Adding a Compressor circuit to a TS overdrive?

[Mr. Lime]

I saw the schematic of a Rothwell Love Squeeze Compressor which is pretty much a TS circuit.
Now I'm wondering if it would be possible to tweak the values of the compressor to fit the TS circuit very close and switch the output and the added compressor path.

If we take a look at the Love Squeeze, the most difficult part seems to be the feedback of the first stage which is limited by the 5k1 resistor parallel to the 10k pot.
The common TS value is 1M.

I could switch a 5k1 // 10k pot into the feedback but I hope to find a more elegant solution using the orignal 1M pot and just have to switch in the path between R9 and R13.

Any ideas?



Thanks

[blackieNYC]

If you reduce R16 a lot (10k?) can you get the gain of the TS but maintain your compression?  I don't know.

[slacker]

The maximum gain of the IC1A stage is about 150. If you want to use a 1Meg pot then remove R15 and change R13 to 6k8 and you will have about the same gain.

[midwayfair]

The LEDs are only there to clip transients. You CAN distort it by running up the gain, but here's the problem:

The gain pot is only 10K. You can drive the FET up to perhaps 25K (at least, that's about the most I get). To get it to distort, you're going to have to increase the gain pot. But giving it more gain will just trigger the envelope harder, so you have to keep turning the gain up. Eventually, though, your signal will be large enough that you're ALWAYS triggering the envelope, because you won't get any clipping until your signal is easily bigger than the 0.25V drop from the Schottky diodes.

To get around this, you need to move the compression control to be a threshold control, but it would be best to do it in such a way as to be able to reduce the envelope signal to less than unity.

You can also just build GrindCustom's Rotten Comp, which uses an overly large Attack control (100K pot in place of R9) in this same compressor to introduce clipping. Or simply put the compressor before a TS-style pedal.

[slacker]

If the fet only goes up to 25k then what I suggested won't work, I hadn't thought about resistance range of the fet.
What you can do instead is flip the problem round and make the TS mode work with the current setup. There's already enough gain, if you just connect the bottom of C11 to ground you've got a TS but with more bottom end. In TS mode if you change C11 to 10uF you'll have the same bass cutoff frequency response as a standard TS.

[PRR]

You can limit on the diodes, or in the linear limiter.

Trying to get it to do "both" may be really tricky.

[samhay]

^ You can limit on the diodes, or in the linear limiter.
Trying to get it to do "both" may be really tricky.

Maybe worth trying a feed-forward approach?

[Mr. Lime]

Thanks for your interest, guys!

Well, I would like to add the compression circuit to my "Tiger Driver".
My priority is the overdrive sound, I just like the idea of changing the circuit to a compressor with a flip of a switch.
It's not necessary that the compressor works 100% in a useful range.
I would be very glad to see a schematic of your suggestions, maybe it helps other people too, which are interested in this topic..

Here's my TS alike circuit: (not the updated one, but pretty close.



Thanks!!

[Mr. Lime]

How about switching R15 5k1 parallel with the 1M Gain pot and adding a 250k "Bloom/Attack" pot instead of R9?

The "compression" control in compressor mode would have a strange sweep but would it work so far?


Thanks

[Mr. Lime]

I decided to flip the problem and using the current setup.
The Tube Screamer has a gain of  1 + (551K/4.7K)= 107 at the first gain stage.

If I'm right, the gain on the Love Squeeze is 1 + (5000R/22R) = 228 which would mean that it's double the gain of the TS?

A 4PDT Switch would change the modes.
I am not sure about the second stage and it's tone control, how much would a tone control like a James tone stack influence the compression circuit?

Am I fine using a 100k Pot at the output for both modes?


Thanks

[slacker]

The love squeeze schematic you posted has a 5.1k resistor in parallel with a 10k pot setting the gain so it's about 3.4k in total not 5k making the maximum gain about 150. A 5k pot should work fine though. Why do you have a 500k pot in there as well? the 5k will do the job for both modes.
You might not need to disconnect R9 from the opamp it will probably work if you leave it connected, that saves you a switch as you can just switch between the outputs.

[Transmogrifox]

You have the right idea.  Really all you need to do is keep the original lovesqueeze circuit in tack, but change the 100 uF cap to 10 uF to match the TS 720 Hz cutoff.

In compression mode you switch in the FET, in TS mode you just short across the fet to ground or maybe reduce the mode switch to an SPST to do this and just leave the envelope detector active.

In either case you need the gain of the envelope detector so maybe you should replace the op amp with a common-emitter BJT gain stage with gain of ~4 or 5.

Then the only switch you need to change modes is to short the FET.

Having the high frequency boost of the TS will actually add a pre-emphasis curve not entirely unlike the JoeMeek compressor -- might be an improvement.

The TS tonestack de-emphasizes this, so it is in your best interest to just pull off the envelope detection with a BJT gain stage to control the FET, then just short the FET when you don't want compression.

I hope that makes sense.  If not I can sketch a schematic of what I mean.

[Transmogrifox]

here's the idea:


The 5k resistor is something you substitute with a pot.  Looks promising using an amplitude modulated 800 Hz tone in LTspice.  Adjust the .47u cap to taste for attack/release time.  The original 10u was pretty slow but then maybe it needs a bleed resistor.  The 0.47u simulated as still being nice and slow (like 100ms attack).

