Sudden malfunction of certain channels

[dmitri]

If it is an ARM based then the MUXer will be the same to check soldering.
If besides Aux4 and Aux5 you have Aux6 and Aux7 false triggering then I would also suggest to check soldering of the pin 17 on the ARM chip and the resistor R13.

[derbstens]

It's my second MD I received built by you a while ago.
I think I mentioned the type in an earlier post.
Right now I test switching to aux 6...

[derbstens]

...which seems to work, although it's not in the very same context as yesterdays rehearsal.
What is hard for me to believe is that two inputs in a row can go bonkers without external cause.
If I understand you right, you think it's a hardware issue too?
I still have my Atmega 56 input Kit MD, which should work ok. Can I copy my settings to the other MD?

[dmitri]

...which seems to work, although it's not in the very same context as yesterdays rehearsal.
What is hard for me to believe is that two inputs in a row can go bonkers without external cause.
If I understand you right, you think it's a hardware issue too?

Based on your problem description, it seems so. Electronics do some time fail for various reasons although I cannot remember any MegaDrum failing so unfortunately you might be the first. If you send the board back to me, I can troubleshoot it and repair if needed.

I still have my Atmega 56 input Kit MD, which should work ok. Can I copy my settings to the other MD?

You can of course copy settings but because of slight differences between ARM and Atmega versions you may need to adjust these settings.

[derbstens]

It happened again. The next Aux channel is kaputt.
The channel went bonkers during a gig on my birthday party two weeks ago, mercifully everybody has had a drink, so noone was upset of the hiss that was constantly emitted...

So now I would like to send the board to you Dmitri so you can have a look what causes the malfunction.
At the same time the Megadrum has become an important tool for my band to rehearsal and we cannot afford to skip a lot of dates.

As i am interested in the faster arm version running at 120 MHz I wonder if it is possible to purchase that board, swap it with the broken one and put this one instead of my atmega 644 in my old Megadrum Kit version, once it is repaired.
As far as I can tell the connections on the boards haven't changed...

[dmitri]

If it is a very old Atmega644 kit then it may have different inputs pinout. Please, take a picture of the board so that I could confirm that you can swap it directly without rewiring input jacks.

[derbstens]

You soldered it in Dec 2012, I'll make a photo later, when back to rehearsal room.

[derbstens]

On the other hand I seem to have messed the whole input order up anyways, so a rewiring of the inputs would be nice to do.

I have now engaged and plugged in the Atmega MD and I feel there is a pretty considerable difference in speed and ease of use.

Am I seeing this right, the first ARM is 6 times as fast as the Atmega644 and the recently introduced ARM@120MHz doubes speed roughly?
I'm sure that it translates not proportionally in reducing latency but can you make an approximate estimate on how much is gained?

[dmitri]

According to my measurements at viewtopic.php?f=3&t=2079, the latency on STM32F103 is better than on Atmega644 by about 2ms - 1.2-1.8ms on average. STM32F205 may be further 10-20% better.

[derbstens]

Great!
So if compatibility with the rest of the connectors is not a problem, I'd like to order my third Megadrum Brain together with the positional sensing board.
Here is a picture of the Atmega board: https://www.dropbox.com/s/m4hzq0i2sgxvq ... .jpeg?dl=0
I hope it works as I'm iPad only in rehearsal room.