Moldover's Mojo

[KG]

I chose 73 rather than 1 or 35 as when used to control volume
the loss is imperceptible.  OK I need to close this down now and move on to other things.

Best,

Kenneth

[moldover]

Bravo!  I am a happy camper

Is the bulk pricing you sent me before still valid?

( ! )

[KG]

Great!  And yes.  Should I print up some UMC inside stickers ?

Kenneth

[moldover]

Hey Kenneth,

I'm on the road now, getting more time to perform with the UMC-32.  It seems that the bouncing problem has returned.  Maybe it was never totally resolved, or maybe it has come back due to some change in the last firmware you sent me.  The problem is not as bad now, maybe 1 out of every 10 movements.  I can try the capacitor thing again, but it didn't seem to help the last time around.

Do you have any ideas?  It seems like the simplest thing, but can you send me a wiring diagram for the capacitor?

Thanks,
( ! )

[moldover]

P.S.  This problem is happening with the two types of gaming switches I shipped to you, so perhaps you can duplicate it with your own hardware.

( ! )

[KG]

Hi Moldover,

OK this should help some.  Although those switches you sent have a nice and low GramForce to actuate they tend to waiver between states if you hold it in the middle -- which in this case makes debouncing irrelevant.  A quality tactile switch will not do this and incorporates a mechanical form of anti-hysteresis.  I'd recommend using a Sanwa if this is a big problem for you. http://techtools.myshopify.com/collections/diy-kits/products/arcade-buttons  Even at 2mS they seem to work pretty good with this firmware.

In this build there are 4 debounce times to choose from, 2mS, 5mS,  10mS,  20mS.  I think 20mS is pretty useless but it is there as an example.  Even with 20mS of debounce applied to your switch there is still some noise that can creep thru.
Keep in mind that the longer the firmware has to wait for the switch to settle the less responsive your control will be -- no way around that.

Also, not sure who brought it up but I added a feature on these input types that will keep the CC# and shift the MIDI channel (except for channel 16) when the shift pin is engaged.  So you could call it the return of the shift pin.

For the config utility.  There is a autofill feature based on the last control "touched".  To activate simply right click anywhere on the background of the dialog app.

Best,

Kenneth

[moldover]

Hey Kenneth,

Wow, I wasn't even aware of the wavering between states in the middle thing.  I've got 24 switches in front of me that I just carefully tested with the previous firmware.  About half of them work flawlessly, the other half have varying degrees of wavering. 

I've tried those Sanwa switches and pretty much every other kind of gaming switch I could find.  The one's I'm using I feel are far and above the best for performance.  I think I'm going to try and buy a new batch and return the switches that don't work to my standard.

I'm not so sure denouncing is my problem but I spent a little time with the new firmware and config utility.  I tried to test the new debouncing choices you added but the config utility seems to crash before loading a config with 5ms debounce.  I have limited time and tools while I am traveling this month but I can test more later.

Thanks!

Best,
( ! )

[KG]

I'd really like to hear more about the config utility crash as it is the same transfer code as before.  The next time it locks up please post the debug_log.txt that it spat out in the same directory.

Best,

Kenneth

[lqbert]

Moldover,

First off, much props to what you're doing with your music and Controllerism.

In a previous post you mentioned that you're using the Spectrasymbol SoftPots in the Mojo, I believe.  Did you compare them to the HotPot line?  If so, can you tell me why you chose SoftPots over HotPots?

Thanks,
lqbert

[moldover]

Hi lqbert,

Thanks for the props.  This forum is for the UMC-32 and this thread is for working with it's developer KG.  Your question is more appropriate to the controllerism.com forum.  Please repost it there, Email me a link and I will reply.

Best,
( ! )

[moldover]

Hi KG,

We've built several Mojos, all using the exact same configuration, but we're having a recurring problem with some of the units.  Analog inputs seem to get "stuck".  It is easiest to notice if you move a fader quickly from the top of its travel to the bottom.  Most of the time it will "stick" at a value higher than zero.  Usually it hangs at a number close to zero (16, 6, 4 etc.) and never goes there.

The first time it happen I solved this by swapping in a different UMC-32, but I have just tried 3 different UMC-32s and it's not helping.  Do you have any idea what might cause this kind of behavior or how to troubleshoot it? 

Thanks,
( ! )

[KG]

Hi M,

Please find attached the latest beta, there are some improvements in how sampling is performed.
Let me know if this helps.

Best,

Kenneth

[moldover]

I tried the new firmware and it doesn't seem to help.  Perhaps it is useful to add that this "stuck" behavior only happens with RAW analog inputs.  AVERAGED analog inputs do not show this behavior. 

Any other things I could try?  I find it strange that it happens on only some of my controllers and not others.

Thanks,
( ! )

[KG]

Just to clarify a couple of things.

1) Is this stuck behavior happening with a standard potentiometer or with the Spectra softpot ?

2) When you say that it is happening on only some controllers, do you mean the UMC ?

3) When you say "Stuck" does it simply hold its value or does it eventually ramp up or down to the target value over a period of time ?

Also...

4) If you hook up only 1 of the known "sticking" pots to a known "sticking" channel on a known UMC32 can you replicate
it that way.  (This test would also entail setting the other unused channels to off).

and

5) Does changing this pot yeild identical results ?

Best,

Kenneth

[moldover]

1) Is this stuck behavior happening with a standard potentiometer or with the Spectra softpot ?

Both standard pots and softpots

2) When you say that it is happening on only some controllers, do you mean the UMC ?

I am using UMC-32s to build 10 Mojos all with the same components.  By "some controllers" I meant "some UMC-32s".

3) When you say "Stuck" does it simply hold its value or does it eventually ramp up or down to the target value over a period of time ?

It eventually ramps to it's target value.  It can take anywhere from 1/2 a second to as long as 5 seconds.

4) If you hook up only 1 of the known "sticking" pots to a known "sticking" channel on a known UMC32 can you replicate
it that way.  (This test would also entail setting the other unused channels to off).

Yes, I get the same behavior with this configuration using only one slide potentiometer.  The average time it "sticks" is shorter, but it is still very problematic.

and

5) Does changing this pot yeild identical results ?

Yes, I tried 3 different pots with the same results

Of the 7 UMC-32s I have with me, 4 of them exhibit this behavior and 3 of them do not.  Is it possible that my working with them somehow altered or damaged these 4?  Do you have any other ideas about what might be causing this behavior?

Thanks,
( ! )