Compression Sustainer

[tv]

You don't need a quad opamp (074) to do a fullwave detector comp. It is easily doable with a 072, second opamp being used as inverter. I -made- and used such fullwave det. things ... some 20yr ago..

In your present config/schem, all the second opamp in 072 does is sitting there being a "virtual ground" Vcc/2 generator, which is not needed in a simple circuit like that. Except that it will eat batteries twice as fast compared to what would a single 071 eat. Namelly, a single 071 is all what this circuit really -needs- ...

For a comp that is plugged into a guitar amp, a fullwave detector is absolutelly not needed.

For a bass instrument (DI'ed) or bus compressing, that's another story, but as I said, I found it passable on bass if a halfwave det. was used -but- done in-phase..

Other thing I found on bass was that the distortion-cancelation on FET killed the vibe, so I removed it on the unit that I used. I think I still have some small proto PCBs somewhere. The "final units" "evaporated"..

 

[livingnote]

You mean a distortion cancellation similar to the one used on the SSL "talkback" compressor?

I was thinking about the extra opamp being not too bad an idea, if a little overblown like you
say, but doesn't it at least help fine-tuning-wise to go ahead and use it? (Though the battery
thing makes major sense but I'm not a battery nut myself, wanting to go with 15V and relays
to boot). What I do have here is a pack of 10 AA batteries that has a 9V clip on the end, that
lasts pretty much for ever on one of these...I learned that one doing my old guitar mask...

It had a little 20W amp, speaker and MP3 player on it, which was what really
got me back into electronics in the beginning. 8AA batteries, and it lasted forever...

 

[tv]

The two 2M2 (iirc) resistors that are connected D-G and G-to-time constant. If you either omit the D-G resistor or make the other (going to G) variable (f.e. a ), you can remove or influence distortion cancellation mechanism (i.e. make it variable). This you could accomplish by using f.e. (a) a 2Mohm pot in series with 220K resistor or (b) a 1Mohm pot in serise with a 1M2 resistor (you get the idea, mix-n-match to get the most "feels good under fingers" combo of values).

If current consumption is not a problem, then having the second opamp as a "virtual ground" generator could help in case if the circuit is powered by a not-so-good walwart - resulting in slightly better psrr. However, a drawback is (per my observations) an annoying power-up opamp transient that happens when you power-on the device and it gets injected into the audio-path via biasing resistor for the "audio" opamp..

But in this or similar circuits, I don't think that this has any -other- benefit at all (except if the makers of original device bulk ordered TL072s and it was necessary to "somehow use" the other half). I also did that in preamps that I had in my "work" basses, used duals at first because I had a pile of these - but later did it "proper" with single because with first gen pres the battery life wasnt really good..

If you have a strong-enough battery, what I said doesn't matter at all. Hope it helped some, though.

 

[livingnote]

Interesting, I was wondering where that sharp transient came from. It almost makes me want to
implement the relay as being time-delay to avoid it.

I saw the thing with using an opamp to stabilize DC voltage first-off visiting the page of an ex-Scientologist
who took great pleasure in explaining why their e-meters suck from A to Z, and continued to design
one that works waaay better, making PCBs and selling it to DIY. Dianetics is just basically Jungian stuff etc.
repackaged to look like L.Ron invented it (which is like saying Ray Kroc invented McDonald's), but it was
interesting to know that human body resistance can vary from a few hundred ohms to 100k ohms within
about 30 seconds or less. This is inexplicable through "sweaty palm" galvanics, as a human's bodily DC
resistance will tend to actually increase when under stress, when it should be decreasing.

i.e. - Psychic resistance to a key thought translates to physical ohmic resistance. Crazy.

It is also the root cause of some people dying on sight when they get electrocuted, and others hanging in
the line for a while and coming out yelling "sonofa!"...has little to do with how dry your skin is.

So he built this thing that had to be 10'000% accurate even in the tiny ranges, and temperature-compensated,
free from long-term drift, and fast as in "true read". - it's chock full of constructions like that.

 

[okgb]

The Girl, How much for the girl ?

 

[tv]

Using an "virtual ground" opamp (or some other circuit doing the same) is required when you -need- a good, low-impedance ground for reasons -other- than just biasing. I mean, there -are- cases when you need such circuit.

But using a relay for a turn-on transient suppression is a bit much (for Christ sake). I'ts like using a hammer when all you need is a toothpick. The "simplicity itself" attached below will get rid of some of that transient, by the fact that this is a RC time constant circuit and when a device is first powered, the positive rail will gradually go from zero to Vcc, thus giving less chance for the transient to occur at the output.

