then were shipped to customer and showed the described anomalies. Looks like infant mortality, but needs further investigation of any damages inside the FPGA IC itself.

[wwanalim_intel]

Hi JunL, 

For further step on device level root analysis, there are a few information that we would like to have if possible.

1) We would like you to erase and reprogram the Cyclone 3 to get the checksum value then we will compare with the mfg checksum value.

2) Is there any electrical functional issue on the Cyclone 3?

3) Is it possible for you to share the identity of the end customer?

It would help us if you can provide those above information but if cannot then you can inform us back. 

Thank you. 

[JunL]

Hi, 

Please see the highlighted test in red below:

1) We would like you to erase and reprogram the Cyclone 3 to get the checksum value then we will compare with the mfg checksum value.

NW: Sorry, but the two defective devices are no longer on a board, and we do not have a jig for programming FPGAs stand-alone. This could be done by Intel support anyway on receipt of the devices.

2) Is there any electrical functional issue on the Cyclone 3?

NW: One FPGA shows a repeated reset at about 45°C device temperature (actually measured on the PCB, so temperatures could be different by +/- 5°C), below this temperature, the device performs normally.

The other FPGA doesn’t operate at all. Therefore, JTAG is also not working.

Again, the other circuitry on the 2 control boards did not show any problem, and both boards did work right after replacing the defective FPGA with new ones. So, the evidence is very likely pointing to a problem within the two FPGA ICs.

3) Is it possible for you to share the identity of the end customer?

NW: Unfortunately, we cannot share the end-customer information. All we can state is that it is for the commercial use.

Thank you.

JunL

[wwanalim_intel]

Hi JunL,

Thank you for the detail explanation. We will be sending an email for you to fill in the form. Kindly check inbox or spam folder in your email.

[JunL]

Good morning, 

Thank you for your information.

JunL

[wwanalim_intel]

Welcome JunL,

[wwanalim_intel]

Hi Junl,

Did you received the email?

[wwanalim_intel]

As we do not receive any response from you on the previous question/reply/answer that we have provided. Please login to ‘https://supporttickets.intel.com’, view details of the desire request, and post a feed/response within the next 15 days to allow me to continue to support you. After 15 days, this thread will be transitioned to community support. The community users will be able to help you on your follow-up questions.