ESP8266 Community Forum

- Sat Jul 30, 2016 1:44 am #51753 I can't guarantee success. I am still working through everything, and there's a lot I don't know. I yet may have to settle for USB low-speed. *EDIT* If I do that, I won't need the low-latency interrupts.

Basically, if I want full-speed, there are many roadblocks, this interrupt thing is only one problem.
If I want low-speed, there are many fewer roadblocks.

- Sat Jul 30, 2016 1:53 am #51756 Have you seen viewtopic.php?f=9&t=3979

- Sat Jul 30, 2016 2:24 am #51759 Whelp that's deeply concerning. How in the world do people do their software WS2812B stacks I wonder... There is one convenient thing though. With USB, if something does go wrong, the host typically retries.

I'm a little bit scared of trying to modify the vecbase stuff though I did not see this thread. Thank you for finding it!!

People are using NMI in different contexts and it is rather confusing.

Once NMI is triggered further NMIs are disabled until return from the handler

That sounds rather promising.

Tunnel vision guys...

It's in the SDK documentation that the new PWM implementation uses NMI's. Maybe other code as well.

If you want precise PWM, use the PWM implementation in the SDK, that's precise BECAUSE it's use NMI's.

Sounded promising... Sadly, the code posted by @rhaenel on page 6 would be quite infeasible for me since it would take so very long to initialize.

This is looking less and less likely.

- Sat Jul 30, 2016 4:15 am #51763 Have you seen this project? viewtopic.php?f=11&t=10239
This guy has apparently figured out how to read USB packets using interrupts...