Really Cheap Compressor  [documentation]

Original effects with schematics, layouts and instructions, freely contributed by members or found in publications. Cannot be used for commercial purposes without the consent of the owners of the copyright.
User avatar
Rolbista
Solder Soldier
Information
Posts: 178
Joined: 01 Feb 2011, 19:56
Completed builds: Mosrite Fuzzrite clone
Location: Poland, Gdynia
Has thanked: 4 times
Been thanked: 3 times

Post by Rolbista »

mictester wrote:You could do that, but you'd have to match everything and the chances of finding two opto-couplers with exactly the same characteristics and matching LEDs and so on are very small!
Yeah, it would be very hard to match them but I guess one could use and audio interface to run left channel through it and then right channel recording them back to the daw and recombining both track back to stereo. Maybe not drums, but general mixbus compression? Just an idea to toy with

User avatar
Dix
Information
Posts: 13
Joined: 12 Mar 2016, 20:59
Has thanked: 2 times
Been thanked: 7 times

Post by Dix »

OK, apparently I'm missing something here, so hopefully someone can give me a lesson in what that is.

First off, I'm pretty much a novice in circuit design... I can understand what changing component values should do for the most part... but beyond that it gets fuzzy fast. :mrgreen:

In other words I can build/troubleshoot a project from a schematic, & mod 'em a bit, but not much past that.

Also, I've only recently gotten into using LTspice to simulate circuits so maybe I've got something screwed up there... hard tellin' not knowin'.

Anyway, here's my delemma...

Modeled up this project in Spice to check out the various mods mentioned thoughout this topic to get feel for where I may want to end up at for values.... I found the "improved version" schematic in the "other" topic... adding the caps for a "pre" & "post" high freq. emphasis to reduce noise (or thinking maybe I'd just use a cap across the input resistor for a "bright" switch)

Image
Adding an extra capacitor across the 220k input series resistor, and adding a capacitor across the output level pot will give pre-emphasis on the way into the circuit, and de-emphasis on the way out, thereby maintaining an overall flat frequency response:
The problem I ran into is that the 10n cap across the output pot doesn't seem to do anything in the sim... I'm getting the "pre-emphasis" in the freq. response trace (or so it seems), but nothing about the response trace changes with or without the de-emphasis cap.

Here's a screenshot of the circuit in the sim...
circuit.png
And the results... sustain pot modeled as "off" (max resistance)... output taken as if "level" pot is on full.
output.png
output.png (6.52 KiB) Viewed 3947 times
V(out1) BLUE - C17 & C18 both removed (original schematic)
V(out2) YELLOW - C17 & C18 added ("improved" version)
V(out3) RED - C17 (10n) only connected - ("de-emphasis" cap only)
V(out4) GREEN - C18 (470p) only connected - ("pre-emphasis" cap only)

OK, can someone point out what an idiot I am & tell me what I'm missing?

User avatar
Dix
Information
Posts: 13
Joined: 12 Mar 2016, 20:59
Has thanked: 2 times
Been thanked: 7 times

Post by Dix »

Rolbista wrote:I guess one could use and audio interface to run left channel through it and then right channel recording them back to the daw and recombining both track back to stereo. Maybe not drums, but general mixbus compression? Just an idea to toy with
If you're going to combine a "dry" & "effect" track back together remember you'll need to reverse the phase on one of them as the op amp reverses the phase of the signal in the meat of the circuit's freq range.

Here's the trace from my Spice model with the signal phase plot added (dashed trace) in case anyone is curious...
LDR phase.png
LDR phase.png (7.34 KiB) Viewed 3948 times

User avatar
Dix
Information
Posts: 13
Joined: 12 Mar 2016, 20:59
Has thanked: 2 times
Been thanked: 7 times

Post by Dix »

OK... above question retracted... I think I've found my answer... isn't the "improved" schematic missing a resistor (10k) on the "de-emphasis" part between the 2.2u & 10n caps?

At least Spice seems to think so.

I actually ran into the answer while working on another circuit sim (Gibson Victory Artist bass preamp) that includes a switchable 1kHz midrange notch filter.... it was while working on troubleshooting that sim that I noticed how the low-pass half of it was designed & it rang a bell on this one.

