FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 

PolyDAC Scaling Problem
MUFF WIGGLER Forum Index -> Oakley Sound Systems  
Author PolyDAC Scaling Problem
After putting a few new modules into my cabinet I tried out my PolyDAC. When patching into my 1V/Oct ins on my Classic VCOs I noticed nothing being out put of the VCOs. Messing around with the tuning and lowering the octave of my controller I could hear a high pithed wine. Immediately I thought scaling problem. On opening and going through the set up section I notice that I could not scale down the voltage of the DAC out to +10.66 volts. It seems stuck at 13.66 volts with the jumper in place. At the moment I have only tried replacing the AD8300 but that seems to not have fixed the problem. Is the problem the micocontroller?
Backstory, after firing up the synth after installing one module I shorted V+. I'm a little worried I may have fried the microcontroller on this one. Is there any way to confirm this? The keyboard is varying the voltage albeit it several volts higher than needs be. Any help would be much appreciated.
How many volts should I see out of the DAC out of the 8300?

I'm so incredibly angry I did this. I need to build or purchase a bench top PS for testing purposes.
After some probing I'm seeing around 2v on the DAC OUT. The Output of the first half of the AD712 (pin 1) is only swinging about a volt from 10.43 to 11.31v. Approximately how much voltage is needed to affect scaling. I switche out the mux/demux chip and it seems to not have done anything. How much voltage should be applied to pin 1 of the 4067 to affect scaling? Perhaps I'm following the wrong path and should be looking at demix pins of the avr. The more I think about it I think this avr is fried but how do I confirm this?
I should say that I'm not that familiar with the polyDAC. It's really Paul Maddox's project. I sold the board for a short while and was involved in the first beta tests and edited the documentation for Paul. Then Paul made the project open source and Laurie Elby took up the design and added to it.

In terms of any discrepancies between the schematic and the part list, I'd believe the parts list. eg. scale trimmer is more likely to be 1K.

I don't think shorting out the 15V rail could have damaged the AVR. It's driven from a 5V regulator on the polyDAC board which also has a current limit resistor.

Do the gate outputs work? If they do the AVR is probably working.

I'm not sure what exact voltage you should be seeing from the DAC when in calibration mode (J5 shorted). However, in normal operation it will be the multiplexed signal - ie. a jagged looking sample and hold style waveform and never exceeding 2.5V. That 2V figure may be OK if you are measuring it with a meter but it's hard to tell because the shape is more important rather than the average value.

An output stuck at 13.6V suggests that particular op-amp has 'railed' - ie. stuck at it's positive power supply rail. That means either the op-amp has fried or that the signal driving the op-amp in massively incorrect. It's probably the latter in this case.

The output from U8 could be wrong. I would expect this to be approximately five times that from the DAC. But again it's the shape that's more important rather than the voltage measured by a DVM. Do you have a scope to hand?

Does the aftertouch work OK? If this too is stuck high at 13.6V and the gates don't work then the AVR is probably not working. If the gates do work and the aftertouch is stuck high then I'd suggest a fault with either the 4067 or U12.

Thanks Tony!!!
Gates, pitchbend, modwheel, velocity...all seem to be working properly. I just purchased a new controller so I'm not entirely sure that the aftertouch is set up correctly but using a GM controller patch I don't see the polyDAC responding to aftertouch.

I do have a cheap 20MHz Protek that is okay for looking at waveforms but not entirely great for looking at some other things. I'm quite a novice using one. However probing at U8.8 I see something like you describe. It look like a pulses of three various heights. I just went through a tube of 074 so I replaced U11 with an 054 but it didn't seem to do anything. You mentioned replacing U12 is this correct or did you mean U11 the quad op amp set up as 4 voltage followers?
Right if PB, modwheel, velocity and the gates are working, the DAC, the AVR, U12 and the 4067 are probably working just fine.

The aftertouch should be working if the PB is. Maybe it's not being sent... but it will be worth checking this for sure.

If it isn't working it's maybe worth swapping out the 4067 but I'm not so sure it is that.

What's the voltage of the mod wheel output? It should go from 0 to 10V I think.

What about the pitch bend? Can you get this to zero properly?

What's the output of U17, pin 7? This should follow the pitch CV but backwards, ie. -1V/octave.

Swapped out the 4067 for an OnSemi version I had on the shelf which resulted in similar problems.

For PB I can zero it out and it's swinging from +/-0.7V

Mod Wheel swings from ~0.0-11.1V

The strange thing is the octave seem correct. At U17.7 I'm seeing jumps from around 1V/Oct. -1.70, -2.77, -3.85 (I think this is C1) and -4.91.
Are your pitch CV outputs also moving in the right direction with more or less 1V/octave too?

Is your mod wheel down? The polyDAC adds mod wheel voltage to the pitch CVs if I remember correctly. So make sure any wheel is fully down when you try to calibrate the DAC.

Just to make sure....when you mentioned measuring U17.7 were you speaking of the half of the LF412 set up as an inverting amp for NOTE_CV_0? The IC following the 10K resistor network (RP3)?

I tried calibrating it with and without the controller attached with similar results. And yes I did made sure to have the mod wheel down when the controller was attached.
Paul Maddox is on facebook, and is a very approachable guy. You might want to contact him and ask for some help.

Pitch is generally about 1V/Oct. Rather 1.06V/Oct. It's close enough I guess for most of the stuff I do but it still gets to me. It also bothers me that if in the future the scaling goes completely out of whack I won't be able to calibrate it.
keninverse wrote:
Pitch is generally about 1V/Oct. Rather 1.06V/Oct.

Try this, I'm guessing here, but it's worth a go.

Change R21 to 11K and the trimmer to 2K. It may even calibrate with the trimmer as it is, at 1K.

I'm not sure why the jumper calibration setting pushes the final op-amps into the rail though. For the midiDAC I calibrate using the highest and lowest notes on your controller. If the lowest C note is set to 0.00V the highest one should be 10.00V, ie. 10 octaves above.

I replaced the trimmer and the resistor and I'm seeing the same results. I started a conversation with Paul Maddox and he's suggested taking a look at the S&H circuits to see if everything is working properly. Particularly using a scope to take a look at what's happening with the S&H caps and the outputs of the individual NOTE_Ns op-amps.
Just and update. For whatever reason the note cvs are still sitting at maximum voltage of around 12V after I put the module in calibration. Dave Brown and Paul Maddox helped me check a few things and it seems the CC CVs values are the proper voltages. So I just used those voltages to calibrate and scale the module. I suppose there may be an issue somewhere in the Note CV summing when the module is put into calibration but I don't see any voltage going into the resistor networks after tuning is zeroed...Pitch bend and mod wheel is sitting at zero. Everything seems to be working correctly now and the module tracks 1.0V/Oct so I suppose everything is okay. Thanks for the help!
MUFF WIGGLER Forum Index -> Oakley Sound Systems  
Page 1 of 1
Powered by phpBB © phpBB Group