Made the "really cheap compressor", help debugging fuzz?

[deadlyshart]

Hi guys, I made the "Really cheap compressor (as seen on this page: http://tagboardeffects.blogspot.com/2012/03/really-cheap-compressor.html ) after several people suggested it. Here is a schematic of it: http://www.diystompboxes.com/smfforum/index.php?topic=110831.0

(I drew out the schematic from the vero board just to be sure, and it's identical.)

I assembled it, made my little LED/LDR thing, and I'm pretty sure it works. The signal definitely goes through and gets amplified if I turn up the Level pot. It's a little hard to tell how much it's actually compressing it, but I think it is.

However the most pressing thing is that I'm getting this annoying fuzz if the level pot is anything more than the lowest value. It's not terrible, but it's noticeable and unpleasant.

I have a cheapo kit scope I made that works decently, as well as a kit function generator. Since my unamped guitar seemed to output a ~100mV signal, I used a 100mV amplitude square wave from the Fn gen and put that into the input of the compressor. I used a square wave so I'd be able to see if it was getting distorted, because higher harmonics were getting messed up.

I tried several frequencies in the audible: 100Hz, 1kHz, 10kHz. The sustain knob seemed to not affect any of them at any level, but this is expected I think, because it's just a constant input (not changing amplitude really). Here's what I noticed: the 100Hz wave seems to get the most screwed up, and following the signal path, it actually appears to happen right after it goes through that first 15nF capacitor.

But this actually makes some sense, right? That 15nF input cap is basically a high pass filter, and 100Hz might be low freq enough to get partially killed by it, leaving just its harmonics.

Here are a few pics. Apologies for the crappy quality, it's a phone pic of an LCD screen...

100Hz output right after input cap:



1kHz output at final output jack:



10kHz final output. Some distortion with level pot high:



Still, I'm not sure how much this tells me. I know it's difficult to diagnose something without hearing it probably... Is there anything else I can try? thank you!

[midwayfair]

The level pot's just an output volume. Its only connection is the 2u2 hanging off pin 1 of the IC. So if you've got distortion that changes when you turn that pot, you've maybe either hooked up something wrong or you're distorting the next device in line (i.e., the amp). The output can be pretty high, especially if it's not compressing much.

Some other things:

This design uses an inverting op amp input. These are, for whatever reason, more prone to distorting than the non-inverting side. The RCC distorts and I'm pretty sure that, besides the kind of high gain, this is one cause.

The input impedance drops when the LDR's resistance is low. In fact, it might not be much higher than 220K if you drive it hard enough. 15nF is sufficient to pass all audible frequencies, so you shouldn't see any appreciable phase shift on 100Hz from the input RC filter. There might be another cause but I'm not sure what. Anyway, if you were using a guitar, that lower impedance might load the pickups. A signal generator, though, nope.

[deadlyshart]

Thank you, I'll think about this!

Dumb question, but how often is fuzz/etc just due to messy board assembly? I don't think mine is god awful, but it's probably not great.

Ooh, sorry, one thing I forgot to mention: the board I posted above suggests a NE5532 op amp, but I actually used a TL072 (CP) because that's all I had. Sorry, I'm not familiar with RCC...what is that?

[EBK]

Thank you, I'll think about this!

Dumb question, but how often is fuzz/etc just due to messy board assembly? I don't think mine is god awful, but it's probably not great.

Ooh, sorry, one thing I forgot to mention: the board I posted above suggests a NE5532 op amp, but I actually used a TL072 (CP) because that's all I had. Sorry, I'm not familiar with RCC...what is that?

RCC = Really Cheap Compressor.  Turns out you are familiar with it. 

[deadlyshart]

Derp, I'm an idiot. Thanks.

[Ben Lyman]

Is it a distorted guitar sound or more of a fuzzy background noise along with a "regular" guitar sound? Sorry I don't speak O-Scopish but maybe it's something like envelope ripple?

I had a sound that I would describe as a very tiny man with a very, very tiny kazoo living inside the TL072. He would play quietly along with my guitar and then fade out when I stopped playing.

So, I put a 100nF ceramic cap right across pins 4 and 8 (-9v/+9v) and I mean quite literally on the op-amp, soldered right to it.

It sucked the little guy and his kazoo right into the cap and now he can jam along with me to his heart's content but I don't have to hear him

[EBK]

Question about your fuzzy sound: Does it go away when you play a loud note?  Compressors decrease dynamic range, in part, by making your quiet sounds louder.  Without a note being played, many compressors are quite happy to amplify your background noise. 

