0
\$\begingroup\$

We are using the FPGA ProASIC3E A3PE1500-PQ208. The FPGA got internally short circuited during runtime. The FPGA IO Supply Voltage and FPGA ground are permanently short circuited and it is not reprogrammable after the short circuit as TDO does not toggle.

The most recent code had dynamic switching between the IOs which caused the FPGA device to go out of normal operation mode (This point is mentioned in the FPGA fabric guide). The coding practice till now was not to internally tie up/down the unassigned IOs. Looking at the FPGA fabric details document, the inference was that unused IOs are either given high impedance or weakly pulled up. Now, if several IOs are weakly pulled up to 3.3V it is dangerous and induces a fault in our case.

What is the protocol to avoid such internal IO short circuit faults in PQ208 devices?

\$\endgroup\$
2
  • 2
    \$\begingroup\$ I’d be contacting microchip support first. \$\endgroup\$
    – Kartman
    Mar 21 at 12:41
  • \$\begingroup\$ Thanks, contacted microchip via a support case and got a response from one of their engineers. \$\endgroup\$ Mar 23 at 10:58

1 Answer 1

2
\$\begingroup\$

The unused I/Os being weakly pulled up do not cause any issue.

Please look at the below document https://www.microsemi.com/document-portal/doc_download/129992-hb-proasic3-e-sso-and-pin-placement-guidelines

The SSO(Simultaneously switching outputs) or the SSN need to be avoided in designs that get more complex. In the board design near the FPGA pins, make sure that we follow the guidelines given by Microchip and avoid too dynamic IO characteristic that can cause the device to go outside the normal operation and internally short-circuit.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Out of curiosity, do you know of the mechanism by which SSO violations lead to the failure? Is it standard latch-up due to transients on the supply rail, or something else? \$\endgroup\$
    – nanofarad
    Mar 24 at 19:59
  • \$\begingroup\$ I am not completely sure of various possibilities of failures. It may be latch-up due to transients on the supply rail or even two transistors on the input buffer going to unpredictable state and drawing large currents. \$\endgroup\$ Mar 25 at 1:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.