Bearhug Compressor - schematic, demo, layouts, & forthcoming 1776 PCB!

[rankot]

Stupid me, I left compression pot hanging Now it works fine when I soldered it to ground, but I will try to put a switch to get distortion if needed . And one more thing - you mentioned in previous comments that R4 should be 4k7, while it is 2k2 in all schematics'. If I understood well, you didn't have a file to modify schematic, but this one shall be 4k7 in final design?

[midwayfair]

Stupid me, I left compression pot hanging Now it works fine when I soldered it to ground, but I will try to put a switch to get distortion if needed . And one more thing - you mentioned in previous comments that R4 should be 4k7, while it is 2k2 in all schematics'. If I understood well, you didn't have a file to modify schematic, but this one shall be 4k7 in final design?

It should be 2.7K in the new version. You want it to be the same as the source resistor so that when the resistance of the FET is driven way up, you minimum gain is ~1x. Matching the source and drain resistance ensures this. The dry path is based on the AMZ MOSFET booster -- the FET is just in place of the gain control.

[rankot]

One question about diodes - I put some fake 1N34A, but I believe they are just some kind of shottky - voltage drop on them is 0.33V. Should I leave R7 at 1K or more?

[midwayfair]

One question about diodes - I put some fake 1N34A, but I believe they are just some kind of shottky - voltage drop on them is 0.33V. Should I leave R7 at 1K or more?

I think it would be better overall to replace the two diodes (1n4148 or even 1N4001 are fine) because that puts kind of a hard "lower limit" on how big the signal has to be before it applies any compression (a little better in my experience), but if the compression's too much and you don't want to replace the diodes you can try anything between the 1K and the 1M. 1K is "might as well not be there," 1M is the original setting from the Love Squeeze's rectifier, and maybe 220-470K would probably kind of resemble the diode change, but again ... if you're replacing parts, the diodes are the better part to replace.

[rankot]

OK, I have replaced diodes - could't find 1N4148, but there was a couple of other Si, so I put them. I also put 1M pot instead of 1k for R7 and tried it, definitely almost no difference in compression when turning that pot, but no distortion when it is turned to 0.4-1k, so I finally left 1k resistor for R7.

After that, I tried to put 1M pot instead of R9 and SW1, and I liked it more, because I could really feel the change of decay value! So this could be an interesting and simple mod. And just to mention, I used BC550 instead of MPSA18, since I couldn't find it here. Now, I will box this little bear, and post some images when done!

PS: What will happen if I put bigger (5k) pot for compress?

[midwayfair]

I made the decay a switch to help with recall (i.e. fewer settings to remember) but I have built it at least once with a pot.

PS: What will happen if I put bigger (5k) pot for compress?

5KC might be okay but finicky to dial in. Once it gets higher than 2.2K in the stock circuit, more resistance doesn't really matter: the gain for the envelope will be 1x regardless, but you might misbias the transistor, which could result in worse compression. You could make the collector resistor bigger (like 4.7K) to make it less likely to misbias, but that will mean a lower compression threshold overall, and it already has plenty of gain with the 1K turned all the way to 0. The threshold is already very low in that situation, as the gain factor for the envelope signal will be a whopping 421x at max in the stock design (including the dry path gain).

[ericlaw02]

Hi there, I'm wondering if there's an etchable PCB layout for the 2.1 revision?

I looked on 1776 and Jon's website, but all that I could find was the perfboard layout and the PCB layout for the first version.

[midwayfair]

Hi there, I'm wondering if there's an etchable PCB layout for the 2.1 revision?

I looked on 1776 and Jon's website, but all that I could find was the perfboard layout and the PCB layout for the first version.

selfdestroyer on Madbean might have made one. And sometimes people make them from my perfboard layouts. I'm not immediately aware of one.

[iefes]

Hey, I recently built the Bearhug but I am experiencing some minor issues. In general everything worked right away but I noticed some distortion when playing chords or hard picked single notes. I tried it with Q2 and Q3 pulled or with a 2k7 in Q3s position but theres always the same distortion going on. That's why I'm quite confident that it is related to Q1. When all trannies are put back it's mostly audible on low comp settings but I think thats just because the gain gets lowered initially when a strong signal enters the circuit on higher comp settings.
I also tried another BS170 I had laying around and also fiddled a little bit with the voltage on Q1s drain. Currrently it's set around 5.2V. I tried a higher resistor value (3k3) for R4.
None of the attempts described above helped

I made two minor changes to the circuit: I included a compression indicator LED circuit using three resistors, two caps, an LED and a 2N3904 but the distortion is also audible when this circuit is disconnected from the main circuit.
I used a 25kB poti for the Level control, just because I didn't have a 10k. But I cannot imagine this is the reason I am getting a slightly distorted signal, what do you think?