[PRR]

If your LDR is working, there is No Way this thing is distorting.

LDR usually goes down to 2K Ohms. We have 220K input resistor. So the gain can go down to 1/100.... output far smaller than input.

[Groovenut]

In my experience with this circuit, distortion comes from 2 places (if the LDR is functioning properly). Conduction of the LEDs and overshoot of the signal transient. The LED issue can be solved by placing a 10nF cap in parallel with the LEDs. The overshoot issue is most often caused, in my experience, with an LDR that has a very high dark resistance or a slower than "normal" reaction time. You can try substituting another LDR that measures ~ 1M dark or possibly using a smaller parallel resistor than the 2M2. The parallel resistor will of course lower your overall gain potential but you may not need all that gain to get the circuit to behave like you wish.

Also from what you said in your first post, I don't think your LDR is working, hence your just getting gain and clipping the rails of the opamp. This circuit has very noticeable compression when functioning properly.

My 2 cents

[deadlyshart]

Is it a distorted guitar sound or more of a fuzzy background noise along with a "regular" guitar sound? Sorry I don't speak O-Scopish but maybe it's something like envelope ripple?

I had a sound that I would describe as a very tiny man with a very, very tiny kazoo living inside the TL072. He would play quietly along with my guitar and then fade out when I stopped playing.

So, I put a 100nF ceramic cap right across pins 4 and 8 (-9v/+9v) and I mean quite literally on the op-amp, soldered right to it.

It sucked the little guy and his kazoo right into the cap and now he can jam along with me to his heart's content but I don't have to hear him

Ohh, sorry, I should have been more clear: the guitar sound itself doesn't seem very distorted, it's just that it has this fuzz when it's not playing. When it's playing it seems to actually sound pretty clean.

Hmmm, I have a cap across the rails, but I guess sometimes the location of it is important too?

[deadlyshart]

Hey guys, some more info--

I think my DIY optocoupler is working, mostly at least. I unsoldered it from my circuit to test it (...as I probably should've done before). It's just two back to back green LEDs, in several layers of heat shrink tubing, with a LDR across from it. In ambient, it's about 1Mohm, but if I cover it even slightly from the ambient light it goes up to about 2Mohm.

I applied 5V across the LEDs side of it in series with a 200 Ohm resistor. This put about 2.4V across the LED pair. There actually seems to be some asymmetry, because with it biased one way, it has 10mA running through it, but 13mA the other way. Similarly, the LDR resistance drops to 2kOhm for one bias (the same 10mA I think), but ~400Ohm for the other bias.

So it's definitely working, but is the asymmetry bad? I feel like I messed up enough stuff that it might be worth rebuilding the board.

[duck_arse]

the asymmetry is not as bad as all that current you're running thru the leds. as long as you have two leds, and they light and the ldr changes resistance with the lighting, you opto parts are good to go.

I had a talk-radio station instead of a kazooo man, came in as the notes decayed - I found I kept leaning in trying to hear what was being said all the time - that's the path to madness. in my case, it coinsided with the opto driving opamp coming out of signal clipping. and my "cure" was to put the ldr right close against the opamp pins, with the leds as far away from the opamp as they'd "go".

is your board soldered or breadboard? can we see pics of it, either way?

[EBK]

So it's definitely working, but is the asymmetry bad? I feel like I messed up enough stuff that it might be worth rebuilding the board.

This thought springs to mind:

Jim Jarmusch once told me "Fast, Cheap, and Good...pick two. If it's fast and cheap it won't be good. If it's cheap and good it won't be fast. If it's fast and good it won't be cheap." Fast, cheap, and good...pick (2) words to live by.

Keep at it, and eventually your efforts may pay off.

[deadlyshart]

Question about your fuzzy sound: Does it go away when you play a loud note?  Compressors decrease dynamic range, in part, by making your quiet sounds louder.  Without a note being played, many compressors are quite happy to amplify your background noise.

That's a good point... it's hard for me to tell, but I don't think it's going away when I play, but it's definitely not really noticeable while playing, because it's so quiet compared to the guitar notes... maybe I'm complaining about nothing? I can't tell...

It's hard for me to tell if it's doing anything, on the other hand. I had my friend turn up the sustain knob while I played the same thing repeatedly and it was difficult to say if anything was happening. I'll bring it to a jam today and see if I notice it/its absence. I might just have crappy ears and its effect would be obvious to anyone who knows.

