DIY-RE Colour Palette 500 Series Distortion Effects module

[nevetslab]

I recently took in a DIY-RE Colour Palette 500 series module, one of many products that mate up with the variety of API-format ‘lunch box power supply units that might have API 550 EQ’s, Compressors, and a wide range of non-API modules on the market these days. DIY-RE is DIY Recording Equipment, which offers very nice quality modules in kit form that you can build up following their decent destructions to guide those who at least know how to solder and follow directions, and if lucky to get it done right, it should work.



This Colour Palette module has three slots on the PCB, which had a 15IPS module, a TM79 module and a DOA-RED25 module. The 15IPS module has a 50Hz 2-pole HP Filter with a nominal 3dB head bump at 60hz, and a stacked array of diodes ahead of the gain stage. It produces soft symmetrical clipping along with the head bump typical of 15IPS tape machines. The TM79 module is interesting, in that it has a pair of small bridging transformers 1:2 to 2:1 between gain stages, with a pair of MosFET clamps in series across the middle of the xfmrs. It clips asymmetrically, offering a much different harmonic structure when driven hard. The DOA-RED25 module has a discrete Op Amp module, same pinout as an API 2520 Discrete Op Amp on board, which when overdriven, yields hard clipping. So, this module is set up with three pots, one per module to dial in the amount of overdrive you want, followed by a gain makeup pot.

The gain makeup pot should be +/- 6dB, with a 12dB range. But, I find it instead only having a 10dB range, and while mid-position on it’s detented pot is Unity Gain, it instead has -6.5dB attenuation at full CCW and +3.5dB at full CW.

My client, who’s one of the guitarists on the Late Late Show on CBS, managed to mix up a pair of output xstrs on the discrete op amp, and a of resistors elsewhere, so easy fixes. I only got a chance to play with the module this past weekend to see just what it does.

Began with powering it up, connecting my bipolar 15V supplies, balanced in and balance out connectors to the 15 pin edge connector, all having EZ-Hook connections. After verifying no surprises, I fed it 400Hz 0dBu signal and saw output on the main module, before plugging in any of the ‘colour’ modules. One by one, each worked fine. I took screen shots of the unit under signal drive, first at 500hz, then at 100Hz, showing the Distortion Waveform with the Overdriven sine signal, as well as it’s harmonics on the spectrum analyzer.



Then, set up the Bode Plotter to run frequency response plots on the modules, first at unity, then under 2dB Overdrive, then at 3.5dB Overdrive. Of course you can really crank up the drive, but it doesn’t have the correct resistor values in the Gain makeup stage.



I'll have to add the schematics on the next post, as I've hit the limit on this post.

I did get some of the schematics, though not that for the RED-25 Discrete Op Amp. I might get around to drawing it up if I have enough time before my client comes by to pick it up.

[Helmholtz]

that you can build up following their decent destructions

Is this an intended pun?

[nosaj]

Is this an intended pun?

Not if I was building one

nosaj

[nevetslab]

Is this an intended pun?

Yes.

[J M Fahey]

Very interesting distortion behaviour.
The first screen capture shows a "tubier than tube" almost perfect frequency doubling, riding on fundamental remains, practically what we expect from toobz (and not always get), all that without visible clipping ... incredible.

Then waveform gets hairier, of course.

Waiting for your schematics to see how they do it.

EDIT: oh, NOW I see it, it´s not a modular Guitar effects box but something designed to imitate old recording equipment quirks ... including tape saturation ... WOW!!!

I had thought all 3 captures came from same module driven harder and harder, now I see they are different.

Still interested in them, but specially the first one

[Helmholtz]

almost perfect frequency doubling

I see frequency tripling as result of symmetrical clipping. (Doubling results from asymmetrical clipping.) Also shows in the spectrum.

[nickb]

I don't see doubling or tripling. I see the fundamental with varying degrees of harmonics but the fundamental still dominates. I suspect you might be getting confused by the scope shots of the error signal.

