Very simple compressor. Works well IMO

[MartyMart]

No problem Brett, I was interested in the circuit and I'm pleased that I "understand" what's going on !!
I've had the 100k in series from early on, in an attempt to "load down" my HB gtr, I dont have the problem
with a strat or Tele, perhaps it would be better at 200/300k, but it's working well and any more
de-soldering will ruin my veroboard !
I like the Jfet BJT amp idea too, let me know if you have an improved idea with this,  great


Marty.

[David]

Hey, guys!  Somebody try this with a bass, please...

[MartyMart]

Yup, works well with bass too ! ( deluxe P-bass )
Perhaps just use bigger in/out caps like 470n or even 1uf poly's

MM

[Arn C.]

Just out of curiosity, what would happen if another of the same IC was piggybacked on the existing IC?

Peace!
Arn C.

[David]

Yup, works well with bass too ! ( deluxe P-bass )
Perhaps just use bigger in/out caps like 470n or even 1uf poly's

MM

REEAAALLLLY...  Hmm...  Maybe I'll just build this puppy and sell the Alesis 3630 I just got yesterday...

Hey,  MM!  What are you running the Aussie into?  Next question:  how does it handle slapping and popping?
Another request (I haven't got a breadboard right now):  how does it do as a peak limiter instead of a sustainer?

[psw]

Hope you guys don't mind me resurfacing this thread, but the links given seem to be broken  

Does anyone have the schematic on this one? (looking into such a compressor for the sustainer project over at Project guitar now on it's 108th page! http://projectguitar.ibforums.com/index.php?act=ST&f=16&t=7512&st=1605#entry289046)

Pete

[markm]

I was after the schem too and only found the one you already have.
I did a layout for this comp and used Torchy's Vero layout for some of the additional componenets.
Check it out; 
http://aronnelson.com/gallery/album15
The PNP is in my Layouts Gallery too so, maybe this can be of help.

[KerryF]

Mark, how does that Aussie comp sound?  Does it sound like a normal compressor?

Did you get my PM?

[markm]

Mark, how does that Aussie comp sound?  Does it sound like a normal compressor?

Did you get my PM?

I like the sound but, it needs some type of recovery for the high end. Though this can be a typical problem for compressors,
it does lack some sparkle.
PM'd you back just now 

[brett]

Hi
sorry about the disappearing schemo.  I changed internet provider.
The file is somewhere at home, but it's so simple that anyone can make one in 5 minutes.

Here's the different sections needed:
1.  A 386 op-amp, set up for moderate gain (100 would do).  The application note shows how.
2.  A mosfet (BS170, 2N7000 etc) with its drain connected to the op-amp input and the source to ground.  This mosfet turns on when the output is high enough, and some signal is conducted to ground*.
3.  The gate of the mosfet is connected to a (+) of a small electrolytic capacitor (say 2.2uF).  The (-) of the cap is earthed, and a resistor is connected in series with the cap.  This resistor "bleeds" voltage from the cap and sets the "release" time.  It could be a 100k pot or a resistor between about 10k and 100k.
4.  A schottky or germanium diode rectifies the AC output of the op-amp (see 6 belowie after the optput cap) to the (+) of the cap.  Unbanded end to the op-amp and banded end to the cap.
5.  A very small linear pot (1k or 5k) used as a voltage divider.  lug 3 goes to the op-amp output (after the output capacitor).  lug 1 goes to ground.  lug 2 goes to the unbanded end of the diode.  This pot controls the amount of compression. 
6 (optional).  A small pot (10k?) connected as a variable resistor in series with the diode (this sets the attack time, but affects the compression a bit too, IIRC).
 
How it works:
With small signal input, the mosfet is "off", and all of the signal goes to the op-amp, which puts out maybe 1 or 2 volts. 
With a large input signal, some of the output gets to about +3V, and a bit less gets past the diode and charges the cap.  If the cap and gate get to 2.5 V, the mosfet conducts some signal to ground and lowers the output.  No matter how large the input signal, the output never gets above +/- 3V.

* this is why there's some treble loss.  The input impedance of the 386 is only 50k to start with.  Once compression starts it falls even lower, loading the pickups and reducing treble.  A simple JFET or BJT buffer at the input would totally solve this problem.  I should have thought of it earlier.   

[markm]

Thank you Brett,
I'll have to mess around with it a little bit and perhaps incorporate the buffer into an updated layout.
My hats off to you for going out on a limb to create a unique Comp circuit.
I do think there is alot of potential here with this one.

[brett]

Hi
I knocked out a schemo:

I think it's about right.
There's a few components that might be tailored to taste, especially that 47k resistor.  47k give about 0.6 seconds release time.  That sounds about right, but it might be out by 2x or 3x.  IIRC I originally had a huge resistor that held for 30 seconds or something ridiculous.

Almost all of the components could be substituted with other values.  Except:
The 1M tie-down resistor on the gate of the JFET.  It needs to be kept high (>220k).
The 0.1uF cap between the JFET and 386.  It fixes the bass frequencies that are trimmed (0.1uF= about 30Hz and less), preventing compressor "breathing" or "wobbling".
The 1.2k resistor between pins 1 and 8.  This resistor needs to be 2k or less, otherwise there won't be enough gain to turn the mosfet on.  Too low (100 ohms?) and it might get weird.
Although the original showed a 1N4148/1N914 diode, but a Schottky or a Ge would be better because they have less voltage loss and make the compression more controlled.

Just about everything else is quite tweakable.
cheers

[psw]

Thanks for the "improved" design Brett. I think the buffer is a good idea.

As for my interest, I suspect I would need a significant change to be able to drive an 8 ohm sustainer with it, but the ideas are intriguing and certainly food for thought...

Well done, I love the simplicity of it... pete

BTW... the infamous Project Guitar DIY Sustainer Thread can be found at...
http://projectguitar.ibforums.com/index.php?showtopic=7512&pid=289349&st=1635&#entry289349

[Noplasticrobots]

The link to the schem isn't working (404 Error) but I'd love to check out a simple compressor!

[amz-fx]

One thing to consider is the body diode of the mosfet....  if the input is allowed to swing more than -0.65v then it will begin to clip. 

A series resistor after the 0.1uF and before the mosfet connection to pin 3 might be a good ideas....  try 4.7k

regards, Jack

[billings]

Why hasn't anyone run with the TL072 idea?  I don't see any particular reason why it wouldn't work for this application.  I know very little, however.

Wouldn't the op amp clip if the input drops below ground anyway?

[markm]

I did a layout for the new tonight,
I haven't tried it out yet but, never-the-less, here it is.
Any takers wanna try it out?

[timo]

Any idea on how the circuit would react to being operated at 18 volts? Of course using 386:s that can handle it (LM386-4 or JRC386BD) and sufficient caps. At least the Ruby amp has been reported as being "cleaner" @ 18V

[Noplasticrobots]

Wow, this looks too simple! I can't wait to build this one! No specifics on the JFET or MOSFET? Would a J201 and 2N7000 work fine? I don't have any Schottky's lying around so I'm gonna try mine with a 1N34A and report back here with the results.

[markm]

Ya know, I didn't mean to actually have the 1N34a in there, I was going to experiment with the germ diode but, it's not what the schem calls for. Think it'll work