[Mr. Lime]

Hi Transmogrifox and thanks for yours suggestion!

Meanwhile I drew up a updated schematic.
I added the "attack" pot from the Rotten Komp and changed the switching.
As I am not a big fan of LEDs I would like to use 1N914 in pairs, not sure if they fit the compressor mode too?
Maybe it's worth switching them with the modes..

I always liked James tone stacks and the AMZ BMP TS does pretty much the same but has an additional mid control.
It's also possible to set that EQ flat.

If the BMP doesn't work so well with the compressor, I would wire a simply stupid tone control 2 infront of the output.

Your schematic looks impressive, maybe I should give this arrangement a shot! 
I am a totally noob at compressor circuits but the idea of the BJT gain stage seems cool, if it really works well.
At which point would you short the FET to ground, at R2?

Here's what I came up with:
Maybe it's time to get the ideas together, I am glad for any suggestions! 


R10 and C7 may be not needed, the Attack control in TS mode seems a bit strange too 

[Transmogrifox]

*EDIT*

I originally marked up your schematic, but looking at it decided it easier to update mine to include your ideas.  This better explains what I am thinking about this:


Main points:
1/(2*pi*R2*C2) = 723 Hz, which is right around where the TS does the 720 Hz emphasis
1/(2*pi*R1*C1) = 6.77 kHz, which more closely matches the relationship between the 500k pot and 52 pF cap.
R21 and C11 added to de-emphasize the 720 Hz boost, in similar manner to how TS does this.  Otherwise you will find you keep the tonestack balanced toward the dark side all the time and lose some of its useful range.

Hopefully that all is helpful.  I ran a circuit simulation on this to make sure the compression range seems about right.  I think it's pretty good here -- I might actually mod my Vox V847 (TS work-alike) to implement this, or maybe just build this circuit with the BMP stack because it looks really useful to have a compressor and OD built into one box.  I can see myself rarely bypassing this and more frequently just switching the OD/Comp switch.  I love compression on clean guitar.

This is a really sweet idea.  I'm glad I tracked onto this thread.  Nice work! 

[Mr. Lime]

Thanks a lot, Transmogrifox!

Over night I drew something similar based on your idea, now I updated it to your new schematic.
I like compression pedals for clean sounds too and think that this is going to be a really nice effect which is super versatile.
On my drawing, I added a boost trim to let the second op-amp gain stage act as a recovery stage after the BMP tone control, should be useful somehow.
Now it's got a variable gain boost from 2- 11.

I also thought about switching between 1N4148 and LEDs because I have something like the "son of a clay jones overdrive" in mind.
For additional Fuzz like sound I found the AMZ tone clipping circuit, maybe worth a try as it's switchable.

If everything seems alright I start to create a stripboard layout that I'll post here.
Please feel free to add any idea.

[Transmogrifox]

Looks pretty good other than U2 needs an R + C going to ground off the inverting (-) input to make the boost/trim do something.  Maybe try something like a 22k + 100n.

The cap could be smaller, say, 4.7n to 10n if you wanted a treble boost.  Maybe even a switch to switch between "bright" and normal boost.

Nice idea on the tonestack clipping.  Have fun building this !

[Transmogrifox]

One thing I forgot to mention is add a 22k resistor in series with Q1.  I noticed in simulation when this goes saturated it may forward bias and put a low impedance load on the op-amp causing some distortion, muddy up the sound.  The 22k will keep load impedance to the op amp high enough to control this.

Also I did some .wav file simulations in LTSpice to listen to it.  The TS overdrive sound is spot on.

For the compressor .wav simulation, I noticed that even with 2x 1N4148's in series there was still some distortion on the compressor.  Removing them or adding an LED in series eliminated the distortion.  Maybe you would want to consider using a dpdt to remove the clipping diodes with the same action as shorting the FET.

Another trick would be to add back-to-back LED's in series with the diodes and use the switch to short them during overdrive mode.   

Either way, leave a jumper in your layout to play with this so you don't have to kludge it if you find the diodes give more distortion than you like during compression.

[Mr. Lime]

Thanks for your great simulations, here's my idea for the diode problem.
It's an On-On-On switch, it allows TS mode operation with 1N4148, TS mode with LEDs in the middle position and Comp mode with the LEDs.

I think a treble boost switch would be nice but for my taste it's getting too much, we've already got 5 pots and 2 switches.
The modded BMP tone control can be adjusted to a treble boost too.

Now I'm thinking of a name for this pedal..  
Maybe "Lovedrive", this would share a part of it's name with the circuit it's based on.
One of my favorite bands have a song called Lovedrive too  
https://www.youtube.com/watch?v=itUkcJdQc8Q

Anyway, here's the schematic: (corrected)


Well, if you enjoyed this circuit maybe my other thread could be interesting too.
It's a CMOS overdrive with octaver blend, also something new but it is in the middle of development.
http://www.diystompboxes.com/smfforum/index.php?topic=109817.msg1005230#msg1005230

[Transmogrifox]

R18 (22k) should go at the base of the BJT.  In its present location it acts as a voltage divider with the downstream 10k and also doesn't entirely solve the problem.

Stick it right between where R2 and BJT base connect and you have a circuit that looks like it will do what you want.