Gibson Victory Artist mid notch
mid-notch.png
mid-notch.png (2.97 KiB) Viewed 3933 times
So I went back to this project & added a 10K resistor just ahead of the 10n cap & presto... the "de-emphasis" appeared in the output trace.

Now, that does create about 6dB of loss at the output, but considering that this circuit seems to be capable of providing near line-level output (if you have a hot humbucker driving it) that shouldn't be a big issue.

Here's my model of the original...
orig.png
orig.png (3.44 KiB) Viewed 3933 times
And the "fixed" de-emphasis...
fixed.png
fixed.png (3.49 KiB) Viewed 3933 times
10n didn't seem to remove all the "pre-emphasis"... but 22n modelled a perfect match of the original output.
results.png
results.png (7 KiB) Viewed 3933 times
BLUE V(out3): Original Circuit
YELLOW V(out4): "Improved" Circuit
RED V(out7): "Improved" CIrcuit w/added 10K resistor
GREEN V(out2): "Improved" Circuit with 10K resistor & 22n cap
(Red & Green traces have the input voltage doubled to raise the traces by 6dB for ease of comparison)

User avatar
Gobi
Information
Posts: 16
Joined: 06 Feb 2015, 22:35
Has thanked: 3 times
Been thanked: 5 times

Post by Gobi »

Interesting, but im really wondering if LT spice will emulate the LDR LED combination correct?
Could you upload this file ?

User avatar
Dix
Information
Posts: 13
Joined: 12 Mar 2016, 20:59
Has thanked: 2 times
Been thanked: 7 times

Post by Dix »

If I'm getting what you're asking.... I doubt it.

First thing you'd need is a model for the LDR.... I have no idea if Spice even could model an LDR... well, I suppose someone with way more expertise & knowledge than I have could likely come up with a simulation for one, but that's something that's over my head.

What I have listed as "LDR1" is actually just a plain 1Meg resistor which models the circuit output as though the "sustain" pot was at minimum. (even though the screenshot of it above shows the pot drawn as at full)

I only added that part of the circuit to the sim so I'd have a test point to see what the voltage & freq response of the circuit driving the LEDs was.

At any rate, if anyone wants the LTspice sim I did to play with I've attached it. I also included the model files for the TL072 opamp, which LTspice doesn't have in it's default library.

There are 3 models in the sim... the original circuit, and 2 versions of my "fixed" improved version... one with the de-emphasis as 22n that I finally arrived at & a third version with that cap at 10n as per the original "improved" version... that produces a roughly 5dB peak at about 5.5kHz... which would make for a nice "bright" switch if a switch was added to select between the two.

Also, on the resistor I added to get the de-emphasis part working I lowered the value to 9K to better match the output curve of the original... I plan on using 2 18Ks in parallel (because I'm anal that way)... but really 10k is fine... the difference is only about 0.5dB across the useable spectrum.

Only thing I'm wondering about at this point is that with the "improved" version the freq response of the circuit driving the LEDs now has high freq boost starting at about 600Hz & rising up to about +10dB at 6K-ish... how that may effect things I'm not sure... or if it even matters.
comp.png
comp.png (10.11 KiB) Viewed 3916 times
Attachments
LDRcomp.ZIP
(3.99 KiB) Downloaded 150 times

User avatar
Gobi
Information
Posts: 16
Joined: 06 Feb 2015, 22:35
Has thanked: 3 times
Been thanked: 5 times

Post by Gobi »

Thanx a lot!!!!
My english is not verry good, I'm from Germany.
I'm brand new with LT spice, so for me it will be interesting to play around, as meanwhile i'm kind of familliar with this ciruit.
Give me a few days.....
As far as I understood, you are more interessted in the frequenzy stuff, so it might work without emulating the LDR.

User avatar
Dix
Information
Posts: 13
Joined: 12 Mar 2016, 20:59
Has thanked: 2 times
Been thanked: 7 times

Post by Dix »

Gobi wrote:Thanx a lot!!!!
My english is not verry good, I'm from Germany..
Don't sell yourself short, your English is fine... I'm just a simple dumb*ss mountain redneck & I can understand it... that being the case, you're good.

