Orange Squeezer Distortion

[saxtim]

Hi,

Built and orange squeezer last week and it generally sounds really good, however I have one problem.

If i hit a string or a chord too hard it distorts.  I'm using a fender strat w/single coils.  I set the trimpot to the point where basically there is maximum compression - if I back off this point the distortion became more noticeable and made itself present at softer volumes.

I don't really like this distortion, I don't think it sounds that great (its got that sorta sound when you play a CD through a stereo system too loud and all the high notes distort, know what I mean?)

Did the originals do this? All my voltages look good and I'm sure the circuit is working correctly, so I don't think it's a debugging thing.  I'm using 2n5457s for the FETs.  How can I mod this circuit to get rid of the distortion?  

thanks

tim

[RDV]

Did you use the external Volume, or internal? Also, which layout did you use? Mine is the GGG external Volume version, + I splurged for the RTS PCB and I couldn't be happier with mine.

Regards

RDV

[saxtim]

I used an external volume control - however it doesn't matter where I set the volume control, it always distorts when I hit the strings hard.

I used my own layout.  

thanks

tim

[RDV]

Is the other half of the op-amp tied to ground? Different op-amp than JRC4558? Which schematic did you use? Ge diode? I can't make mine break up even with EMG PUs. There's also supposed to be quite a bit of variance between J-Fets.

Regards

RDV

[bwanasonic]

I have seen this issue mentioned before with the OS. I am a bit suprised you are using single-coils as I assumed it was a hot pick-up issue. I have not been able to recreate this distortion with either single-coils or humbuckers. I used the same JFETs. Maybe the Opamp? Is it possible you are overdriving the next stage in your setup with the OS?

My Os build: http://home.earthlink.net/~bwanacentral/html/o_squeezer.html

Kerry M

[saxtim]

Is the other half of the op-amp tied to ground?

This is probably the problem - I've got pins 5,6 and 7 connected to each other, but these aren't then tied to ground - is that what I need to do?  Connect these 3 pins to ground?

thanks
tim

[RDV]

Is the other half of the op-amp tied to ground?

This is probably the problem - I've got pins 5,6 and 7 connected to each other, but these aren't then tied to ground - is that what I need to do?  Connect these 3 pins to ground?

thanks
tim

Yes. Check out the wiring layout at GGG, it shows which pins are connected to ground.

Regards

RDV
http://www.generalguitargadgets.com/v2/diagrams/orangesqueezer_lo2.gif

[Mark Hammer]

There are several sources of what sounds like distortion on a circuit like that.

One, of course, is the dynamic limitations of FETs.  You may want to consider reducing the gain of the op-amp a bit since that might be pushing the FETs in unpleasant ways.

Another source of apparent distortion, which isn't really distortion comes from envelope ripple.  The envelope follower in any compressor attempts to balance out responsiveness, in terms of tracking instantaneous changes in level, with smoothness.  Maximizing one will invariably impede the other so they aim for a compromise.

When the tracking of the envelope follower is not smooth enough, the little variations introduced as the note trails off result in small fluctuations in sweep of the control element (in this case a FET) that sound like a burr added to the note and very much like distortion.  This is more true of designs that use a half-wave rectifier circuit (which both the OS and Dr Q do) than those using a full-wave rectifier, and also more true of designs that use a speedy FET or transistor control element (like the DQ and OS) than those using a more sluggish optical control element.

Just for the heck of it, increase the value of the 4.7uf capacitor to ground just after the diode.  You can either tack on another cap (e.g., 4.7 or even 10uf) or replace the one there.  If the "distortion" appears to be reduced with only that change, then that is your problem.  You can also consider increasing the value of the 100k resistor to 330k or even 470k to extend decay time a bit more.  A slower decay will generally increase the apparent smoothness of a simple half-wave rectifier circuit like that in the OS.

