Bearhug vero - no compression!

[mysticaxe]

I've built a Bearhug vero based on the Tagboard layout (http://tagboardeffects.blogspot.com/2013/01/bearhug-fet-compressor.html) and I don't seem to be getting any difference in effect regardless of knob setting.  Actually, it looks like one of the 10K resistors in parallel to the high/lo switch is missing...  The hi/lo switch position doesn't change the sound either, so while that may be one problem, it isn't the main one.  My only substitutions were using 1n34A's for D3/D4 and tantalum caps for the 1uF's.  The one on the input has the - side towards the input and the + towards the BS170 (switched based on midwayfair).

Based on a couple ideas from midwayfair, here is what I have (all DC):

Q1 - D - 5.8V
       G - 2.1V
       S - 3.3V
Q2 - C - 7.4V
        B - 1.2V (drops to ~0.8V when playing)
        E - 0.72V
Q3 - D - 90mV (jumps to 180 mV when playing)
       S - 0V
       G - 55mV (drops to -0.6V when playing)
D3 anode - 9mV (goes to 2-3V when playing)
     cathode 45mV (goes to -1.5V when playing)


2N5457 drain to ground resistance is 370 ohm (goes up to 700 ohm when playing).  Gate to ground is 560k (drops to 70k when playing)

The compression knob is 1kB - when it's all the way one way, the 1-3 resistance is 0, the other way it's 0.97k.  Turning the knob doesn't seem to change the current on the Q2 emitter.

I've verified all the components (twice) and reflowed a couple joints (which is why the voltages are a little different than what I told midwayfair).  I also accidentally reported the voltage on the wrong side of D3...  Sorry!

Pictures to follow very shortly.  Thanks in advance for any help!

[mysticaxe]

Pictures.  Volume knob works fine.

[mysticaxe]

Picture of the solder side.

[midwayfair]

The 47uF cap on Comp lugs 1&2 is backwards in the layout.

Deciphering vero is a pain. I'll let you know if I see anything else.

[mysticaxe]

Thanks.  Flipped the 47uf around.  Didn't solve it.  Do the voltages looks reasonable?

[midwayfair]

Do the voltages looks reasonable?

Actually, Q1 looks wonky ...

My Q1 is D: 7.8v, G: 2.8v, S: 2v
My Q2 is C: 7.8v, B: 1.04v, E: .5v, so yours is close enough.

The fact that your collector and drain are different voltages definitely means something weird has happened. They are both connected to the 9v via a 2.2K resistor -- they should have the same voltage.

You should be getting something closer to -4v on the cathode of D3 even with 1N34A. Heck, it's 3.5v in mine with LEDs
instead of the schottky diodes.

What's the audio path doing? Is it still distorted the whole time? You might want to replace Q1 with a socket and see if it's just a bad MOSFET. Maybe do the same with Q2 since it doesn't seem to be amplifying ... I've fried chips and transistors soldering them before, so this wouldn't be the first time.

[addshoe]

The layout has been updated for the flipped cap...http://1.bp.blogspot.com/-5a-ds9F4WR4/USFXXCQweLI/AAAAAAAAB08/A0ZHTAQeABk/s1600/Bearhug-FET-Compressor.png

Are you missing the 4.7cap at the far right? Should be two there, on at the comp 1&2 line and a second at the sw2 line.

I am building this project too right now, going to wire and test tonight so I will post back with my results.

[midwayfair]

Are you missing the 4.7cap at the far right?

It's a tantalum cap in his build.

[addshoe]

Are you missing the 4.7cap at the far right?

It's a tantalum cap in his build.

DOH! How did I miss that! See it now on second glance.

[mysticaxe]

OK.  It's taken me a while (I got sidetracked on finishing an Echo Base, plus had work travel).

I socketed Q1 and Q2 and reflowed the solder joints around the compression pot (both on the board and pot, they looked OK, but I thought what the hell!).  It looks like my Q1 was the culprit of the low drain voltage.  Now here are the readings:

Q1 D - 7.2V, G - 3-2.5 V - pretty much always falling when probed, S - 2.3
Q2 (tried 2, both were about the same) - C - 7.3V, B -  .9-1.2 V (drops to about 0.8V when playing), E - 0.4V
Q3 - D, S - 0, G - 6.7 mV, drops to -2V when playing
D3 - Around -3.5 V after chord

[midwayfair]

OK.  It's taken me a while (I got sidetracked on finishing an Echo Base, plus had work travel).

I socketed Q1 and Q2 and reflowed the solder joints around the compression pot (both on the board and pot, they looked OK, but I thought what the hell!).  It looks like my Q1 was the culprit of the low drain voltage.  Now here are the readings:

Q1 D - 7.2V, G - 3-2.5 V - pretty much always falling when probed, S - 2.3
Q2 (tried 2, both were about the same) - C - 7.3V, B -  .9-1.2 V (drops to about 0.8V when playing), E - 0.4V
Q3 - D, S - 0, G - 6.7 mV, drops to -2V when playing
D3 - Around -3.5 V after chord

It's definitely working if D3 is dropping that low. Is it still distorting?

[mysticaxe]

No distortion that I could hear!  Through my little home practice amp, I couldn't hear any real difference between the hi/lo levels or anywhere on the compression knob.  It's almost like its at a fixed ratio...

[midwayfair]

No distortion that I could hear!  Through my little home practice amp, I couldn't hear any real difference between the hi/lo levels or anywhere on the compression knob.  It's almost like its at a fixed ratio...

It is a fixed ratio. The comp is a threshold control.

The compressor limits peaks. It happens very quickly -- between half and one and a half cycles of any given wave or "faster than your ear can hear the transients," and the decay is short enough that you'd only notice it carrying over between notes in fast runs or when strumming chords.

If you measure the resistance between the gate and source of Q3, you should see it go from about 300Ohms to ~1-2K. This is actually a pretty substantial reduction in gain.

[mysticaxe]

Ahh.  I guess that explains it!  I don't think my 2n5457 got that high on strumming (maxed out at about 800 ohm).

[midwayfair]

Ahh.  I guess that explains it!  I don't think my 2n5457 got that high on strumming (maxed out at about 800 ohm).

Hrm, that is a bit low. Mind you, I don't think anything is wrong with your build. I've had a couple bad batches of 2N5457s recently not working properly in other designs. But you could try two things: Replace the 1n34A with the Schottkys (the lower Fv really will help), and/or try a different FET (e.g. 2N5952) if you want to see if it might give you a little more compression. Mind the pinouts. The diodes are the main thing for maximizing the voltage drop.