I'd really appreciate any more ideas :-) Thanks a lot for this! The circuit kinda opened the technology of compressors for me.

cheers, iefes

Edit: Forget to tell, I built the V2.1 version on perf and used the schematic on Joshs page for reference. The original perf layout is slightly modified because I used a quite large trimmer and wanted to include a tone circuit first.

[Flaconsius]

I made the V1, and slowly was changing the values to the V2.
I don´t have much experience with Compressors but I find it very Subtle.
I had 3.7V in Q1´s Drain, so changed R6 to 62k (V1 Schem & Layout) and Drain´s Voltage went up to 4,7V.
I perceived a little more Compression but when play Chords or Strum hard with Humbuckers, I got some Overdrive, nice but don´t know if I wil want it always.
Another Issue is that when the effect is Turned On or Off, there´s a strong Pop noise.
Maybe, reducing the value of R1 both things will resolve. 

[blackieNYC]

Having a problem.  No sound, or at least very little sound - as I have turned my amp up really loud and there is some coming thru.  I don't think it is responding to the compression pot. voltages (does anyone have verified, working voltages for this at hand? We have to start posting that info in these threads. If I get it working....)
BAD voltages:
BS170   S:4.5v   G:4.0v  D:5.1v
MPSA18  E: .3v    B: .8v   C: 8v
5457    all zero or less than a millivolt    signal or no signal

All stock values for V2, as found here:http://1776effects.com/wp-content/uploads/2015/03/Bear-Hug-V2-1.pdf

I've torn this apart completely, twice, and measured all resistors, checked all parts values.  And left it alone for a day (crackerjack troubleshooting trick, or a means to suspend loud cursing fit.) Swapped all three transistors.  Almost no sound?  As in massive gain reduction?

[PRR]

> BS170  ...  G:4.0v

That has to be a lot lower. Like 2V. So you end up with several V from Drain to Source.

[blackieNYC]

OK, that's got FET flip written all over it.
Yes.  thanks.  And I will post "good" voltages:
BS170: D:5.3v   G: 3.3v   S:4.3v
MPSA18  E: 30 to 500mv (comp pot)   B: .6v   C: 7.2v
5457    all zero or less than a millivolt    signal or no signal

[rankot]

These are my values:
BS170: D:5.65V G:1.9V S3.0V
BC550c: E:0.06-0.46V B:0.60-0.96V C:8.2V (variations due to Compress pot position).
2N5457: 0V everywhere

I suppose my pedal works fine.

[blackieNYC]

Mine is not making me happy.  Very little compression, more like a little attenuation as you reduce the comp pot to 0 ohms.
I have the voltages 2 posts above.
The FET resistance is 300 ohm and goes up to 1500 ohms approx
The voltage at C9 goes from -.2 to -1.3v

I've tried R3 up to 4.3k, upped C9 to 10uf, swapped in schottky diodes.  At the max comp setting I get an unpleasant compression with an initial peak coming thru that sounds a bit sitarish. That same characteristic is there at the middle settings, just less of it.  I've gone to a humbucker guitar,  and get can get up to 2200 ohms across the FET and the same C9 voltages, but this is just not happening. I am getting no worthwhile compression at any setting.
As a reference, how close should this come to carbon copy?  I mean Custom Comp- which I imagine is much like the dynacomp. This was described as being less squishy than a dynacomp or ross, but I get nothing here.  Are old Dire Straights' sounds squishier than this one can offer?  But I think something is wrong, it does not sound good.

[midwayfair]

Yours is definitely not working properly. Your voltages are fine and your FET resistance at idle is normal. However, at max comp settings you should be able to easily exceed 5K of resistance and multiple volts (basically about 5, at which point the rectifier is just clipping.

If you socketted the FET, try removing it and checking that you get unity volume.

If you do, replace it with a 330R or 270R (or 300 if you happen to have it). You should get a ton of boost, and probably some clipping with humbuckers if it doesn't clip your amp.

I think a couple pages back I had a full diagnostic procedure for the sidechain, but I guess I should also say to double check the pinout of Q2. If it's backwards it will still amplify but only a tiny bit, which would explain why the compression pot still has some effect but isn't completely crushing your signal at higher settings.

Pictures might help. (I might have also missed which layout you used -- the PCB or my perfboard. I don't think I ever updated the perfboard layout to the trimpot version.)

The attack is very fast, especially at high comp settings, but the effect should be audible on pretty much every guitar by halfway on the comp knob. At the highest setting even a telecaster will usually be completely crushed. If you want I can send you some direct-to-interface recordings at various settings using whatever the closest guitar I have is so you have something to use as a sound reference.