[saxtim]

I've now fixed the problem somewhat - I've tied the other half of the opamp to ground.  This greatly reduced the distortion - however it's still present sometimes when I push it too hard.  

One, of course, is the dynamic limitations of FETs. You may want to consider reducing the gain of the op-amp a bit since that might be pushing the FETs in unpleasant ways.

How do reduce the gain of the op-amp?

Just for the heck of it, increase the value of the 4.7uf capacitor to ground just after the diode. You can either tack on another cap (e.g., 4.7 or even 10uf) or replace the one there. If the "distortion" appears to be reduced with only that change, then that is your problem. You can also consider increasing the value of the 100k resistor to 330k or even 470k to extend decay time a bit more. A slower decay will generally increase the apparent smoothness of a simple half-wave rectifier circuit like that in the OS.

I will give this a and get back.

thanks
tim

[saxtim]

Just for the heck of it, increase the value of the 4.7uf capacitor to ground just after the diode. You can either tack on another cap (e.g., 4.7 or even 10uf) or replace the one there. If the "distortion" appears to be reduced with only that change, then that is your problem. You can also consider increasing the value of the 100k resistor to 330k or even 470k to extend decay time a bit more. A slower decay will generally increase the apparent smoothness of a simple half-wave rectifier circuit like that in the OS.

Ok, I put a 10uf cap in parallel w/the 4.7 after the diode - this didn't improve the distortion.  I'd like to experiment with reducing the gain of the op-amp, though I don't know how.  Advice would be appreciated

thanks!
tim[/quote]

[smoguzbenjamin]

Increase or reduce R9 to change opamp gain... I'm not sure if its a bigger or smaller resistor you need, I always tend to use the inverted input

[saxtim]

Increase or reduce R9 to change opamp gain... I'm not sure if its a bigger or smaller resistor you need

Thanks, I'll have a fiddle with this.  Anyone got a definate answer as to whether or should increase or reduce this resistor to reduce the gain? (resistor connects from pin 2 to pin 1 of the opamp)

thanks

tim

[smoguzbenjamin]

Just give it a go mate, change the value by 50 - 100k and see what the difference is. just turn your amp DOWN and the increase the volume, just in case you create a huge output that can damage your ears (or your neighbour's temper )

[saxtim]

Did some experimenting - reducing the resistor reduced the gain.  I ended up going with the 220k anyway though, as I realised I hadn't rebaised the trimpot since I tied the other half of the opamp to ground. I've now got it working distortion free (after a bit of fiddling with the trimpot) at a compression level that I like.  

tim

[RDV]

Did some experimenting - reducing the resistor reduced the gain.  I ended up going with the 220k anyway though, as I realised I hadn't rebaised the trimpot since I tied the other half of the opamp to ground. I've now got it working distortion free (after a bit of fiddling with the trimpot) at a compression level that I like.  

tim

Cool!

Regards

RDV

[smoguzbenjamin]

Congrats tim I'll be making a squeezer in the future aswell. Doh! I also wanna build a dr quacky and a woody. Damn :mrgreen:

[Mark Hammer]

The capacitor suggestion I offered was intended to address something that some folks *describe* as "distortion".  Clearly that wasn't the problem in your case.

The modded version of the OS that I made for myself is drawn and posted at generalguitargadgets.  One of the things that is awkward about the stock OS is that compression amount and overall level are not separable.  In the modded version I made, the gain of the first op-amp (hence compression amount) is variable, and a second op-amp compensates for the altered gain in order to keep volume level up.  You may want to look at that for some ideas.  Sounds like you are on the right path, though.  Congrats.

[caspercody]

Mark

Anyway to get a copy of your mods?

[GibsonGM]

I like the OS too, but have found that lack of separation to be limiting.  +1 for seeing how you did that, if you can find it, Mark!     

[bluebunny]

I think this may be the version at GGG that Mark mentioned:



Here's another one of the same vintage, drawn by Charlie, that also mentions Mark's mods: