Simple Transistor Compressor

Stompboxes circuits published in magazines, books or on DIY electronics websites.
User avatar
mictester
Old Solderhand
Information
Posts: 2887
Joined: 11 Sep 2008, 20:29
my favorite amplifier: Mesa Boogie, Roost Sessionmaster, AC30
Completed builds: Hundreds! Mostly originals, a few clones and lots of modifications.
Location: Somewhat closer to Amsterdam than before!
Has thanked: 27 times
Been thanked: 784 times
Contact:

Post by mictester »

No unusual ICs required! Uses just four cheap transistors! Most of you will have all of the parts in your junk boxes!
SimpleCompressor.jpeg
SimpleCompressor.jpeg (32.4 KiB) Viewed 5026 times
The input impedance was originally designed to be about 10k as it was going to be used after another effect pedal. If you want to connect it first in line, before all your other toys, you'll need a higher input impedance: increase the two input bias resistors to 330k or 470k and reduce the input capacitor to 100n or even as low as 47n. Because the circuit uses a crude transistor shunt for level reduction, you'll find that there's a quite tiny amount of second harmonic distortion added, but it's not objectionable.

The transistors used are not critical - any PNP and NPN medium gain silicon types will work well. The coupling capacitors can be bigger or smaller by a factor of two, without affecting its operation. Many of the component values were "chosen" just because they were lying around on the bench! Feel free to experiment with the values - you may find some magic values that really suit your guitar and your playing style!

In the next few weeks, I expect to see this in a hand-painted diecast box, full of "mojo" resistors and capacitors, with the faces of the ordinary, cheap silicon transistors sanded, lots of potting compound spilled inside and for sale on Ebay for $400!!!
"Why is it humming?" "Because it doesn't know the words!"

User avatar
Broadcaster
Information
Posts: 19
Joined: 15 Aug 2007, 21:09
Location: Here, there and everywhere
Has thanked: 20 times
Been thanked: 4 times

Post by Broadcaster »

So, the LFC style reduction seems to work indeed... you are productive I have to say! I should build this just to tame my curiosity.

User avatar
mictester
Old Solderhand
Information
Posts: 2887
Joined: 11 Sep 2008, 20:29
my favorite amplifier: Mesa Boogie, Roost Sessionmaster, AC30
Completed builds: Hundreds! Mostly originals, a few clones and lots of modifications.
Location: Somewhat closer to Amsterdam than before!
Has thanked: 27 times
Been thanked: 784 times
Contact:

Post by mictester »

Broadcaster wrote:So, the LFC style reduction seems to work indeed... you are productive I have to say! I should build this just to tame my curiosity.
Build it by all means - it works as well as a Dynacomp! If you want to use it connected directly to your guitar, you'll have to increse the input impedance as described!
"Why is it humming?" "Because it doesn't know the words!"

User avatar
branleur
Information
Posts: 18
Joined: 14 Nov 2009, 21:11
Location: belgium
Been thanked: 3 times

Post by branleur »

Thanks for sharing :D I just need a simple compressor without OTA for another project. It seems I lucky 8)


EDIT: Has anyone have a EHX LFC corrected schematic

User avatar
mictester
Old Solderhand
Information
Posts: 2887
Joined: 11 Sep 2008, 20:29
my favorite amplifier: Mesa Boogie, Roost Sessionmaster, AC30
Completed builds: Hundreds! Mostly originals, a few clones and lots of modifications.
Location: Somewhat closer to Amsterdam than before!
Has thanked: 27 times
Been thanked: 784 times
Contact:

Post by mictester »

branleur wrote:Thanks for sharing :D I just need a simple compressor without OTA for another project. It seems I lucky 8)


EDIT: Has anyone have a EHX LFC corrected schematic

I haven't got the EHX LFC circuit, but look at the next thread!
"Why is it humming?" "Because it doesn't know the words!"

User avatar
earthtonesaudio
Transistor Tuner
Information
Posts: 1244
Joined: 28 Jan 2008, 04:00
Completed builds: Metal Simplex, Fuzz Factory, two Fab Echos-modded, Noisy Cricket, Earth & Space Wah, TS-7-modded, passive xover/splitter box, opamp fuzz
Location: Bloomington, Indiana
Has thanked: 11 times
Been thanked: 73 times

Post by earthtonesaudio »

It would probably be worthwhile to swap the first stage for something biased a little more conventionally. Actually the part values make it look like Q1 was supposed to be an NPN emitter follower... is that a typo?
rocklander wrote:hairsplitting and semantics aren't exactly the same thing though.. we may need two contests for that.

User avatar
mictester
Old Solderhand
Information
Posts: 2887
Joined: 11 Sep 2008, 20:29
my favorite amplifier: Mesa Boogie, Roost Sessionmaster, AC30
Completed builds: Hundreds! Mostly originals, a few clones and lots of modifications.
Location: Somewhat closer to Amsterdam than before!
Has thanked: 27 times
Been thanked: 784 times
Contact:

Post by mictester »

earthtonesaudio wrote:It would probably be worthwhile to swap the first stage for something biased a little more conventionally. Actually the part values make it look like Q1 was supposed to be an NPN emitter follower... is that a typo?
Not at all - it was originally built that way. However - an NPN would work, though it would be better to have some gain in the first stage.
"Why is it humming?" "Because it doesn't know the words!"

User avatar
DougH
Transistor Tuner
Information
Posts: 1087
Joined: 18 Aug 2007, 04:53
Has thanked: 4 times
Been thanked: 22 times

Post by DougH »

mictester wrote:
The input impedance was originally designed to be about 10k as it was going to be used after another effect pedal. If you want to connect it first in line, before all your other toys, you'll need a higher input impedance: increase the two input bias resistors to 330k or 470k and reduce the input capacitor to 100n or even as low as 47n.
With a BJT it's not a good idea to raise the base biasing network resistors too high. For one thing, the input impedance of an emitter-follower will always be dominated by (hfe + 1) * Re (Art of Electronics 2nd ed. pg. 66). So scaling up those base resistors really doesn't raise the Zin of the follower. It just reduces the bias current supplied to the base. Secondly, a good rule of thumb is to keep the parallel combo of the base resistors much lower than hfe * Re, in order to bias it correctly (Art of Electronics 2nd ed. pg. 70).

So a BJT emitter-follower with high value base resistors still has a low input impedance and may not bias correctly. A better solution is to use a device with a high Zin like a JFET for example, for your input buffer.

Your compressor circuit looks nice. A few years ago I came up with a 2 transistor/MOSFET pin diode compressor for guitar. It wasn't the greatest compressor in the world but was fun to play with and a real simple circuit.
"You have just tubescreamered or fuzzfaced yourself " -polarbearfx

User avatar
DougH
Transistor Tuner
Information
Posts: 1087
Joined: 18 Aug 2007, 04:53
Has thanked: 4 times
Been thanked: 22 times

Post by DougH »

Oh, I just noticed- your input stage is a gain stage not an emitter follower. :oops:

Same basic principle applies though. For higher Zin use a different device (MOSFET or JFET).
"You have just tubescreamered or fuzzfaced yourself " -polarbearfx

User avatar
DDD
Information
Posts: 5
Joined: 27 Dec 2007, 17:31
Location: Almaty, Kazakhstan

Post by DDD »

The device seems to produce considerable distortion:
1. The very first stage has too much gain to process the guitar attack without severe clipping
2. The BJT "shunt" usually begins to distort signal at approx. 0,6 Volts peak-to-peak.
So, those two factors should introduce considerable non-linearity.
At the same time the first stage bias and operation points are very unusual.

User avatar
Zodd42
Breadboard Brother
Information
Posts: 73
Joined: 06 Jan 2010, 14:25
Completed builds: None
Been thanked: 1 time

Post by Zodd42 »

DDD wrote:The device seems to produce considerable distortion:
1. The very first stage has too much gain to process the guitar attack without severe clipping
2. The BJT "shunt" usually begins to distort signal at approx. 0,6 Volts peak-to-peak.
So, those two factors should introduce considerable non-linearity.
At the same time the first stage bias and operation points are very unusual.
So, does this thing work or not?
I'm buying the parts tonight, wondering how to wire this thing, not great with schemos yet . . .

User avatar
DougH
Transistor Tuner
Information
Posts: 1087
Joined: 18 Aug 2007, 04:53
Has thanked: 4 times
Been thanked: 22 times

Post by DougH »

Always try stuff on the breadboard first before committing to building it...
"You have just tubescreamered or fuzzfaced yourself " -polarbearfx

User avatar
blueduck577
Information
Posts: 5
Joined: 03 Jan 2009, 07:08
Has thanked: 1 time

Post by blueduck577 »

are those caps supposed to be 47u or 4.7u?