In fact, your English is better than some natives of this country... but don't get me started on that rant. :mrgreen:
Gobi wrote:I'm brand new with LT spice, so for me it will be interesting to play around, as meanwhile i'm kind of familliar with this ciruit.
Me too... if something like Spice had been around 20 years ago I might have gotten more into electronics than I did back then... I've been playing with it like a kid with a new toy for about a month.
Gobi wrote:As far as I understood, you are more interessted in the frequenzy stuff, so it might work without emulating the LDR.
That is correct... basically I wanted to make it so it added as little of it's own "color" to the tone as practical (although, the "bright switch" option I mentioned above has some appeal as an addition). I also liked the idea of the noise reduction aspect of the "improved" design, just couldn't seem to get it to work properly until I was messing around with something else that clued me in as to why.

Actually, just after I uploaded that version, I played around a bit and added a "de-emphsis" to the LED driver circuit (10K resistor follwed by a 10n cap to ground just ahead of the LEDs) to remove the treble boost from that part of the circuit as well.... so now even that part of it has the same freq response as the original.... not sure that really matters much in the grand scheme of things but I did it anyway.

Here's what that looks like if you want to add them into the circuit (R60 & C35)...
led.png
led.png (2.42 KiB) Viewed 3902 times
sust.png
sust.png (6.01 KiB) Viewed 3902 times
At this point I'm happy with the freq response... my final build will likely include a second opamp... half of that being a buffered bypass (not a fan of unbuffered bypass unless I'm sure it's never going to be the 1st effect in the chain)... & the other half as a simple phase inverter on the input.. that way the output is back in phase with the input.

User avatar
Gobi
Information
Posts: 16
Joined: 06 Feb 2015, 22:35
Has thanked: 3 times
Been thanked: 5 times

Post by Gobi »

I really feel ashamed.... AGAIN my sketch above contains a mistake. Would please someone of the admins delete them... dont know, why I could'nt do it.

I found the mistake, when I tryed to rebuild it - now it should be right:
Image
(there is now need fot the red LED)
hope this will be the last time... sorry again

User avatar
karul
Cap Cooler
Information
Posts: 504
Joined: 02 Jan 2011, 20:28
Has thanked: 597 times
Been thanked: 259 times

Post by karul »

Gobi wrote:my sketch above contains a mistake. Would please someone of the admins delete them... dont know, why I could'nt do it.
go in user control panel, select manage attachments

User avatar
kco_no
Information
Posts: 3
Joined: 26 Apr 2016, 12:38
Has thanked: 1 time

Post by kco_no »

I don't get what does that dashed line mean.
Can You please explain? Thank You

User avatar
Gobi
Information
Posts: 16
Joined: 06 Feb 2015, 22:35
Has thanked: 3 times
Been thanked: 5 times

Post by Gobi »

It's a stereo potenitiometer (the german way to show it - how is it shown in the states?) . The blue dots are showing the direction (if one gets up the other one gets low)

Extra patience is needed for a very clean 12V - otherwise you will hear a hum on full full sustain level

User avatar
kco_no
Information
Posts: 3
Joined: 26 Apr 2016, 12:38
Has thanked: 1 time

Post by kco_no »

Thank You for your reply. I'm actually from Italy and i've never seen that kind of notation before.

So while turning right the SUSTAIN potentiometer, it'll also reduce the input level. I was wondering about having separate potentiometers for INPUT and SUSTAIN.
Is that in your opinion a good/bad idea?

Anyway thanks for your design. I'll post a picture of my build as soon as it's ready :)

User avatar
Gobi
Information
Posts: 16
Joined: 06 Feb 2015, 22:35
Has thanked: 3 times
Been thanked: 5 times

Post by Gobi »

Remeber: this is just my personal modifikation! The original ciruit is from mictester!
In fact, the changes are not very big - so you cuold try both.

