Orange Squeezer Compression problem

[fab672000]

Hi, I'm trying to make the feedback compression working on my Orange Squeezer (tonepad version http://www.tonepad.com/project.asp?id=5)
Here's the schematics:

and here's my board (perfboard):

I use 2N5457's, one 1N34A germanium diode, all components standards, no mods except one 500k pot replacing the 100k fixed resistor, I have set up the bias trim to approx 1.7V, I hear that it has an effect (decrease) on the gain of the op-amp as RG explained in his full OS explanation,
The problem is that I don't hear any significant compression when testing. Even more, I decided to replace the 100k resistor with a 500k pot to try to increase the decay/release and see if I hear something -> absolutely no hearable difference.
I spent many hours trying to figure out what's wrong in this feedback part of the circuit, the problem is that I don't know what tension I should expect on this path...
I'm a noob, so I don't really understand how to debug this, though I already tested all connections, all resistances values (also checked them with my multimeter) and capacities values-> all correct.
I check all pinouts on the data sheets of 2N5457's, I check the RC4558 opamp (it works), the diode orientation, the polarized caps orientation, I also checked all paths, to see if one connection is absent, and did some basical short-circuits testing, nothing wrong apparently but something must be wrong :-(
It works as a preamplifier for now, not like a comp.


Anyone having an idea of what's happening ?
EDIT:I measured all the dc voltages I can measure on the transitors the Ic's as aron suggested :
DC Voltages measured in Volt from ground to point, with jacks connected, but not playing any sound:
at +9VDC = 8.92 V
at  bias trim 10k = 1.72V
Q1 (2N5457)
G=1.72V
S=2.73V
D=8.88V

Q2 (2N5457)
G=0V
S=1.72V
D=1.72V

IC1 (RC4558P)
P1=8.31V
P2=8.31V
P3=0V
P4=0V
P5=4.79V
P6=4.89V
P7=4.85V
P8=8.89V

D1 (1N34A)
A (anode, the non-band end) = 0V (in VAC measurement, with my thumb tapping on the tip, I get 20mVAC approx)
K (cathode, the banded end) = 0V (in VAC measurement, with my thumb tapping on the tip, I get 1mVAC approx)
Finally I tested the Voltage of polarized capacitors I get a positive value at the + side and 0V at the grounded side, seems ok ?

[The Iron Chef]

nothing wrong apparently but something must be wrong :-(
It works as a preamplifier for now, not like a comp.

I have the same issues. I built 2 of these and was very pleased with the build of both boards. However they both perform in the same way. Just as described in FAB's post above.  I am a mechanical guy not an electrical guy so my diagnostic skills ah, non existant suck. Sorry, no extra credit here for pin voltages or pics .........

[fab672000]

Good to see I'm not alone Iron Chef 
BTW, I have made a mistake naming Q1 and Q2 ; in fact Q2 on the schematics above is Q1 in my measurements (drain connected to +9V power supply), I mixed myself with what i saw on the guitar gadgets osq board, sorry.

Also, I have setup my trim to 2.4V to compare with the suggested OSQ values (as it is seems quite similar except the 1M5 resistor pull down resistor at the input)
And I get values very similar to recommended values at Q1(Q2 on my board) and Q2(Q1 on my board)
the difference is that the recommended source tension for the 'variable resistor' FET is  S=2.6V (Q1 on the tonepad schematics above)  but mine is equal to the trim tension (which is 2.48V in my second test).
you can find the recommend values here :
http://www.diystompboxes.com/smfforum/index.php?topic=63789.0

I would add that my conditions of testing when playing were :
direct plug to my motu ultralite digital audio interface, 1st with the effect bypassed, 2nd with the effect on

Trying to play with abusive dynamic changes (right hand soft and strong attack, legatos, tapping in clean direct sound and so on).

I also verified that I hear clear differences with different  compressors in my pod XT pro gear.

To sum it my DIY OS currently has a cool preamp sound but no compression effect, it is as if the feedback part would not do its job.
I read the debugging faq and tested the polarized caps with the suggested tips, seems ok.

I hope I did things right when I furthered as many details as possible as aron suggested on his recent post about debugging,
if for any reason though, you think I shouldn't put particular details, please feel free to remove them, and I apologize in advance for any problem related to it.

