lower latency debounce algorithm

[e00E]

There are several ways you can do debouncing. This page outlines some of them.

The way Kaleidoscope currently does debouncing is called "defer" in the link. (I determined this by glancing at the code in the file keyscanner/ATmega.h.) The downside of this approach is that it incurs a fixed a latency. The latency is around 5 ms according to a comment in the code.

I would like to use a lower latency way of debouncing. The one called "eager" in the link. This mode has no extra latency.

[obra]

In the past, when we've played with eager, engineers have reported phantom press issues related to faint electrical noise on a circuit board design, which has definitely made me skittish. I'd love to see it become configurable, though. In the keyscanner we use for the Model 100, we actually built out support for state-machine based keyscanner designs, along with a ton of recorded test data from questionable switches.

…

On Mon, Aug 26, 2024 at 9:37 AM Valentin ***@***.***> wrote: There are several ways you can do debouncing. This <https://docs.qmk.fm/feature_debounce_type> page outlines some of them. The way Kaleidoscope currently does debouncing is called "defer" in the link. (I determined this by glancing at the code in the file keyscanner/ATmega.h.) The downside of this approach is that it incurs a fixed a latency. The latency is around 5 ms according to a comment in the code. I would like to use a lower latency way of debouncing. The one called "eager" in the link. This mode has no extra latency. — Reply to this email directly, view it on GitHub <#1435>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAALC2EOOPUTOLLPM2TMV5DZTNKS3AVCNFSM6AAAAABNELOWSWVHI2DSMVQWIX3LMV43ASLTON2WKOZSGQ4DOMRXHA2DGOA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>