For me, it works perfect now - in chase one has a very clean and quiet 12V! Otherwhise you will get a lot of noise

User avatar
phatt
Transistor Tuner
Information
Posts: 1224
Joined: 04 Aug 2010, 05:17
Location: Morayfield SE Qld AU
Has thanked: 8 times
Been thanked: 343 times

Post by phatt »

Gobi wrote:I really feel ashamed.... AGAIN my sketch above contains a mistake. Would please someone of the admins delete them... dont know, why I could'nt do it.

I found the mistake, when I tryed to rebuild it - now it should be right:
[ Image ]
(there is now need fot the red LED)
hope this will be the last time... sorry again


The noise issue is not due to a bad power supply,, your gain is off the scale.

The circuit is dependent on the ratio of the *Series* resistor (Your 250k pot and 10k resistor) at pin2
And the 2m2 resistor from pin2 to pin3. that ratio sets the gain.
with 250k pot at zero the series resistance is then 10k which dramatically increases the gain. [smilie=a_holycrap.gif]
The other 250k pot (Dual pot) then switches on the Leds harder and so a balance if found.

BUT with no signal the gain will rise up to a much higher value and noise will be the result. :twisted:
That circuit may even become unstable and squeal.
It's not something I would build as it's highly noise prone.
I have built this circuit so I'm very aware of how it works.

With that series resistor at 10k you also make the input Z very low and than may have a negative result on the signal
Can I also suggest removing the extra indicator LED as it's likely of little use and may interfere with the compression.

Phil.

User avatar
Gobi
Information
Posts: 16
Joined: 06 Feb 2015, 22:35
Has thanked: 3 times
Been thanked: 5 times

Post by Gobi »

Thanks Phil for your suggests.
The indicator LED is useless, thats right - it's a left over from my older circuit. Because of the high gain you could hear noise of the power - I added something like this:
Image
and it's much better now.

Belive me, it works fine - the only problem left over ist a little crackle at low notes witch i didn't realised at first time- I gues you are right with the 10K resistor at the input. I will change it next time....

(by the way: it´s important to run the circuit with 12V - it works also with 9V, but very different - amzingly it's louder than - seems to be because the LEDs shine less)

User avatar
phatt
Transistor Tuner
Information
Posts: 1224
Joined: 04 Aug 2010, 05:17
Location: Morayfield SE Qld AU
Has thanked: 8 times
Been thanked: 343 times

Post by phatt »

All good,, Yes that filter helps but just the same I've found it's better to remove potential issues by backing off the gain.
Agree this circuit works much better at higher voltage,, mine runs from 20Volts and dead silent.

That distortion you mention is (For me at least) the whole magic of this compressor if you do your research that is as close to the classic Valve amp rattle I've heard.
ANY old dirt pedal can do distortion that's the easy bit,,,but few can do the magical triode rattle. listen to some old Shadows tracks,, clean?
Well not really there is quite a bit of edge on some notes.
The untrained ear would assume a clean guitar but it's got that little edge reminiscent of some famous valve amps. try some Rolling Stones songs it's more obvious then,, less clean.

Now Add in the bonus of compression you have as close as one can get from one chip to the classic Valve era. Not perfect but who cares costs less to build than a new set of strings. :D

So yeah for such a simple basic circuit it just leaves all the other Comp pedals sounding like crap,, super clean is way way wrong for good R&R guitar.

OH yeah,, If that low distortion is too strong try lowering the value of the input cap,, I think mictester mentioned this point a few pages back.

Phil.

User avatar
kco_no
Information
Posts: 3
Joined: 26 Apr 2016, 12:38
Has thanked: 1 time

Post by kco_no »

I was wondering: shouldn't that 1uF capacitor be reversed? The one near the trimmer

User avatar
Gobi
Information
Posts: 16
Joined: 06 Feb 2015, 22:35
Has thanked: 3 times
Been thanked: 5 times

Post by Gobi »

:oops: you are right..... a mistake again.... should be more concentrated , sorry

User avatar
uki
Information
Posts: 1
Joined: 20 Aug 2015, 23:25

Post by uki »

Hey thanks for this !!

I just build one and it is really good ! "Is it on?", but miss it when it's off! That is exactly how it is !
Attachments
Compressor_.jpg

Post Reply