Hi,

I have designed and built an acoustic guitar preamp. It has a bunch of EQ and gain sections, focused on acoustic guitar.
(Previous thread of a previous version).

The first stage has a JFET and 2 BJT's (EasyEDA schematic below), that honestly I don't really understand. I stole this design from a Fishman Mini amp. This has worked in previous prototypes, but my most recent prototype, I've tracked it down to Q1 is not passing signal. Using a signal generator and scope, I get signal at pin 3 of Q1, but not at pin 2. So that makes me think the switch is not switching. Disclaimer, I'm a mechanical engineer and I don't really understand transistors that well. In the original Fishman circuit they used a 2N4392, but this is very hard to source, so the PN4392 is a good alternative that has worked previously. Also D1 and D2 were not in the original circuit, but someone suggested I add them.

I'd like to better understand what these 3 transistors do, and also how to troubleshoot why signal is not passing through Q1.
VCC and VEE are +/- 15VDC using a Traco 2W TEC2-0923 from 9v.



2nd picture is the original Fishman circuit as reference.

 

Have you tried changing Q1?

 

D2 is back to front. Just remove it and see what happens.

 

Have you tried changing Q1?

yes. And I think jfets are more sensitive to electrostatic discharge so I was careful. No change.

 

D2 is back to front. Just remove it and see what happens.

First thing I tried was clipping both D1 and D2, essentially removing them since they weren’t in the original circuit. No change.

but I wonder if I have one if them backwards, is that what your saying?

 

Exactly!
D2 is incorrectly polarized

EDIT: the original is in a TO18 package, the substitute in a TO92. Are you sure you didn’t swap terminals?

 

Exactly!
D2 is incorrectly polarized

Ok good to know. I think I understand why. However, that’s not the problem.

 

Although since D2 was backwards… could it have fried other parts, effectively shorting Vee to the signal line?

 

With a DMM measure the DC voltages with respect to ground for all the nodes around the transistors

 

EDIT: the original is in a TO18 package, the substitute in a TO92. Are you sure you didn’t swap terminals?

well no. Transistor pinouts are very confusing to me because they aren’t consistent. You’re referring to the PN4392? Or the BJTs too?

I’ll get a PCB photo later…

 

 

Q1:
1 Drain 14.38
2 Source 2.9
3 Gate -0.03

T1
1 Emitter -14.4
2 Base -13.76
3 Collector 2.9

T2
1 Emitter -15
2 Base -14.4
3 Collector -13.76

 

Q1:
1 Drain 14.38
2 Source 2.9
3 Gate -0.03

T1
1 Emitter -14.4
2 Base -13.76
3 Collector 2.9

T2
All ground. That's probably the problem? Because I hooked up D2 backwards?

Q1 and T1 have the voltages I would expect. But how can T2 be all ground, when its collector is connected to T1 base and its base to T1 emitter?

 

Q1 and T1 have the voltages I would expect. But how can T2 be all ground, when its collector is connected to T1 base and its base to T1 emitter?

Yes that makes no sense. I'll re-check tonight.

 

Don’t forget that it’s connected as a source follower, so there will be no voltage gain.

 

I'm also hoping to learn what these 3 transistors do. JFETs (Q1) are voltage controlled, so apply a voltage and it should turn on and apply gain, right? But the BJT's are current controlled, and where they are placed, I just don't understand what T1 and T2 are there for.

 

The JFET is used in linear mode, so there isn’t a question of its being either ON or OFF. As a source follower, the source follows the gate, with an offset of Vgs. Vgs is negative for a JFET so the source will be at a higher voltage than the gate. 2.9V in your case.
The two transistors form a constant current sink, which sink a current of Vbe/R4. As Vgs varies with source current, having a constant source current keeps Vgs constant.

 

The constant-current from T1 and T2 keeps the bias current of Q1 basically independent of the wide variation in the Q1's threshold Vgs between units.

 

I fixed voltages in post 12, they seem to make sense

 

I fixed voltages in post 12, they seem to make sense

Seems like it is working.