[George Giblet]

Firstly there is no "correct" value for the Q1 source voltage.  The reason a pot is used is to *change* that voltage to suit the *particular* JFETs you have in your hands.  Voltages in the 1 to 2V range seem about right for the JFET part you have chosen, but could be outside this range.  The bottom line is don't set this voltage to something that someone else hase used - their JFETS aren't yours!

The voltages you measured look pretty good - so I'm having doubts something is wrong.

The best way to test the compression is working is to play through the unit then adjust the 10k pot.  You should find a region where the volume is loud and the trimpot has no effect, then you will find a point where the pot has an effect, the beyond that point the volume gets lower and lower.  If you can get this behaviour it's highly likely the unit is working.  The "correct" point is around where the onset of volume reduction occurs.

The os is a quite a subtle compressor.

[fab672000]

Thanks for your reply George!
it indeed seems to work, and I get a change of perceived volume when setting up the 10k trimpot, thanks also for the hint on determining the correct point, I use all standard components with no substitution,  and for me the volume starts to lower around 1.75V, still when there and testing different configurations for the trim, either I am at the limit of transistor response, hearing some cuts and plops, or, when raising the trim voltage and so the perceived volume, I don't hear any compression.
Maybe it is because it is very very subtle indeed

[selectortone]

I built one of these using a GGG board and schematic.

It's quite a subtle effect compared to say a Dynacomp and I've found that the amount of compression depends a lot on the type of pickups in the guitar.

With P90s or humbuckers it's more obvious than with a Strat.

Try boosting your signal with a pedal in front to see if the pedal is actually compressing.

Here's mine (on the right):

[fab672000]

Thanks, I'll try this too.

BTW, your pedals are AWESOME !!!
Do I need special material to achieve as beautiful results as yours, just curious ?

[George Giblet]

One thing which could make the unit more subtle than it should be is if the 4u7 cap on the source of Q1 is faulty, has a broken leg, or not soldered correctly.

[selectortone]

Thanks, I'll try this too.

BTW, your pedals are AWESOME !!!
Do I need special material to achieve as beautiful results as yours, just curious ?

Thanks!

I print my designs onto photopaper, stick them down with double-sided tape and then use automotive clearcoat (three coats).

[fab672000]

I tested this for the 4.7uF cap:
2.47VDC on the + pole
when rapidly disconnecting the power, i get 1.4V rapidly then a slow decrease from .2 to 0V, it seems ok.
I also tested with a bridge humbucker and it seems to compress, but isn't it due to the higain humbucker nature more than the compressor's compression ?
I also thought that maybe the diode voltage drop could be a reason why it does not work that good ?

Thanks.

[fab672000]

I made new tests, and indeed it works much better with humbuckers, legatos parts are more hearable, and when i switch it off and on rapidly i clearly see a difference now !!
Thank you very much folks !
ps: I think I'm going to build an MXR dynacomp now, will be more obvious to hear something 

[oldrocker]

IMHO the subtle effect compression on the OS is the beauty of it.  I have built other comps but I always like my OS the best.  In fact I now use it on my bass when I play.  I won't play at a gig without it.

[fab672000]

I'm no expert but I can easily believe this  because after what I heard from it today, it's magic, it compresses legatos beautifully but not enough to change radically the color of what you're playing, I' ll keep that one too 

[fab672000]

Okay my  baby is in the box !
some pictures as promised:



Let's Rock Baby!

[oldrocker]

Nice job man.  I wish mine looked that good.

[The Iron Chef]

thanks all for your advice,
Nice job Fab.
This is my effort. I tried to copy an EH enclosure
&nbsp

[fab672000]

Nice job man.  I wish mine looked that good.

Thanks !
I'd like to thanks selectone for his hints about glossy photo paper and varnish coats, I just applied that ;-)
Nice Job  Iron Chef, I love  shiny metal tops too !

[selectortone]

I'd like to thanks selectone for his hints about glossy photo paper and varnish coats, I just applied that ;-)

Hey, you're welcome - just passing on one of the many tips I've learned while lurking here 

Nice job on your pedal!