What you need to figure out is first the value of the R - according to the current consumption of your circuit, there will be a voltage drop across that resistor. Figure out the smallest value that works satisfactory.

You can also split the R into two resistors - IOW, you could cascade two RC cells and f.e. use 2 x 22 Ohm + 2 x 100uF...

KISS, cheap, Just Works, survives accidental AC ...

The only thing you could add is a Zener diode (a 15-16V one) to put things "under control" for good...

 

[livingnote]

Ah of course - tame the lion FIRST, and you don't need the big exoskeletal corset anymore

Makes sense - CRC, what else. And standard Lukas stompbox behaviour when not copying other
people's work - of course the ramp-up through the current-lim resistor makes perfect sense, without
sacrificing on-time delivery of electrons when the circuit need to shovel some kinda transient.

It's even dead easy to mod that way with the layout topology as it is.

Ah Greg - yeah, the girl is finished with her education and I think she's back in Switzerland now...
guess that would screw prices up something rotten...specifically once you get past the "entry price"...

But you know...where there's love, who needs money?

Well ok yeah as long as you're swimming...

 

[tv]

Acually, it is a RC-RC, not a CRC..

It behaves like a sort of ... doubled time-constant or a time-delay circuit. For simple circuits like that (with low-ish current draw), you can get-by only with using passive components, which is cheap enough and Works. This is also recommend with low-voltage circuits, because you can make such CR-CR..CR circuits with a low voltage-drop (i.e. using low-value resistors and elcos).

For larger current draw, you can use a capacitance multiplier, which will also exhibit "soft-start" behaviour, but will have larger forward voltage drop (and some other not-so-good behaviour) so it's not as suitable for use with battery powered devices. But, such circuits are cheap, small, and if you are clever, you can cram some additional fuctionality into it..

--------
OT, but since you asked for photos (which I dont have btw.) - I recorded some loops thru a prototype built on your board. If you are curious, I can send it via email.. but it's not "for public".

 

[livingnote]

Ah - right, RC with the R first. It always reminds me of a fountain and a bucket.

Capacitance multiplier? Ok gotta go google that one.

And sure - really curious to see what does that stuff that ends up on my PCBs

 

[tv]

So that you know - I finally took a look in my "box of oldies" and dug out two compressor boards that didn't get trashed.

One is missing half the components, but the other just missed the input cap (however I killed the IC - but as a consequence now it has a socket). This board also has Roederstein RDE electrolytics, mind you.

One thing that stood out is how similar the time constants values in GR circuit are to the "mictester's" circuit - one board has the same values (2K2 attack into 2.2uF tantalum and 1Meg release, the other has 1uF tantalum, for whatever reasons I did that .. 20yr ago). The shorter was optimized for bass.

What I meant to say is that there must certainly be something "going" on with this attack/release values. "Sweet spot"?

 

[livingnote]

Now I got clearer info on the caps I was using here - the one cap that really rocked on the input
was some generic MKT cap, really nice. What I'm gonna try is to build the whole thing out of MKTs
now and compare it back to the MKT+MKS2 version, but the guy over at Segor was talking a clear
cut himself re MKT for audio stuff.

 

[tv]

Hey, Lukas, is this thing still alive?

One thing to remember is that "guitar electronics" isn't "audio" (while technically of course it is audio)...

Nothing beats personal experience, though, and personal tonal/hearing preferences come into play -much-stronger- when one listens to himself while playing an instrument. Little things you wouldn't even notice when listening to playback can make much difference. So called "response".

For me, the small MKTs "just work". Slightly more nasal and thinner than their more expensive brethren, but this isn't exactly a bad thing for guitar.. ymmv. Or perhaps I'm just too set in my ways and too lazy to experiment with cap swapping anymore.

 

[livingnote]

Yeah - been twiddling around with it now for a while and possibly the MKS's have their virtues as
coupling caps but that MKT on the input just does this "clarity" and "crispness" thing that triggers
the "wow what opamp is that" question...pretty cool considering it's so cheap. Could of course also
just be the way it works as an LC with my concrete guitar's pickup but I guess there's something
to the guy's opinion.

That with the attack cap is funny, I used, out of necessity, a 3.3µ, gonna have to compare that
to a 2.2 next time around...

 

[okgb]

The boards , where are the boards ?

 

[tv]

For circa 15yrs I swapped caps fiercely and settled for the cheap ones...
The "nasal focus" is where it's at (imo)..

 

[livingnote]

The boards , where are the boards ?

Now that I can ship them with the caps that rock, they're coming