Calibration issue

[flel]

Ok so to narrow down if this is indeed a calibration issue I created a table for each of the CV output channels (columns) for the lower 4 and upper four octaves. These are the results. The first number (only recorded for the lower octaves) shows the current CV value (noted from current values screen) and the second number shows the midi note registered by uTune. 

Clearly, the CV values are off for track 1. Does this confirm that it is a calibration issue?

Interestingly, CV out track 1 "recovers" in the last two octaves whereas CV out track 4 performs great for the first 7 octaves but falters at the 7th octave. Are these calibrated separately somehow?

---

Posting values from the calibrate menu for the two with wrong outputs (tr1 and tr4) and a reference (tr2)



I am not brave enough yet to change anything here or to start twisting any potentiometers on the back. I do have a 1104X-E oscilloscope I can patch directly from CV1 out so if necessary I can probably fix this. Does it look like I need to follow the steps outlined in the calibration section of the manual?

EDIT: staring at this a bit longer, I may not need to touch any of the potentiometers since it looks like the zero point is calibrated decently OK? This is also confirmed by measuring using the scope since I am getting e.g. 4.92V where it should be 5V for tr1 (which if off pitch) versus 4.98V for tr2 (which is on pitch). So I could fix this by just changing the values for 2-8V in track1 and the values for 7-10V for track 4?

[mgd]

I have calibrated all my 3 CV16 using a decent (5 1/2 digits precision) Voltmeter. All I had been doing is change the numbers on the calibration screen. I doubt it generally would be a good idea to mess with any potentiometers unless you really know what you are doing.

What I found is:
- the calibration values are similar but not identical for all 16 channel and across my 3 CV16 NSA
- stability of my tuning (aka the calibration) also depends on the quality of my power supply and whether the total power load of all the modules that share the power with my NerdSEQ is nearing the maximum load of the power supply. [I found that staying below 75% max load is a good upper boundary - your mileage may vary].

When I had to deal with tuning issues on my NerdSEQ I got rid of them by replacing power. After that I did another (and so far last) calibration of all CV16 and am happy with the result since then. That's now a couple of years ago.

I don't recall that I ever calibrated the CV1-6/MOD1-6 of the main module. At the time I was using solely the CV16 [writing about this I actually may look at them and check :-) ].

Kind regards,
Michael

[XORadmin]

Just use a decent multimeter (3 digits after the dot for all ranges) as described in the manual. Calibration happens per octave and per output (which explains different numbers per output and per octave) so indeed it can be fine for all octave except for one on one output if not calibrated well.
A regular multimeter can improve things as well if you don’t have anything else.

Also I can see that someone changed the calibration values already at least once, so there is a
good chance that this didn’t happen completely as it should be.
And don’t twist the potmeters on the back, that is not necessary.

I am sure it’ll be fine again after fixing the calibration.

[flel]

I only have a 5$ multimeter and a 500$ oscilloscope... perhaps I should get a better multimeter?

I realized that I could actually calibrate the outputs simply using the uTune, which worked perfectly. I trust it more than I trust my Siglent oscilloscope for voltage detection and was able to get it "good enough" tuned in 12 notes per octave. Perhaps I'll try setting that to 24-EDO or something even more granular and see if it still tracks.

I bought the NerdSeq used, so perhaps someone did play with the calibration settings before I did? I wonder if the pots got turned because i am consistently getting negative voltages at 0V using the oscilloscope, but this might be a measurement problem? I imagine a good multimeter would be better for this.

Do you know if there is a multimeter probe adapter for 3.5mm TS modular cable? That would make the measurement a lot easier. Currently using one of these for the oscilloscope which is quite useful: https://www.threetom.com/product/wiretap...t-adapter/ .. if only the voltage readings were reliable.

Thanks for the support and the amazing machine!

[XORadmin]

The adapter together with the scope should be fine! (If it got 3 digits after the dot)
Negative voltage is definitely possible and I am sure the potmeters have been turned slightly. But don’t touch them.
Just change the value to get 0 volt.  And for every value it’s always the first reachable value above the desired one (eg for 1 volt it could be 1.002 V if the one before was 0.999V)

Don’t use the uTune for it.

[flel]

So the issue is that if I tune based on voltage, uTune doesn't properly quantize. I believe that uTune is working properly because if I send CV from other devices to uTune, there are no issues.

Oddly enough, lets say I send a C-5 (note 60): measuring voltage from another sequencer gives voltage slightly below 5V (4.98V) and quantizes to note 60, whereas bringing nerdseq down to 5.01V the quantization drops to note 59.

Edit: picture of the pots in the position I found it.

[XORadmin]

Forget the pots. It can be any position as it can turn 24 times.

Measure again while the cable is connected and check if the voltage drops. If, then you got a hardware problem. Are there differences with the uTune inputs?

[flel]

>Measure again while the cable is connected and check if the voltage drops.

What do you mean "white the cable is connected"? I've only tested with the cable connected? There do seem to be minor voltage fluctuations when measured using the oscilloscope, but maybe this is noise?

[XORadmin]

>Measure again while the cable is connected and check if the voltage drops.

What do you mean "white the cable is connected"? I've only tested with the cable connected? There do seem to be minor voltage fluctuations when measured using the oscilloscope, but maybe this is noise?

I mean measure the voltage with and without the uTune connected. I am curious if there is a significant voltage drop.
In which range are the fluctuations?

[flel]

So the range on the fluctuations were in the 0.01V range, which seemed too high. However, I suspected that the continuous measurements of the oscilloscope wasn't sufficiently accurate for static measurements, so I bit the bullet and bought a proper 4.5 digit multimeter. Indeed, after doing this I was able to successfully calibrate the NerdSeq CV outputs. The uTune "estimates" were within range (and more useful than the oscilloscope) but a well-calibrated multimeter was needed for reliable measurements.