M-Audio Keystation 49 MK3 recognized but no MIDI events received #73
Comments
Nephiel
changed the title
|
Do you have enough power ? Please check inTransfer() return value. Lines 377 to 379 in 808b5a1 ↓ insert #ifdef - #endif |
|
Here's the dump with check in place: Steady stream of Also tested power draw with a meter, Arduino Pro Micro + USB Host board + Keystation MK3 only draw about 60 mA. |
|
Hmmm |
|
Did a solution to this ever get found? I think I'm having the same problem with an Arturia Beatstep. Here's the output if I enable ENABLE_UHS_DEBUGGING and EXTRADEBUG: `MIDI Init STEP1: MIDI Start Interface descriptor:
(ugh, the >'s break code formatting, but hopefully you get the gist!) |
|
Try commenting out the following line in Usb.cpp. Like this. rcode = dispatchPkt(tokIN, pep->epAddr, nak_limit); //IN packet to EP-'endpoint'. Function takes care of NAKS.
if(rcode == hrTOGERR) {
// yes, we flip it wrong here so that next time it is actually correct!
// pep->bmRcvToggle = (regRd(rHRSL) & bmRCVTOGRD) ? 0 : 1;
// regWr(rHCTL, (pep->bmRcvToggle) ? bmRCVTOG1 : bmRCVTOG0); //set toggle value
continue;
}
if(rcode) { |
Success! Sending and receiving MIDI note on/offs now seems to work! Thank you! :D Are there any drawbacks to this modification -- will this have a negative impact on the of other devices? (I've tested it with an Akai APCmini, and that seems to work, but maybe there is a case where it causes a problem?) I note that although noteon/noteoff messages seems to work fine in both directions, that using MIDI.sendClock() to send clock to the Beatstep doesn't seem to work -- is this perhaps handled in a different function that needs to be similarly modified? I have also tested with an Arturia Keystep, and that DOES receive the MIDI clock OK. [Edit: actually, upon further testing, I'm finding the Beatstep behaving strangely in other ways too - and am starting to think this is to do with the Beatstep configuration itself. I'll explore some more with this over the weekend and see what I can figure out] |
|
Sorry, I made a mistake in one place, please try again. This problem is on the USB_Host_Shield_2.0 side, not the MIDI driver.
Some devices have problems with the initial toggle, and BeatStep is one of them.
Usually there is no problem. |
EDIT -- see below! Thanks -- I've changed my copy of the library as you suggested here. It doesn't seem to make any difference that I can see, though. The Beatstep still doesn't respond to clock - was this meant to solve this problem? I've been through the settings in the Beatstep control app, hoping to find something like a 'MIDI clock mode' option, but had no such luck, so I don't know what else to try here..
Would it be better to solve this problem by applying that pull request to my copy of the USB_Host_Shield_2.0 library rather than commenting out the two lines you suggest? Edit: so to answer my own confusion! I've applied the above pull request to the USB_Host_Shield_2.0 library (https://patch-diff.githubusercontent.com/raw/felis/USB_Host_Shield_2.0/pull/438.patch) and the Beatstep now seems to be receiving clock! Steps to apply the patch for anyone who wants to replicate this:-
The other part of the puzzle here was that the Beatstep needs to receive an external MIDI start message before it will be clocked externally! |
|
@doctea |
Many thanks for your help, I understand now! Yes it seems to solve the problem and should be adopted :) |
|
@Nephiel
Unfortunately, that is current consumption under stable conditions. This keyboard draws a large current when it is powered on. I will say it again. You must connect via a powered USB hub. I've wasted my money, you know. 😭 |
This MIDI keyboard model is class compliant and seems to be recognized, but no MIDI events are received at all.
Here is the USB_MIDI_dump with
#define DEBUG_USB_HOST:Tried all of the keys, wheels, fader and buttons, but nothing appears on the dump after this point.
USBH_MIDI_desc output:
And the output of
lsusb -vvv:Let me know if I can help get this keyboard model supported.
The text was updated successfully, but these errors were encountered: