Tone Sucking Ross Comp Advice Needed

[KazooMan]

I built a Ross Comp following the info from RG's GEOfex website a while back.  I was never very impressed with the compression I obtained.  I  also built the DOD 280 "optical" comp from the Tonepad site and had a bit better results there.  I recently resurrected these pedals from the pile to use the DOD 280 to have a go at the plexiglass surround lighting effect (finally got that to work great). 

So I returned to the Ross to see what it could do.  After a lot of tweaking and various swapping of parts I had screwed up the board so bad that I etched a new one.  I moved most of the parts over from the original board, but used a lot of new parts including all new transistors.  The IC is socketed and I have tried several.  I have the true bypass with the Millennium board for the indicator LED. 

The compression works "okay", but this pedal is an absolute tone sucker.  The tone is fine on bypass (thanks to the Millennium) but when the effect is engaged it sounds like playing through a bottle of marshmallow fluff.

Any hints on where to start looking to solve the problem?  I have assembled many pedals and lots of tube stuff and after going over the board many times I am confident that there are no (obvious) problems with the build.  I need some expertise with the basics of the circuitry and where the tone sucking may arise.  BTW, I am a synthetic organic chemist by training, not an electrical engineer!

[jkokura]

Have you played with the trimpot to get a better sound much? I don't seem to have the problem with my Ross - in fact I find it more transparent than my Orange Squeezer which is pretty dang clear also. I would guess that there is a problem somewhere, cause that's not supposed to be what it sounds like, and that it would be worth it for you to try and fix, cause there are so many Ross Clones out there that people love, so it's not likely that the Ross Compressor design is what you don't like, it's more likely a mistake or error. Perhaps if you went through the debugging process described in the sticky we could help - I've built two of these from Tonepad's site and like them both!

Jacob

[Processaurus]

Mine sounded bad, tone suck, until I noticed with a buffered pedal in front, it sounded much better.  I tacked the GGG discrete JFET buffer onto the front end.  If that doesn't help, you may have a wrong part somewhere (especially check the value of the cap to ground right at the input of the effect).

[KazooMan]

Thanks for the suggestions.  I'll check out the value of the caps.  I have played with the trimmer and all I can do is change the tone sucking setting into one with some distortion.  I should say that I broke a pin off the trimpot I used in the original build when I was desoldering it to move to the new board.  The only trimmer of the proper resistance I could find locally is an 18 turn Cermet.  Overkill for the intended use to be sure!

[trjones1]

I personally just don't think the Ross Comp sounds that good for precisely the reason you said.  It doesn't sound transparent to my ears, just dull.  After a couple years of disappointment with mine I converted it to a Dynacomp (Tonepad's project file shows the differences in the circuit) and I've been happy getting "that Dynacomp sound" ever since.

[Mark Hammer]

The majority of compressors DO have the net effect of seeming to take away some of the presence.  That's not a result of a "bad" design so much as a byproduct of the compression process which is not compensated for.

Ask yourself "Where does the treble live?", and the answer is "At the beginning of the note".  Now ask yourself "Where do compressors exert their greatest impact?", and the answer is "By clamping down on the dynamics at the beginning of the note".

One of the things that some compressors have tried doing, is blending back in some of the unprocessed signal, in order to keep that extra bite that gets all packed down by compression.  Another thing I might note, and certainly not as any sort of scolding, is that the less compression one applies, the more of the treble will seem to be preserved.

Finally, one thing you can easily do with the Dyna/Ross design is stick a capacitor in parallel with the 10k resistor that comes just before the Volume pot.  That will act as a Q&D treble enhancement.  Not a stupendous one, I have to admit, but easy to do and dead simple to tinker with to get it to your taste.  Start with a .022uf cap, and work from there.

[zombiwoof]

Has anyone noticed that BBE has a new comp out to replace the defunct Orange Squash that is a Ross clone?  It's called the Bench Press.  Just thought some might be interested in a new Ross clone on the market.

Al

[KazooMan]

An update.  Thanks for all of the suggestions!

First off, I did a careful check of all of the components and learned a valuable lesson (one that did not solve the problem, unfortunately).  When I made a new board and moved components, I installed new transistors.  I take care to use a heat sink when I solder them in and I was not comfortable with how they would withstand the heat of unsoldering so I grabbed my bag of new ones  Well, there were supposed to be five 2N5088's on the board, and I had pulled out a bag of the same, but one of the transistors was NOT a 2N5088.  It was another NPN transistor, with similar, but different, specs.  My bad for not checking every part before installing.  Someone else's bad for putting the wrong transistor in the bag

I installed another 2N5088 and, sorry to say, had the same tone sucking result.  Good lesson learned, but still not happy.

I tried many alternative caps in parallel with the 10K resistor before the volume pot as Mark suggested, but without any improvement.  I reread RG's description in the build instructions on how the compressor works and noted the comment on the cap/res combo that sets the high rolloff.  I tweaked this a bit also with no better results.  This isn't on a breadboard, so I am slowly but surely making a mess of things.

I am coming to understand Mark's comments on how a compressor functions and I think I am just not comfortable with the impact on the tone.  The DOD 280 is better in this regard, but there is still some compromise. 

