I've been working on a guitar preamp that starts with the circuit below from a Fishman Loudbox amplifier. I've replicated it but it's not working. I substituted TO-92 versions for the transistors and I'm running +/-15v on the rails. Using transistors at the input like this is unusual. I'm trying to understand what the transistors are for. Q1 (n-channel JFET) looks like a switch, but I'm not sure what Q2 and Q3 (both BJT's) do. I have considered trying to bypass the transistors by removing Q1 and or all 3 of them and jumpering pins 2 and 3 on Q1.

Looking for advice.



Below is an example of the input of another guitar preamp that uses +/- 15v on the rails, but without the transistors.

 

Q1 is a source-follower to provide a low-impedance drive the the op amp active-filter circuit (that filter needs a low-impedance source for proper operation due to the negative feedback to the input).
The op amp filter below that doesn't, since the signal goes into the (+) input with no feedback to that pin..

Q2 and Q3 form a constant-current, high impedance bias load for Q1's source (current is approximately 0.7V / R5).

Have you double-checked all the wiring and pinouts for the transistors and op amp?

What is the DC voltage at Q1's source (pin 3)?

 

A very simple simulation in Eagle. The source (pin 3) looks like it is sitting at -14.6v. So there's about 29.4v across Q1, it's rated for 30v across both Drain-Gate and Gate-Source. 30v acros rails should be fine right? The original Fishman circuit only had 28v, I'm using a DC/DC converter (Traco TEC 3-0923) that outputs +/-15v.

 

I used a TL071 for my guitar amp. I will show a circuit diagram later.

 

Below is my LTspice simulation:

Your sim seems to show an enhancement-mode MOSFET rather than a depletion-mode JFET, which could account for you simulation results.

 

Spice tells me the Source voltage for the 4392 is 1.8v

 

Spice tells me the Source voltage for the 4392 is 1.8v

I get the same value.
It means the Vgs for the 2N4392 is -1.8V for operation in the linear mode at the 6.6mA drain current.

Have you simulated it with the AC signal to see if it works?

 

Yes, I've done extensive AC simulations for all the frequency adjustments, it does what I want in that regard. Not as much in DC simulations. Starting from that Fishman circuit above, I added a state-variable notch/HPF as well as modifying the mid control to have a gain, frequency and Q knob. I also added a Volume and Boost knob at the end.

I designed it up in Eagle and printed prototype PCB's. Larger board is the signal board, smaller is power supply and footswitches. First batch had 5 minor mistakes that I have modified in hopes that I can still get it to work. At this point, I know I have +/-15v at the rails. I have bypassed the footswitches (taking IN and OUT directly to the signal board), only using the smaller board for power. But I can't get any sound to come out. I have swapped op amps, those aren't the problem.

So I'm in full on troubleshooting mode, now having eliminated the obvious stuff. I suppose I need to probe around and figure out where there is and isn't signal.

 

I suppose I need to probe around and figure out where there is and isn't signal.

Yes, just follow the signal through the circuit until you find where it stops.
Do you have an oscilloscope to do this?

 

The 5 mistakes I made on the PCB's:
1. signal board didn't have a ground between the rails. I jumpered it.
2. toggle switches had too small a footprint between pins. I had to bend the pins a bit.
3. Lower row of pots hit the dual toggle. I enlarged the Boost and NotchQ holes and bent the pots out a bit.
4. 3PDT's were rotated 90. I really think this is an Eagle model error. Desoldered and soldered in correctly.
5. There was a trace that crossed a ground on power board. I cut the trace and jumpered.

 

Yes, just follow the signal through the circuit until you find where it stops.
Do you have an oscilloscope to do this?

I do have an oscilloscope, 2 actually. But I don't really know how to use them, I need to learn.

 

I do have an oscilloscope, 2 actually. But I don't really know how to use them, I need to learn.

Any questions about that, just ask here.

 

Hello,

The attached PDF might give you some insight on the use of oscilloscopes.

Bertus

 

It would be very difficult to do any electronics design, test, and repair without an oscilloscope.
The oscilloscope is your number #1 goto test tool. Learning to how to use it should be top priority.

 

Back to original post. Why use transistors instead of op amps there? That is very unusual in a guitar pedal.

 

Back to original post. Why use transistors instead of op amps there? That is very unusual in a guitar pedal.

I don't exactly know what the designer was thinking, but there are several possibilities.
1) In this case, it looks like the circuit would need three op-amps, with the first being no more than a buffer to provide a low-impedance drive to the filter.
A JFET source follower can be a buffer.
But by the time you have added a dual transistor constant current source as the source-follower load, a single op-amp would be fewer components.

2) A JFET can be lower noise than a op-amp buffer.

3) Some people say that they prefer the sound of transistors over op-amps. Maybe they can hear the difference

4) Marketing. You can make all sorts of spurious claims!

 

3) Some people say that they prefer the sound of transistors over op-amps.

Based on my experience with various strata of people and their guitar pedals, amps, etc., that is the #1 reason.

ak

 

Many years ago the electric geertar preamp used vacuum toobes. Then transistors were invented so they used germanium transistors that produced a certain distortion (fuzz and overdrive) that some people liked. Then silicon transistors were developed that sounded different to old germanium transistors. Then opamps were developed that produced extremely low distortion that some people did not like so they messed up the circuit to add old-style distortion.

 

So the top circuit in post 1, would it work if I removed Q1, Q2, Q3 and R6? The signal would be at 0V I think, midway between rails.
And jumper R2-C5

 

So the top circuit in post 1, would it work if I removed Q1, Q2, Q3 and R6?

Depends on what you mean by "work". It would be a drastic change to the input of the filter stage. Audio will come out of U1A, but it will be significantly different from that of the original circuit.

ak