This week on the Breadboard - FET / Germanium Boost

Chuck D. Bones

Circuit Wizard
This one has been festering for a couple of weeks now. I finally landed on the right combination of gain, headroom and freq response. Many JFETs will work, but I had a pile of 2N5246's so I ended up using one of those. Voltages are marked in purple. These voltage yield the maximum clean headroom. The JFET I selected for Q1 has Vp = -1.4V. Q2's HFE is just over 100 and the leakage is 130μA. Use whatever NPN Ge you like. Q2's bias is very stable in this circuit. R6 can be tweaked to dial-in Q2's collector voltage. All caps are film except C2 is tantalum, C5 is silver-mica, C6 & C8 are aluminum. With the VOICE switch in the center position, freq response is flat and gain varies from 11dB to 40dB. It runs clean with GAIN below noon. At the higher gain settings, bass & treble roll-off a little bit. When the VOICE switch is in the down position, the circuit becomes a treble booster. Bass rolls-off below 450Hz. C3 determines the roll-off freq. It runs clean with GAIN below 10:00. When the VOICE switch is in the up position, the gain is increased 10dB across the freq range, with a gradual bass roll-off below 200Hz. C4 can be increased for a fatter bottom-end in the high-gain setting. I used 2.2uF because I was going for loud and distorted but not muddy. Maximum mid-band gain is a respectable 50dB. The high input impedance prevents pickup loading and plays nicely with other pedals.

FET-Ge Boost v2.2.png
 
Wow. This is really impressive! Great work with the frequency response design. Those gain figures are wild. You mentioned stable bias for Q2—is this the booster companion to the servo fuzz face?
 
You could stack it with whatever you like. This circuit did not need a servo because it is inherently stable. The large voltage drop across the emitter resistor (R9) and the relatively low resistance of the bias network (R6 & R7) makes Q2's bias point insensitive to variations in HFE, Vbe and Iceo (leakage).
 
Last edited:
This one has been festering for a couple of weeks now. I finally landed on the right combination of gain, headroom and freq response. Many JFETs will work, but I had a pile of 2N5246's so I ended up using one of those. Voltages are marked in purple. These voltage yield the maximum clean headroom. The JFET I selected for Q1 has Vp = -1.4V. Q2's HFE is just over 100 and the leakage is 130μA. Use whatever NPN Ge you like. Q2's bias is very stable in this circuit. R6 can be tweaked to dial-in Q2's collector voltage. All caps are film except C2 is tantalum, C5 is silver-mica, C6 & C8 are aluminum. With the VOICE switch in the center position, freq response is flat and gain varies from 11dB to 40dB. It runs clean with GAIN below noon. At the higher gain settings, bass & treble roll-off a little bit. When the VOICE switch is in the down position, the circuit becomes a treble booster. Bass rolls-off below 450Hz. C3 determines the roll-off freq. It runs clean with GAIN below 10:00. When the VOICE switch is in the up position, the gain is increased 10dB across the freq range, with a gradual bass roll-off below 200Hz. C4 can be increased for a fatter bottom-end in the high-gain setting. I used 2.2uF because I was going for loud and distorted but not muddy. Maximum mid-band gain is a respectable 50dB. The high input impedance prevents pickup loading and plays nicely with other pedals.

View attachment 18468
How about This :

Ge-Fet Boost Mockup Pedal.jpg Ge-Fet Boost Mockup Pedal.jpg
 
Last edited:
A couple of days ago, Cooder asked me if MP38A would work in place of 2N1308. Turns out the answer is "yes." I tried 2N1308, MP38A and AC127. The 2N1308 had the highest HFE (104) and lowest Iceo (0.088mA). The AC127 had HFE = 72 and Iceo = 0.22mA. The MP38A had HFE = 61 and Iceo = 0.17mA. In-circuit, their collector voltages were within 80mV of each other. There was no need to tweak the biasing. They all sound good; there might be some subtle differences. Hard to tell for sure.
 
And here's the Mountain Oyster Germanium Drive aka GE-Fet boost. I thought my name choice sounded a bit more gutsy and tasty ;)
Gutsy and tasty a drive'n boost with attitude it is. Noice.
MP38A Ge trannie and a 2N5457 jfet that had about the right Vp (-1.3 V)

NBWAHoy.jpg


FYfxhwc.jpg


sBttF4T.jpg


JwSJytT.jpg
 
I don't breadboard enough things, so this seemed like a nice easy one to throw together. I'm sitting on an assortment of JFETs and GE transistors so this gave me a chance to try a few different ones out. I REALLY like the way this sounds. I settled on using a 2SK30A-D w/ a Vp of -1.4V and an Idss of 2.31mA. To dial in the voltages I changed R4 to 7.5K and R5 to 1.5K. That gave me 5.2V at the drain and .80V at the source. I tried out MP38As and some Flat Hat transistors and had a bit of interesting (to me) results. I used varying HFEs from the 50s - 100s and for some reason I have one MP38A with an HFE in the 60s that comes across as way more aggressive than some of the ones with higher HFEs (leakage in general is pretty low for all the ones I tested). Voltages are 3.08V at the base, 6.06V at the collector and 2.96 at the emitter. I was lazy and just threw a trim pot on there for the gain knob so it goes it goes from boost to overdrive with the most minuscule of a change, but that'll get fixed when I put it together for real. This circuit is simple enough that I may want to give a go at putting together my first PCB so long as you don't mind. I'm sure I'll post dumbass questions about it in a different forum. Thanks for this one, I was pretty excited with how it sounds.
PXL_20220327_220846202.jpg
 
Glad you like it Flemming. Clean breadboard! As you found out, the circuit is very flexible and will work with a whole range of JFETs & NPN Ge transistors. If you want to design a PCB, go for it. Cooder was the first out of the chute with one. Maybe you'll be next.
 
Back
Top