[Helmholtz]

I don't see doubling or tripling. I see the fundamental with varying degrees of harmonics but the fundamental still dominates. I suspect you might be getting confused by the scope shots of the error signal.

I was referring to the first scope shot. It shows a symmetrical (soft) clipped signal and a second signal (error signal?) with a dominating 3rd harmonic as shown by the number of zero crossings. The spectrum reveals higher odd harmonics and low level even harmonics as well.
(Perfectly symmetrical signals don't contain even harmonics.)

[nickb]

I was referring to the first scope shot. It shows a symmetrical (soft) clipped signal and a second signal (error signal?) with a dominating 3rd harmonic as shown by the number of zero crossings. The spectrum reveals higher odd harmonics and low level even harmonics as well.
(Perfectly symmetrical signals don't contain even harmonics.)

Don't get all upset - I was replying to JMF not you HH This is where the linear thread view can hinder.

[Helmholtz]

What I meant to say is that the equation: nicely rounded symmetrical signal = prevailing low order even harmonics = tube sound - is wrong.

[nevetslab]

15IPS Module & TM79 MNodule 1dB O/D screen shots

This morning I've lowered the drive level thru the two 'colorful' modules, having seen what they do if cranked. I never did invest in a small power supply mainframe that accepts API and other similar formatted modules, or I already would have had this back home to patch into my work station to give a listen to what it does on various audio tracks. All I have here at the shop is my bass, though I do have an SVT4-Pro I could patch it into, which I'll get to before my client stops by to pick it up.

I'm waiting for his contact info to speak with one of the chaps at DIY-Recording Equipment regarding the Gain Trim range that seems wrong, though they did get the Mid-Point gain trim pot setting to be at Unity Gain, so it matches when you switch the module in / out. They just didn't get the min/max range centered, so you'd have +/- 6dB spread, offering more gain make-up ON the module, instead of having to do that post-patch in the audio chain. I've varied the circuit values some by calculations, but haven't thought it thru enough to arrive at the proper formula to set the min, mid and max attenuation values properly. After that, it's a simple matter of re-setting the gain in the follow-up amp circuit.

Here's this morning's screen shots....beginning with the 15IPS module, then the TM79 module. I set the effects' level where it lowered the output 1dB, then increased the gain trim to restore it. The scope shots show the Output waveform and the Distortion waveform (both around 0.3% THD via Amber 3501a), then the harmonic spectrum (via HP 3580A):



When I ran the frequency response plots using a Vellman PCSGU250, with the Amber 3501a serving as the balanced receiver, I had the 80kHz LPF engaged, hence the HF rolloff on the plots. The Vellman's signal to the DUT was by way of my HP 353A Attenuator/Xfmr patch panel, and that rise at the HF end is from unterminated xfmr. The Sine signal source is the B & K 1027, which I was using for both Pink Noise & Sine during listening tests.

Interesting behavior on the TM79 module, where at minimum signal drive (not increasing the drive level to the module), the HF bandwidth is substantially greater than when you crank up the drive level. Look at the other graph in the first post to see the difference.

[nevetslab]

Diff Freq Distortion 2dB O/D on 15IPS, TM79 modules

I haven't used the B & K 2010/1902 Analyzer that much lately, as my plotter pens have dried up, and they're OMG expensive these days to replace.
But, after manually sweeping the 15IPS & TM79 modules watching their individual harmonics over frequency (one of the neat things it offers), I then changed to Difference Frequency Distortion mode, where it puts out two equal level tones, and you can set the Diff Frequency on the dial of the Distortion unit. I set it for 700Hz and 800Hz, and ran the spectrum analyzer out to 2khz, since you will see harmonics on both sides of the two fundamental tones. Big difference between these two modules in this mode. I set the two modules up for 2dB Overdrive, engaging only one module at a time on the images.