Marvel Drive Debugging

[jb83]

I just finished this one, and it sounds stellar with the exception of the normal volume.  With the master at about 2, I can crank the treble volume and get nice volume and saturation.  However, when I turn the treble volume to zero and crank up the normal volume, it feels like I'm getting about 75% of the volume and gain that I am getting on the treble side.  There is also a general "flabbiness" to the distortion.  Here are the voltages:
Q1: 4.1
Q2: 5.2
Q5: 5.34

I know J201's are fickle as far as gain, so do you all think I should just switch out transistors?  If so, do I just switch out the Q1, or are Q3 and/or Q4 part of the equation as well?

Here is a link to the schematic: http://cl.ly/image/1Z1a2h1p1g3T

I've tweaked the trip pot on Q1 (the one associated with the normal volume) every which way to get the most volume and distortion, but, at best, I'm getting what I described above.  Thanks in advance, and I'm glad to be part of the community!

[jatalahd]

Hi jb83 and welcome to the forum. I am not an expert with JFETs, but I will try to answer anyway.

Assuming your voltage readings refer to drain voltages, then those look ok, but don't give that much information, since the voltage from gate to source V_GS also matters quite a lot, since it relates to the biasing point of the JFET.

If there is a difference between treble and normal volume, then in my opinion only Q1 and Q2 are concerned. JFETs Q3 and Q4 form a high-gain cascode ( see page 5 of the application note: http://www.ti.com/lit/an/snoa620/snoa620.pdf ) and Q5 seems just to be a buffer with minimal gain (since the source resistor is not bypassed with a capacitor). The treble and normal signals from Q1 and Q2 are mixed together in the junction of R5 and R7 and from there the mixed signal is amplified further.

To me it looks that Q1 and Q2 use the basic JFET self-bias, where the bias-point is determined mostly by the source resistance. It seems that Q1 and Q2 have been deliberately biased into different quiescent points, which will cause a gain difference. BUT, If Q1 and Q2 have similar characteristics (same pinch-off voltage V_GS(off) and same Idss) this difference in gain should not be so disturbing as you mention when using J201 (because it typically has a relatively low pinch-off voltage).

It is true that each JFET is an individual and the pinch-off voltage can vary a lot. Therefore, if you want to get clarity about the characteristics of your JFETs, a simple multimeter measurement will reveal the differences. To measure the pinch-off voltage, connect the drain directly into the 9V battery, gate to ground and source to ground through a 1Meg resistor. Measure the voltage over the 1Meg resistor and since it is a N-type JFET, put a minus sign in front of the measured voltage. This measured voltage (with the minus sign included) is a good approximation of the pinch-off voltage of the given JFET.

So if you want to be "analytic", select a pair with similar pinch-off voltages as Q1 and Q2 and tune the trimmers at the drain so that the drain voltage in both JFETS is around 5.5 - 6.0 volts. If this does not correct the imbalance between the treble and the normal, then I would assume the gain difference is intentionally designed into the circuit.

Maybe someone else here has some more info to share about this circuit or the J201 in general  ... ... ...