My attempt at a Tube Screamer Work-Alike

[Bill Mountain]

This is something that had been floating in my head for a bit.  Last night I finally got the chance to play with it on the breadboard. 

I had recently played a TS clone that sounded pretty good but was a little light in the lows.  Being a non-inverting opamp stage, the input signal is mixed with distorted signal but the mids are so pronounced that sometimes the lows get lost.

The idea was simple.  I wished that I had more control over the volume of the effected signal when mixed in with the input signal.  Loving simplicity in my pedals, I opted to limit the number of controls.  I could have done a clean blend or change the corner frequency of the clipping but I had hoped that by simply bringing the mids-boosted single a little lower in the "mix", it would hopefully give the effect of boosting the lows and avoid the need for a bunch of blend and master volume knobs.  I could also simply boost the input signal to get the kind of mix I want but that's a project for another day.

The next part of my idea was simply a pipe dream but it might explain some of my design choices.  I wanted to make the venerable discrete tube screamer (um...work-alike).

I set up Q1 (2N5457 biased to about 6V on the drain) by ear to be unity gain when mixed with the 10k mixing resistor.  I then wanted to take the distorted signal from Q1's source so as to not interfere with the "clean" signal.  The idea was to use a single FET or BMP clipping stage for the distortion stage to maintain phase when mixed but with the way Q1 is biased the source signal is not as strong as I'd like and the FET and BMP stages didn't have enough gain on their own to get the character of distortion I wanted.  I also tried using a single drain resistor for both transistors as a simple form of mixing the signals.  I'd seen this done when mixing channels on tube amps.  It technically worked but didn't sound that great.  It's worth experimenting with in the future though.

I then tried some CMOS inverters because I had the chip on hand and as expected, one stage wasn't enough.  But two were!  I tried it with diodes in the feedback loop of the second inverter but it just felt squashed.  Being two inverting stages I then tried it off of Q1's drain to maintain phase but then the "clean" signal suffered.  I'm guessing it was working too hard to feed the mixing resistor and the CMOS stages.  So I simply created one more inverter stage to correct the phase when mixing.

The end result?  It sounds pretty good but a little noisy on the breadboard.  I need to try it boxed up at gig volume to see if I can live with the noise.  The CMOS stages are understandably hot so I'm going to possibly pad it down some on the output.  With the introduction of the hex inverter, I could possible replace Q1 entirely and rework the mixing using extra stages like the ROG UBE Screamer.  It wasn't my original intent to use the CMOS chip.  I'd still prefer to use a transistor but I need to experiment with what would work best.

As always, my fingers work slower than my brain so I don't always make sense when I write.  So if something isn't clear please ask for an explanation.

[GibsonGM]

Well, it LOOKS neat, Bill!  Now - do you have a sound clip? > 

[Bill Mountain]

Here's a suuuuuper quick demo.  I play bass mostly and am not very good at guitar.  My sausage fingers can't fit in the tiny strings.

I recorded with the soundcloud app placing my phone on top the amp.  I used both my bass and my wife's old guitar.  The amp was a small Crate SS combo on the clean channel with the speaker out plugged into my Mesa 2x15 bass cab.  I play each instrument clean and then with the effect on.  I didn't record very loud so you hear me picking the strings and my phone vibrating when the bass is being played.

https://soundcloud.com/bill-mountain/sounds-from-saturday-morning/s-mvs4s