Some more measurements (and possibly a HW bug)


#1

Hello,

I decided to extend a little bit the measurements of the MOD Duo I’d done here:

The additional measurements were performed with MATLAB and a RME Fireface sound card. 2 channels of the RME were used, one for the MOD Duo measurement and the other one as a loopback to get a reference value for the gain and delay. What was performed:

  • Impulse Response
  • Magnitude Response (Sprectrum)
  • Harmonic Distortion

The pedalboard used was initially a simply bypass IN1 -> OUT1 and IN2 -> OUT2. Here are the results:

Pedalboard

1. IN1->OUT1

2. IN2->OUT2

A few things can be noticed here. First of all the delay is indeed around 8ms, as already measured in the link above. Alos, as you can see, there is clearly a problem in the IN1->OUT1 path. A huge distortion is present in the high end (starting at 1kHz) and there is a unexpected gain of 7dB.

In contrast, the IN2->OUT2 path has a much cleaner spectrum. However, the -3dB gain is also unexpected.

The problem could be either in IN1, OUT1 or both. In order to mitigate that, I modified the pedalboard slightly and performed the same measurements for IN1->OUT2 and IN2 ->OUT2, still with only a wire connecting them.

Pedalboard

3. IN1->OUT2

4. IN2_>OUT1

The conclusion from this cross measurement is that there is a problem with IN1, which could be related to the hardware. It would be good to know whether:

  • Someone else also had a problem with IN1 or experienced some strange behavior there

  • Why is there a -3dB gain in the “clean path”

Currently as a workaround I can simply avoid IN1 and add a gain module to compensate for the 3dB difference.

Best,


#2

In1 is known to be contaminated with some hardware noise. Search for the topic “hardware noise”


#3

Hi Azza,

Thanks for the hint. I didn’t know about that. I will send an email to the MOD team.

Best


#4

Here is a link to the discussion


#5

Hello again,

As advised by MOD team, I replaced the components according to https://docs.google.com/document/d/1fXsYRYFi1qsdwhhonAtrNglmctbIObb2VlB68vyFh8M/edit

After doing this I repeated the measurement for IN1 -> OUT1:

As you can see the result is much better. However, considerable harmonic distortions is still there: in the range of -40dB to -60dB. Moreover, the ~10dB difference compared to IN2 -> OUT2 is sill there.

@ricardocrudo : any idea on what is going on? Are there other workarounds to be done?

PS: IN2 -> OUT2 measurements remain roughly the same


#6

@edwillys what is the Duo’s gain setup that you used to make these measurements? Did you run the tests connecting the Duo directly to your PC, or there was an interface in between? What software are you using and what procedure?


#7

The discussion also mentions that some of the noise is due to the audio driver. So there is a software part to the problem.


#8

Hi Ricardo,

After checking the Duo’s settings, input1 (mid) had indeed a different stage than input2 (low). This might explain the gain difference. I’ll measure it again tomorrow and keep you posted.

As I tried to describe in my first post, I’m measuring with MATLAB scripts. The Duo is connected to the PC via a RME Fireface sound card.


#9

Hello again,

Some interesting news. After setting the Stage setting to Low, the measurement is as neat as for input 2. Moreover, the gain difference is also gone.

The problem seems to be related to the Stage setting alone, regardless of the input. For example, if I set it to High, it gets even worse:

Workaround is easy at least: use only the Low configuration. I reckon this might still be a problem for instruments with low impedance.

@ricardocrudo: is there some impedance matching circuitry that depends on the Stage setting or is it purely a SW related issue? In my measurements I connect the sound card directly to the MOD, without any load. Anyway I wouldn’t expect the harmonic distortion to change that drastically.

Update
The noise seems to be coming from input clipping. When I drastically reduce the sine sweep gain (-10dB), the measurement is clean. Nevertheless, if I reduce it just enough for it to be in the same level as the “low stage” setting, i.e., roughly -6dB, this huge distortion is still there, which I wouldn’t expect:

Final test would be to measure the input of the DAC with a scope and see if it is clipping there already. However, I assume I’m too lazy for this :slight_smile: For my use case (guitar), I shouldn’t be reaching levels that high on the input, so it is most likely not a problem.


#10

The main problem is when using microphones directly into the mod (without pre-amp).