Scrub Track number hiccup at Trk3

9801503ca569f94c

If I scrub thru the tracks, when it hits track 3, the LED drops out to black. Then shows track 3 but not crisp as the other tracks. The timing hash marks dim also. It will also do this going from track 4 down to track 3 using the scrub knob.
Even after pasting data into the track, it seems like almost a sync issue the way the LED cuts to black as it hits Track 3 in either direction.

Any thoughts or things I shld look at?

Bill
LoopA#0029

17703227d9536da1

Hi Bill, is this only during runtime or do you see the same behaviour in test mode? The easiest way to enter test mode is to remove the SD card (and restart/reset).

I’m not sure what you mean by “scrub thru the tracks”?

9801503ca569f94c

latigid on
Good call on checking it out in that mode, I had completely forgotten with tunnel vision.
I put it in test mode.
When the top row of key does the cycle or RED, GREEN, BLUE. The keycap in question misses the BLUE flash.
At the same time, the screen seems to have a sync hiccup at that very point, all together.
Obviously this involves doing something to verify things. Thankfully we have other parts outputting successfully to compare to. I think it has always had this, once I did whatever I did.

I’m open to any thoughts on culprit/troubleshooting.

Thank you once again @latigid

17703227d9536da1

Hi Bill,

This could be something to do with the output driver, my guess (and hope!) would be that this is on the current source side. That way you actually have a chance to fix it! (The 595 shift registers are on the inside of the PCB sandwich…)

Still you’d have to separate the upper two boards (from the core PCB). Try your best not to bend the long pin headers.

The first and easiest thing to try would be to replace the LED in question. Is it a cPixel or a superflux LED?
The next is to check the soldering to that LED, so

R23, R25, R26
IC10, pin 12 (this is the blue output)
IC11, pins 17 and 18

For IC10 especially, are there any shorts to the adjacent pins? It might be that the output tries to drive into a pin that is “off” and could try to sink current without any limiting. This could result in the +5V supply sagging.

The next thing is to check the 595 registers, but that is a major operation as you need to desolder all of the keys first.

Best regards,
Andy

9801503ca569f94c

Plugged things in a more thought out manner.
Making sure that there is enough juice as there shld be.
Problem all but gone.

I think I was taxing the other setup with it plugged in with other power needy USB devices.
If it is a case of repair, then I wld submit that my LoopA improvement wld be a shell that can be reopened.
Would that require cables in leu of pins?

I still don’t know how those midiphy folks packed so much capability in this thing.
How did it evolve into this sort of eco system that also talks to the outside world midi?

Seriously, is there a Nobel award category for music?

17703227d9536da1

Hi Bill,
Thanks for the very kind praise! We’ve worked very hard on these devices and it’s wonderful to hear that you are enjoying the features!
For your solution, it’s still a bit puzzling to me that it only hiccups at that one position, but generally a good 500mA supply is needed to drive the LEDs quite hard. I’m glad that you could find something that works! For any future potential repair, it would be required to first open up the case to check parts. In the worst case scenario, all of the Matias switches would need to be desoldered to access the inner PCB surfaces of the upper sandwich. It’s doable but there is always the risk that some parts get damaged, but spare parts are available of course.
Best,
Andy

9801503ca569f94c

I spoke too soon.

Definitely the light blue/dark blue LEDs are not functioning on the button in question.
Just have to decide when to operate on the patient.