Monday, 16 April 2018


Even if most mixers are able to cope with the typical audio voltages a modular synth work with (around 5V peaks), professional line level is defined as typical at +4  dBu (or 1.7 V peak).   So it can be good to have an adaptation module with a bit of attenuation and an output buffer capable to drive the line from the synth to a mixer or an audio interface.
Moreover, using 6.35 mm jacks is more practical to interface with external gears than keeping 3.5 mm jack.

My output module

I got inspired by Make Noise's Rosie : cross fader between A and B inputs, send and return functions and the attenuation stage to match line level.

Make Noise chose to have the levels of the send-return at modular level, to use with effects in the modular synth itself (like their own reverb module).   I decided to go with variable attenuation and gain in the send-return path to be able to use it with external effects like guitar pedals.
As a result, the return jack can still be used as a third potential input for the module, but it might need some attenuation beforehand in order to match the levels of the A-B inputs.

Messing around on the breadboard

The electric schematic is an assembly of three circuits, with slight modifications.
Ouput module electrical schematics

First, the crossfader at the input comes from the panner schematic from R.G. Keen.  I modified the resistors value to match the resistors and the potentiometer I already had. 
Second, the output stage is the output stage from Dintree Output mixer.  I removed the R69 resistor to better match my target output level.
Third, the send-return circuits are Ken Stone's stomp box adapter with variable attenuation and gain.  I removed the 10uF capacitor in the return feedback loop as it created an potentially audible high pass filter at minimum gain.

Finished circuit on the bench

And here is the final layout.  I had to correct a couple of slight mistakes but other than that it went smoothly.

The layout

I elaborated my gain chain to have +4dBu (around 1.7 V peak signal), if I enter a 5 V peak signal on one input and keep level potentiometer at two third.

As there are three op-amps in the audio loop, I made sure they were all in an non-inverting configuration to avoid phase inversion.

I added a peak detector as a simple comparator followed by a rectifier, a filter and the diode.  Classic.  This will detect peaks above 5 V (or 5 dB below the +18dBu clipping level) on the positive side only, but that will do.
Why this value ?  I first decided for a much lower 1.8 V peak.  But I reckoned it was far too low and that I should target a value around +12 dBu.  I decided to change a single resistor and I only have that 15k instead of the needed 12k.

If you're interested in the relationships between V, Vrms, dBu, dBV, I found this very useful website :

View from the back

Ongoing panel

The panel is one of the most cramped I had to do. 
Notice the small id tag denoting this is my 20th DIY project.

Wire mess

Wiring was a mess.

8 HP was a fine choice.   But for the future, I have to resist the temptation to cram as much function as possible into the available space.

I first wanted more vintage look knob caps, but they were too large and I switched back to traditional copies of Davies 1900H.

Finally, here it is in it's natural habitat.

Final view


  1. Thank you so much for sharing !
    I wonder if the protection diodes are the right way on your schematic "output module rev3"?

    1. Thanks for the comment. Yes you're right : protection diodes were reversed. the layout was right though. I have corrected it in rev3.1. Thanks again. Take care.

    2. Is the corrected rev3.1. layout available somewhere? Thanks for the awesome layout!

    3. Hi, thanks for the interest.
      The layout is OK. Only the schematics had the diodes reversed.