User avatar
mictester
Old Solderhand
Information
Posts: 2887
Joined: 11 Sep 2008, 20:29
my favorite amplifier: Mesa Boogie, Roost Sessionmaster, AC30
Completed builds: Hundreds! Mostly originals, a few clones and lots of modifications.
Location: Somewhat closer to Amsterdam than before!
Has thanked: 27 times
Been thanked: 784 times
Contact:

Post by mictester »

blueduck577 wrote:are those caps supposed to be 47u or 4.7u?

They're 47uF. :thumbsup
"Why is it humming?" "Because it doesn't know the words!"

User avatar
blueduck577
Information
Posts: 5
Joined: 03 Jan 2009, 07:08
Has thanked: 1 time

Post by blueduck577 »

OK I just breadboarded it. Some notes:

First I put a JFET source follower on the front end to increase the input Z. Also, I tried building it with a PNP transistor (BC327) in the first stage but it clipped and gated horribly so I put in an NPN (2N4401) to make it an emitter follower.

If anything this compressor is very subtle... I don't have any extra dpdt switches right now to set up a bypass to A/B but I didn't hear significant compression really. I figured maybe my guitar's signal was too weak to get over the diode forward voltage so I replaced the diodes with 1n34a and I still didn't hear much of a change. However when I pinched the leads of the diode that doesnt go to ground with my fingers I did hear the gain reduction...

your thoughts?

User avatar
Zodd42
Breadboard Brother
Information
Posts: 73
Joined: 06 Jan 2010, 14:25
Completed builds: None
Been thanked: 1 time

Post by Zodd42 »

Built a piezo buzzer pickup for my Cort flat top. It hurts to play it through the amp which is why I'm on this thread. To reiterate Blueduck up there, any thoughts on his findings?

I have one of those switches, still trying to track down those caps. When I finish the build I'll post my results.

User avatar
scottywompas
Information
Posts: 25
Joined: 23 Sep 2007, 20:09
Location: Sunny California
Contact:

Post by scottywompas »

How about simple JFET buffer on the front end.

Me thinks that would work, correct?

Scott

User avatar
Zodd42
Breadboard Brother
Information
Posts: 73
Joined: 06 Jan 2010, 14:25
Completed builds: None
Been thanked: 1 time

Post by Zodd42 »

I'm not sure 'Wampas.
Aren't compressors working when you can't hear them though? I mean a good compressor should shape your signal in such a way that cuts through the band unnoticed? Hmmm, my communication skills are subpar today . . .

User avatar
PierreBPSFSB
Information
Posts: 10
Joined: 07 Mar 2010, 16:08
Location: Paris, France

Post by PierreBPSFSB »

DDD wrote:The device seems to produce considerable distortion:
1. The very first stage has too much gain to process the guitar attack without severe clipping
2. The BJT "shunt" usually begins to distort signal at approx. 0,6 Volts peak-to-peak.
So, those two factors should introduce considerable non-linearity.
At the same time the first stage bias and operation points are very unusual.
I also agree !
- Clipping can even rise well before 0.6Vcc, depending the BJT attenuation ratio. (fortunately here, we have some gain AFTER the BJT, which decrease its AC level )
- Bias first stage: also, if we have here a PNP (why ?) at the first high gain stage (# -150 unloaded), it would be better to have a Bias of 470K instead of 22K to +9V, and 2.7M instead of 22K to GND, in respect to Vcc/2 at the collector out (2.7M should actually be trimmable 1Mpot+1.8M, because the high Gain). IMO, a gain of # 20 would be sufficient firstStage.
As someone said, for a guitar purpose, a buffer is mandatory before (JFet or BJT Emitter Follower), because the actual input impedance Zin of this stage is # h11= 20K :cry: if ic=450µA: well below the 2xRbias bridge in //
- Also, at fast attack guitar signal, the "charge injection" of the base current of the gain control would create some large "DC offset transient step" at the BJT attenuator out: for a guitar.. why not like a "slap" ! :?:

User avatar
Hotrats
Information
Posts: 37
Joined: 31 Dec 2007, 01:46
Been thanked: 19 times

Post by Hotrats »

just a few ideas to try here ... (some already mentioned)

- replace the two 22k resistors by 1Meg
- stick a 1k/22uF parallel combo in the emitter circuit of the front PNP device - the stage could be NPN version, no diff
- try Ge diodes instead of Si ... will give lower threshold
- if you replace the 15k resistor by 100k pot that will give you a rough "ratio" control
- the 680r emitter resistor in the output buffer could be a little higher, 4k7 would be fine there (less noisy)
- adjust/reduce input cap to taste (0.02u perhaps ?)

good effort !
HR

Post Reply