As far as the clipping: There is almost always some clipping with the stock circuit with higher output pickups. It's a 9V circuit with anywhere from 6-10x gain at idle (depending on the FET used, but usually closer to 9x) with the stock values, so any guitar pickup that puts out a 1V or higher signal is going to clip the pedal. This would be true with pretty much any compressor except those with a much lower ratio. You either have to accept a small bass cut, accept some clipping, run the circuit on a higher voltage, or modify the drain and source resistors to lower the maximum ratio (it's the drain resistor divided by the FET resistance). However, the distortion is very soft sounding, almost like it's in the background, and I wouldn't describe it as "sitar-ish." If I'm guessing correctly that's the type of distortion I'd expect to hear on a misbiased transistor, so even though your voltages are correct, I'm not sure something isn't wrong. Your voltages are what I would expect using 2.7K on the drain and not 4.3K, for instance.

The carbon copy is a delay, so I'm not sure what that reference is.

Rankot's voltages are pretty close to my working voltages, though I think I usually end up a little higher on the gate. (I usually put the wiper of the trimpot, on that version, near 5.25V).

[blackieNYC]

Correction - MXR custom comp. got it the same time as my carbon copy. Always mixing the names.

Thanks Jon, I'll go thru it again. I'm still on the breadboard, so I can freely try more FETs. I'll try the side chain.  It's good to know you are seeing some measurable misbehavior.

[Flaconsius]

Well, after tinkering the Pedal, I´m back.
Changed R6 to 56k so Q1´s Gate voltage went up to 6,7V. I think it isn´t harmful to any component 
Also left R4 in 2k2 value.
Reduced R1 to 2M7, so the Input Impedance is not so High.
After all these changes, still got the little Overdrive when playing Chords.
John recently said that will always happen, but I´m thinking abot changing something or adding a Switch for having a Clean Comp sound.
Here´s a little Demo. First with the Line 6 POD Farm´s V Comp, and after, with the Bearhug Comp.
Regards.
https://soundcloud.com/flaconsius/prueba-compres


Been tweaking it a little and got what I wanted :-P
Just replaced Q2 for a 5089.
The result is an Ultra Clean (and very Shiny) Compressed Sound.
I am thinking of make my own Board design with both transistors and a Switch to Choose betwen "Slightly Overdriven" and "Chimey Cleany" :-)
Here´s the same Intro for Demo, and, again, first is the VST Plugin VComp, and after, the Bearhug.
https://soundcloud.com/flaconsius/prueba-compre-sor2

[blackieNYC]

Jon, I've got a scope - what should I be seeing at the FET?  what is the "basically about 5..." - where should I be seeing that?

I've got everything you describe - lots of boost with a 330 resistor in place of the D and S of the FET, a faster meter shows me peak resistance there of 3k or more when the FET is back in. Voltage at C9 swings up to -2v or more,  swapping all three Qs again.  Still the envelope is very sitar-like, though I did get the volume out of it

[midwayfair]

Not sure what you'll see on a scope ... However, the FET gate is a peak-to-peak rectified DC signal that appears when you play.

When I measure with my digital multimeter using my strat at full comp, at the gate of the FET, I see as much as 5V on a hard strum and around 3V with moderate playing. (It won't go much higher than 5V because eventually the sidechain will clip, and you'll lose a bit from the decay loading the output of Q2 and the diodes.)

The other things I can think of to cut a bit of distortion are removing the zener (just be aware that at some point, it's possible to pop the MOSFET, so sockets are a good idea), and lengthening the attack to prevent "compression distortion" or whatever you want to call it when a fast attack time rides the wave peaks and makes it sound like distortion (put a resistor between Q2 and its output cap ... 2k2 would be an extra 10mS).

I have a built-up version 2 PCB in my bin. Do you want to borrow it as a reference? (It seems to have a 100K dual gang attached as the volume pot for whatever reason, but that really won't make a meaningful difference.)

It's also simply possible that the pedal simply doesn't mesh with your gear, expectations, playstyle, or whatever. Some amps might reveal shortcomings that others don't, for instance. I understand being frustrated that something SHOULD work, but sometimes, it's just "this isn't the right circuit for my guitars" ... it's why I have a shelf full of compressors, distortion boxes, etc.

---

I may have forgotten to put this in this thread, but soemtimes when I build these for people using higher-output pickups, I put a high-pass on the input. It's usually on a switch, and either a -3dB (1/2 R2) or -6dB (=R2) shelf filter created by putting a capacitor in parallel with the series resistor, usually around 150 or 200Hz. (It depends on the guitar.) This helps stave off some compression at the expense of some volume (but there should still be plenty available).