[JDoyle]

If you followed the GEOFEX layout exactly, unfortunately there is a typo on it - the capacitor on the input of the circuit is NOT 0.0022uF as stated on the schematic at GEOFEX but should be 220pF

I can almost guarentee that changing that cap to the 220pF value will significantly improve your opinion of the tone.

Regards,

Jay Doyle

[Mark Hammer]

Ah yes, THAT one!

Good call, Jay.

[JDoyle]

Ah yes, THAT one!

Good call, Jay.

I am lucky enough to own an original and also have a gracious member send me a factory schematic - otherwise I wouldn't look so informed. 

In my opinion that cap is the ONLY poor design aspect of the Ross Comp - as it just hangs there without a preceding series resistor, the input impedence of the circuit preceeding determines the cut off frequency for that cap - thus Processaurus' improvement in tone with a buffer in front.

One last thing - for more output available, remove/jumper the 10k preceding the level pot.

Also - the 27k in series with the Iabc port (pin 5) from the Sustain pot protects the CA3080 Iabc port from passing too much current to destroy the chip, more than 2ma into that pin will render the chip a pretty little piece of useless, but 27k might as well be a 'belt and suspenders' approach - at 9V V+, the most it needs to be 10k, with 5k being the minimum; lowering it to either of those values will increase the amount of sustain available.

I have a post in the archives outlining all the improvements one can do to improve the noise level, bias it properly, etc. Unfortunately my alma matter (Wake Forest Univ.) just got their asses handed to them by Georgia Tech tonight so I've had a few too many consolation Manhattans to post those same improvements in any useful way. 

[Processaurus]

In my opinion that cap is the ONLY poor design aspect of the Ross Comp - as it just hangs there without a preceding series resistor, the input impedence of the circuit preceeding determines the cut off frequency for that cap - thus Processaurus' improvement in tone with a buffer in front.

I think I may well have been a victim of the .0022 cap, it was my first pedal so I didn't understand the lay of the land very well, and the buffer sounded good so I never thought about it again.  Probably unnecessary.  Though I'm not about to ever open that ratty box up again.  On that one the circuit board got ran over by a truck the first time (last time I put a circuit on top of the car when loading other crap, it had its revenge though, a CTS pot popped the trucks tire), so all the components are 2nd gen.  Lots of the traces are kind of peeling off the board from the trauma.  But it's worked perfectly for 8 years!

[KazooMan]

Thanks for the tip Jay!  I can't swap the cap right now, but I am certain that you are right.  I just checked and I have the .0022 cap.  What is really interesting is that I pulled out my original printout of the GEOfex article and I see that I wrote down the correct cap value in the margin on the parts list!!!  When I populate a board I like to color in the parts I install with a highlighter on the parts layout as I go.  So...... I printed a new set of instructions and thereby missed the cap value change.   

I'll try some of the other mods you have mentioned.

Thanks again., I'll report back with the results.

[BAARON]

You can omit the 220p cap entirely if you want to.  It's not even there in the Dynacomp.

[Mark Hammer]

Thanks for the tip Jay!  I can't swap the cap right now, but I am certain that you are right.  I just checked and I have the .0022 cap.  What is really interesting is that I pulled out my original printout of the GEOfex article and I see that I wrote down the correct cap value in the margin on the parts list!!!  When I populate a board I like to color in the parts I install with a highlighter on the parts layout as I go.  So...... I printed a new set of instructions and thereby missed the cap value change.   

I'll try some of the other mods you have mentioned.

Thanks again., I'll report back with the results.

Yoiks!!  No wonder the cap in parallel with the 10k resistor did nothing; there was no treble to enhance!

[KazooMan]

Well, I tried all of the suggestions.  I removed the cap at the input and added the cap in parallel with the 10 K resistor to the volume pot.  I reduced the resistor from the sustain pot to 9.4 K ohms (I had no 10, so a put two 4.7Ks in series).  I also pulled out two filter caps that are on the Ross but not the Dynacomp.

The tone is "better" but this thing is still ripping most of the treble off the signal. 

I am going to let this project rest a while and get on to some other things (I had to clear my work area to make room to build a Claro walnut end table for the bedroom). 

Thanks again for all of the thoughtful suggestions.

[BAARON]

Is it actually ripping treble out of the signal (actually affecting the EQ of your sound) or is it just dulling the attack and edginess of the note?  If so, you might consider adding an attack control, such as most Ross/Dyna clones incorporate.  It helps you add a little more snap/bite to the start of a note.  Here's an example of one way to implement it.

http://fuzzcentral.ssguitar.com/ross.php (It's the "variable recovery mod")

[JDoyle]

Well - as posted by BAARON you may just not like the sound of a compressor on your signal.

However, if you want to try to get it more to your liking, you can try lowering the value of the 0.001uF cap in parallel with the 150k resistor on the output of the CA3080. As is, only frequencies below about 1kHz get the full gain of the OTA stage, due to that cap. If you lower the value, to say 470pF, more treble frequencies will get the full gain of the stage.

DO NOT LEAVE THE CAP OUT - because the 3080 (all OTAs actually) are current output devices, they can reach into the MHz range with ease, so the cap serves to limit this range and avoid the inevitable oscillation issues it would cause.

Good Luck!

Jay Doyle