[deadlyshart]

the asymmetry is not as bad as all that current you're running thru the leds. as long as you have two leds, and they light and the ldr changes resistance with the lighting, you opto parts are good to go.

I had a talk-radio station instead of a kazooo man, came in as the notes decayed - I found I kept leaning in trying to hear what was being said all the time - that's the path to madness. in my case, it coinsided with the opto driving opamp coming out of signal clipping. and my "cure" was to put the ldr right close against the opamp pins, with the leds as far away from the opamp as they'd "go".

is your board soldered or breadboard? can we see pics of it, either way?

Sorry, I don't understand what you mean by "all that current running through the LEDs"... Do you mean the 10mA/13mA I quoted above? Just to be totally clear, that was when I took the optocoupler out of the circuit and tested it independently, and 13mA through an LED shouldn't be bad, right?

I'm actually getting some local radio when I crank up the level past half or so, now. So you think the reason was that the LDR was physically farther from the op amp pins in the circuit? Why would that cause the radio pickup?

thanks for the advice, I'll look into it. I'll get some pics up here, though you guys will most certainly shudder at them.

[EBK]

I'll get some pics up here, though you guys will most certainly shudder at them.

We promise to shudder politely and thoughtfully. 

[deadlyshart]

Well, you guys asked for it...

Front and back:







A few notes--

You can see that several times the damn veroboard sockets got ripped out, so I had to redo those. I tested the veroboard cuts after I made them, to make sure there was no conductivity between the cuts.

I'm doing that method where the pots are connected directly to the board, so that's how the board is held in the enclosure. It seems to be holding well and the connections for the pots to the board aren't cracking, but if the board was any better it probably wouldn't be a good idea...

You can see that I put the opto coupler on the same side as the solders (opposite side from the rest of the components), this was because I ran out of space on that side. Should've used a bigger board...

I'm also using detachable header pin wires for the inputs and outputs. This has made it a lot easier during the million times I've disassembled it, but I notice that in gut shots of pedals I've seen, almost no one else does... is it bad to do this?

Also, I just noticed, in the veroboard schematic I followed ( http://tagboardeffects.blogspot.com/2012/03/really-cheap-compressor.html ), two of the non-electrolytic caps are maroon, and one is orange. Are the maroon ones supposed to be mylar or tantalum or something? Because, if you look, I used ceramic disk caps for all three of those... is that bad?

[EBK]

You can see that several times the damn veroboard sockets got ripped out, so I had to redo those.

I'll make one more public service announcement encouraging people to stop using phenolic resin boards, mainly for this reason. The epoxy resin boards cost a bit more and are a tad harder to cut, but they are well worth it.

[deadlyshart]

You can see that several times the damn veroboard sockets got ripped out, so I had to redo those.

I'll make one more public service announcement encouraging people to stop using phenolic resin boards, mainly for this reason. The epoxy resin boards cost a bit more and are a tad harder to cut, but they are well worth it.

Ah, didn't know that. It was this one: https://www.amazon.com/gp/product/B019Q0ZTJ6/ref=oh_aui_detailpage_o02_s00?ie=UTF8&psc=1

But for my next project I got this: https://www.amazon.com/gp/product/B00C9NXP94/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1

which is epoxy, right?

(it says "Single-sided high-quality FR4 glass-epoxy circuit board")

[EBK]

You can see that several times the damn veroboard sockets got ripped out, so I had to redo those.

I'll make one more public service announcement encouraging people to stop using phenolic resin boards, mainly for this reason. The epoxy resin boards cost a bit more and are a tad harder to cut, but they are well worth it.

Ah, didn't know that. It was this one: https://www.amazon.com/gp/product/B019Q0ZTJ6/ref=oh_aui_detailpage_o02_s00?ie=UTF8&psc=1

But for my next project I got this: https://www.amazon.com/gp/product/B00C9NXP94/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1

which is epoxy, right?

(it says "Single-sided high-quality FR4 glass-epoxy circuit board")

Yes, that new one looks fine.  Epoxy resin boards are usually green, sometimes yellow.  That particular shade of brown on the board you are using is the telltale sign of phenolic resin (but they were sneaky with that green mask they used on the pad side).

Edit: I just noticed the irony of my suggestion that you build your Really Cheap Compressor using more expensive parts.