SRPP input gain staging Q

[mdcmdcmdc]

Hello smart folks... I've been messing around with a circuit on breadboard and I have it sounding pretty good on the breadboard.



Nothing really new to see, it's Mr. Keen's SRPP from geofx with a resistor added and LND150 depletion MOSFETS instead of JFETs.

The circuit sounds really nice with these devices, but i'm trying to figure out a way to halve the signal coming into the circuit (hence the 10k/10k voltage divider after the cap) while maintaining a relatively decent input impedance.

I've tried running series resistors of various values directly from the input to the "gain" pot, but this current arrangement seems to sound the best overall. I'm a bit out of my depth in terms of what's actually happening before the signal hits the gate of the first MOSFET, so any advice on how to simplify/ clean this up would be much obliged.

[amz-fx]

The base of the BC549 is tied to +9v. No sound will come out of it that way... connect to the source of Q3, or the drain of Q4.

regards, Jack

[mdcmdcmdc]

ack - that's just a typo in the schematic! Everything is working fine on the breadboard.

EDIT - fixed the schem, thank you for the catch!

[ElectricDruid]

You could try sticking a 100K resistor ahead of the A100K gain control. That'd halve the amount of signal the gain control deals with, and it'd be adding to the impedance rather than reducing it, like the 10K/10K divider.

[Ben N]

Seems like a waste feeding LND150s with 9V.

[antonis]

Also, R6 (10M) is unneeded..

[PRR]

.....trying to figure out a way to halve the signal coming into the circuit ...

Why?

This is a HIGH gain circuit. It IS gonna distort.

[Fancy Lime]

I happen to have almost the same thing on the bread board at the moment but with a different gain control. Try replacing R23 and R19 with just one 1M resistor and add a 1M pot wires as a variable resistor in series with C20. Then turning up the resistance of that pot will make the push-pull action less effective and reduce the gain. Less noisy than dumping signal to ground on the input and then boosting the living freck out of what's left.

Edit: or run it at a few hundred volts. Keeps the high gain but increases the clipping threshold.

[mdcmdcmdc]

.....trying to figure out a way to halve the signal coming into the circuit ...

Why?

This is a HIGH gain circuit. It IS gonna distort.

Without substantial attenuation, the range of sounds isn't super usable for me. With the voltage divider in place it has a really nice range and sounds great.

[mdcmdcmdc]

I happen to have almost the same thing on the bread board at the moment but with a different gain control. Try replacing R23 and R19 with just one 1M resistor and add a 1M pot wires as a variable resistor in series with C20. Then turning up the resistance of that pot will make the push-pull action less effective and reduce the gain. Less noisy than dumping signal to ground on the input and then boosting the living freck out of what's left.

Edit: or run it at a few hundred volts. Keeps the high gain but increases the clipping threshold.

I'll give this a go - thanks!

[amptramp]

Your distortion problem isn't at the input to the stage, it's at the input to the emitter follower.  SRPP is a totem pole circuit.  The current in the upper stage moves in the opposite direction to the current in the lower stage due to the bias developed across R20.  If the transconductance of the upper stage is 1000 micromhos, the current in the upper and lower transistors move in the opposite direction by the same amount, which is the most linear choice.  But the difference current between the upper and lower stage has to go through the load.  An SRPP circuit cannot operate linearly without a load with low enough impedance to take the difference current and a BC549A has a minimum hfe of 110. which would give a load of 363 K minimum.  This could only absorb a difference of 24.7 µA to go from 0 VDC to 9 VDC.  You need a lower-gain transistor or a lower emitter resistor than 3300 ohms to keep this circuit linear.  Don't forget, 110 is the lowest hfe for a BC549A.  If you have a BC549B or BC549C, the lowest hfe is 200 or 420 respectively.  Even a BC549A can have an hfe of up to 220 which would double the input impedance.

[antonis]

A low value resistor from Base to GND shouldn't help, Ron..??

[amptramp]

It would help but it would shift the bias point which may or may not be useful.  I have always thought a great fuzz/distortion could be made by hanging a 500 K audio taper pot with the slider going to ground from the SRPP output to ground with some series resistance.  At low resistance levels, the circuit would be linear.  At higher levels, it would be non-linear.  At the full resistance, it would be almost squarewaves.  I would build it into a pedal and call it "Spongebob Squarewave".  But you would have to AC-couple the load to get the full range or the bias shift would be bad news.

[antonis]

It would help but it would shift the bias point which may or may not be useful.  I have always thought a great fuzz/distortion could be made by hanging a 500 K audio taper pot with the slider going to ground from the SRPP output to ground with some series resistance.  At low resistance levels, the circuit would be linear.  At higher levels, it would be non-linear.  At the full resistance, it would be almost squarewaves.  I would build it into a pedal and call it "Spongebob Squarewave".  But you would have to AC-couple the load to get the full range or the bias shift would be bad news.

Is that a promise or what..??

[mdcmdcmdc]

It would help but it would shift the bias point which may or may not be useful.  I have always thought a great fuzz/distortion could be made by hanging a 500 K audio taper pot with the slider going to ground from the SRPP output to ground with some series resistance.  At low resistance levels, the circuit would be linear.  At higher levels, it would be non-linear.  At the full resistance, it would be almost squarewaves.  I would build it into a pedal and call it "Spongebob Squarewave".  But you would have to AC-couple the load to get the full range or the bias shift would be bad news.

I would love to see this drawn up!!

[mdcmdcmdc]

Your distortion problem isn't at the input to the stage, it's at the input to the emitter follower.  SRPP is a totem pole circuit.  The current in the upper stage moves in the opposite direction to the current in the lower stage due to the bias developed across R20.  If the transconductance of the upper stage is 1000 micromhos, the current in the upper and lower transistors move in the opposite direction by the same amount, which is the most linear choice.  But the difference current between the upper and lower stage has to go through the load.  An SRPP circuit cannot operate linearly without a load with low enough impedance to take the difference current and a BC549A has a minimum hfe of 110. which would give a load of 363 K minimum.  This could only absorb a difference of 24.7 µA to go from 0 VDC to 9 VDC.  You need a lower-gain transistor or a lower emitter resistor than 3300 ohms to keep this circuit linear.  Don't forget, 110 is the lowest hfe for a BC549A.  If you have a BC549B or BC549C, the lowest hfe is 200 or 420 respectively.  Even a BC549A can have an hfe of up to 220 which would double the input impedance.

Thanks again for this - in the absence of a low(er) gain bjt in easy grabbing distance, I just flipped the BC549C around (which should reduce it's hfe by around 80-90% iirc?) which 100% worked.

[Fancy Lime]

I've been toying with an idea similar to Ron's but never got around to actually try it out: using an inverting opamp after the SRPP. With an inverting opamp you can choose the input impedance quite freely by using an input resistor of choice. At the same time, a variable resistor in the feedback loop makes a good volume control while keeping the nice low output impedance of the opamp independent of volume setting. Food for thought, maybe.

[mdcmdcmdc]

I tacked a simple gain stage (cribbed from the Spaceman Saturn) in front of it and it's sounding quite nice. I'm trying out the tone control from deadastronaut's Spitfire with it, but I think that might be a bit lossy for my purposes.

[mdcmdcmdc]

Per the chat upthread, I changed the gain from an additional input stage to a variable resistor on the emitter follower and it sounds pretty great.

[antonis]

I changed the gain from an additional input stage to a variable resistor on the emitter follower

Well done but I don't see any Emitter follower with variable gain..