DOD: Discrete Germanium Overdrive Preamp 250

Why? Why not...  just a little experiment.   It sounds killer.

As I’ve been playing with discrete op-amps recently, I thought I might try a germanium one.  I was sure it would be awful but I felt I had to try anyway.  So it isn’t awful - works pretty well in fact.

Being a Distortion + based circuit, the hard clipping diodes do a lot of the work, especially at lower gain settings before the op-amp starts to clip - but at higher gain settings, the op-amp also clips, which is where the germanium should be adding some extra flavour compared to a regular op-amp.   

 

Sziklia pair instead of Darlington, using a silicon/germanium hybrid if you don't have any NPN germanium transistors - it works about the same (I also tested it on the breadboard), although you need a 2k2 resistor to get the collector to bias around 4.5v

 

Notes: 

• I accidentally used a 4k7 resistor on the Darlington emitter on the breadboard, which was subsequently used for the video grabs from my scope below - so maybe a little less opamp clipping than it should have provided

• Germanium PNP transistors:  Fujitsu 2SA267

• Germanium NPN transistors:  Sony 2SD128

• GE diode also works in the opamp - a very minor difference in compression 

• Clipping diodes can be any of your favourite hard clipping combinations (DOD 250 is silicon, MXR Distortion + is germanium).  I liked one of each, or put it on a switch for the best of both worlds

• The small capacitor usually found across the output and inverting input of the op-amp is not needed

• The 2.2n caps are what's found in the EQD Grey channel.  These would normally be 1n on a Distortion +, and a 250 only has a 1n across the clipping diodes

DISCRETE DOD OVERDRIVE ON THE SCOPE

Output

 

Opamp output

Fairly clean most of the way and then starts to clip at the end.  This was using a standard B1M pot.

Same again, but this time with the 3k3 on the collector of the Darlington, and a 4k7 instead of a 47k in series with the gain control.  I also have a silicon/germanium combo for the hard clippers, so the output level is reduced.

 output - quite a decent square wave coming out now.   A C500K pot seems to work best for the gain sweep.

LTSPICE SIM

The images below all use a 500k gain pot, with the R5 resistor changing values.  While the Distortion + uses a 1meg gain pot, it really has no impact on this, as gain increases as resistance is reduced.  

• Distortion + uses a 4k7 for R5, which allows for more gain, which also changes the frequency response at high gain settings (mid-hump)

• DOD 250 has a 47k, which means less gain, but it remains relatively flat by comparison

Increase the input cap C1 to 22n or 47n for a little more bass, and/or increase C5 up to 1u.  I think the stock values are pretty decent, as I usually prefer a fairly flat response.  

This is the response when changing C5 - as you can see, it has the largest effect at the highest gain setting.  In the next two images, gain is swept from 1 - 500k in 100k steps.  R5 is 4.7k

Same again, but this time changing C1 - you can see C1 has more effect at lower gain settings.  Noting that these were both done with R5 at 4.7k, if you are using a 47k resistor in this position, changing C1 will probably provide more benefits if you want a bit more bass.  

DISCRETE DOD OVERDRIVE PREAMP - VERO LAYOUTS