I think only Paul Stoffregen himself can answer this complex question?
Paul we bought loads of 3.6 and now 4.0 with good succes.
We recently converted our C++ application to a javascipt webbpaged based program.
We send/receive millions of messages with 1millisecond interval(very accurate).
2 years ago all USB issues were solved with the genius Serial.send_now() function. with USB analysers we see all the USB packages are send with exactly 1ms interval.
But now with Chrome we have a slow USB return time of 21ms.
Our theory:
After some digging we found ut that javascript(read: html) has a refresh/sample rate of 5ms. Settimeout and Setinterval only works with values above 5msec
Our theory is that windows/chrome use default USB time-out of 16.
16+5 = 21.
Coincidence?
Serial.send_now() doesn't seem to work.
With a serial analyser program we can see the messages are send in chuncks, and/or tripple messages in 1. (instead of every 1msec).
Also we see a strong instability. One day we see USB packets including 100 single messages, and the other day we see USB packets including exacly 1message like it should do.
Did you already play around with Chrome's new USB functionality?
Thx!
Paul we bought loads of 3.6 and now 4.0 with good succes.
We recently converted our C++ application to a javascipt webbpaged based program.
We send/receive millions of messages with 1millisecond interval(very accurate).
2 years ago all USB issues were solved with the genius Serial.send_now() function. with USB analysers we see all the USB packages are send with exactly 1ms interval.
But now with Chrome we have a slow USB return time of 21ms.
Our theory:
After some digging we found ut that javascript(read: html) has a refresh/sample rate of 5ms. Settimeout and Setinterval only works with values above 5msec
Our theory is that windows/chrome use default USB time-out of 16.
16+5 = 21.
Coincidence?
Serial.send_now() doesn't seem to work.
With a serial analyser program we can see the messages are send in chuncks, and/or tripple messages in 1. (instead of every 1msec).
Also we see a strong instability. One day we see USB packets including 100 single messages, and the other day we see USB packets including exacly 1message like it should do.
Did you already play around with Chrome's new USB functionality?
Thx!