Bearhug Compressor - Boost but no squeeze

[alltherightpills]

Hi again y'all,

I've been trying & testing all night to try to get this little comp circuit to squeeze! No luck yet, plenty of boost on tap though.

Everybody probably knows it, but the schematic is http://1776effects.com/wp-content/uploads/2015/03/Bear-Hug-V2-1.pdf
Built on perf w/ this layout: http://effectslayouts.blogspot.com/search?q=bearhug

All parts as spec'ed except:
Q1 is 2N7000 (flipped for pinout)
Q2 is 2N5089
R5 & R7 are 3k (all I had on hand, so a 4k7 for R7 wouldn't set Q1's drain voltage high enough)

Things I've tried:
-different 2n5457 for Q3. Mine measure idle resistance of ~430ohm
-MPSA18 for Q2
-different 2N7000 for Q1
-checked & rechecked components for accuracy
-metered & scraped looking for solder bridges
-reheated everything

Voltages:
Q1 2N7000
D 4.98
S 2.68
G 2.06

Q2 2N5089
E 0
B 0
C 8.27

Q3 2N5457
D 0
S 0
G 0

And just in case its helpful, here are pics of my boost / compressor-to-be. Thanks in advance for any guidance!

[duck_arse]

your Q2 base voltage is bad. bottom right of your board is a resistor should be 10M - yours has a yellow band, looks like 2M7? and the other 10M far right is not visible. are those values correct? I can't read perf solder-sides.

[alltherightpills]

You're right, the visible resistor in the lower right of the board (R1 on the schematic) is 2M7. I forgot I swapped that at some point. Guess I assumed that wouldn't have a huge affect on the circuit.

The other 2 10M resistors are indeed 10M (R2 & R11).

I'll zero in on components around the Base of Q2. See if I can make a breakthrough...

[alltherightpills]

Hit the nail on the head, Stephen. Turns out my Q2's socket had a short somewhere I couldn't see. Replaced the socket and I've got all the squeeze & squish I could ask for.

Side note - This is my 2nd build recently where a transistor socket was problematic. Admittedly, I am just cutting sections off of a large SIP socket from Tayda. I've been doing this for years with no issue, but this has me curious if there's a better way.

Anyone have a different socket solution that they love?

[bluebunny]

Anyone have a different socket solution that they love?

Yeah, "solder".      Seriously, audition transistors (if that's what you need to do) on the breadboard.  Then commit to PCB/perf/vero with good ol' solid solder.

[duck_arse]

haven't you been listening to pinkjimi? he'll tell you about transistors in sockets. geeze, if all fixes were as easy.

if R1 is a pulldown instead of a gate-bias, it probably won't make a diff.

[edit :] if'n youse gots to have a socket, use dual-wipe contact types instead of the machined pin type.