-
Notifications
You must be signed in to change notification settings - Fork 187
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Selftest is not supported on revC #151
Comments
One problem is that there could actually be shorts on the side of pull drivers too. |
There could be shorts on the side of the pull drivers, there could also be a bridge over the 10k resistor, changing pullup/pulldown to push-pull. Some of this can be detected on the other port utilizing the pins-loop and hoping that there is not the same error on the same pins. But in the end you can detect all shorts with current measurement. This is why I wanted to have it on the test jig pcb. |
Just for added information. I just ran into this issue too. The
All other selftests pass. This is consistent on 3 revC1 boards I could get tested. |
Also ran into this in a new board, I had tested before with an older version and it passed.
|
Yes, the selftest is completely broken on revC, please stop reiterating this over and over again. I should probably just remove it. |
@whitequark can this be closed after 13a3509? |
No, we still have to implement some form of selftest on revC. I've renamed the issue. |
@whitequark would you like me to look at this? I think the first steps would be to get the existing tests to be sane on revC? |
@TomKeddie I'd be happy to assist you with that if desired (not that I think you need assistance to do it, but two sets of heads and hands may useful). |
@rwhitby I'll look at this toward the end of the week. I have some fixes already on a branch, I made some changes/hacks to find the shorts in the boards I built. |
Selftest can give false-positive results on revC, see for example https://twitter.com/szeloof/status/1161346534098833408
The main reason is probably that cmos inputs are left floating in high-z and their value is used for the test results. This seems to be a leftover from revA and revB which had hold circuits.
So selftest for revC has to be reworked to utilize the pullups/pulldowns.
For the pins-int test the internal pullups of the ice40 should be used. This is not yet supported in nmigen.
The text was updated successfully, but these errors were encountered: