Nurse Quaky very IC sensitive?

[hilbi]

Yesterday i tried some different ic's in my Nurse Quaky and i experinced that ic's make a lot of difference in this circuit.

First i had RC4558N in it, wich was popping and very noisy but the wah sound was ok, then i tried a TL072 wich had very low volume (50% of unity gain) and hardly any wah effect, then i tried an old TS9 IC (7558 or so) wich was unity gain and a good wah sound, so i kept that one.

Strange that the TL072, original recommended had such a low output.

Anyone experinced this as well with the nurse quaky?

[B Tremblay]

Well, I did.

Even though the Nurse incorporates the biasing componentry from Jack Orman's Dr. Quack, I found that a 1458 or LM358 was needed for pleasing results.

A TL072 did not work for me and the suggestion to use one has been removed from the schematic.

[hilbi]

Thanks B Tremblay.

I used Torchy's vero layout wich is recommending a TL072, i just checked your site and saw that you suggested different opamps. A TL072 is indeed not workable.

Unfortunately i don't have those 14558 and lm358, but i will order them. Meanwhile i use the reissue ts9 opamp, wich is workable.

Just did a search on the forum (sorry should have done that first) and saw that there are more topics about the opamp sensivity of the nurse.

[Zero the hero]

as long as I know, some envelope filters are veri op-amp sensitive.
I had troubles with a DOD440 (the sound was very distorted) and I solved the problem using a LM358 (I was using a 4558).
A friend of mine ad the same issue with a Dr. Quack, and he sonved this problem in the same way I did (he originally used a TL072).
Does LM358 have an higher output voltage swing?

[petemoore]

Yes, and that transistor, the NPN...I socketted that, it can make NQ wayy different, totally revoiced.
 Because it varies the sweep AloT, trying different Q's 'there' [I hafta go look...it's the 'sweep Q'], it can really tune NQ. Even near same Type/Hfe Q's, when swapped there make a fairly profound, or at least noticable difference.
 Going to different gain Q totally changes the freq range/sweep.

[B Tremblay]

Pete's right on the money about the transistor.  I used a grab-bag NPN for a long time, then finally put in a true 2N3904 and was startled by the difference.  A socket is highly recommended for it.

[Mark Hammer]

For me, the question that crops up is whether there is something that can be done with respect to the trimpot or series resistance prior to that transistor?

Incidentally, the old Seamoon Funk Machine (Which I do believe RG drew up and posted at GEOFEX) used a similar type of structure as the Doctor Q, but a different type of bandpass filter, also uses a single NPN as control element to ground.  You will note that this circuit does NOT use any trimpots as the DQ/NQ does, but uses a pair of fixed resistors.

[petemoore]

One thing's for certain, the NQ is ripe for a mod that allows her to get extremely funky or more of a normal nursing type deal.
 Could be as easy as getting it to go nearer one of the extreme settings by choosing a transistor that does that there, then add a cap [B/C?] or gain presetting resistor to change the Q's output, then put that on a switch...
 All I did was plug in different Q's and tune it from there, obviously a different Q could be switched to as alternate, but I suspect there's a way to do it with a smaller switch and fewer nodes.

[hilbi]

mmm the old ts9 opamp (7558 or so) now has the same problems as the rc4558N, first it sounded ok, but yesterday i tried it again and the same troubles arise.

I will get me a 1458 or lm358 asap.

[hilbi]

I replaced the opamp with a lm358.
Replaced the transistor.
Replaced the led's

Still i get distortion and popping sounds especially when i minimise (turn left) the attack (fastest attack) and sensivity pots (maximum sensivity).

Maybe something else is wrong with my quaky?

[onboard]

Hey Mark, would the distortion be a result of the CV from the envelope follower getting into the audio path?

I inserted a gain stage between the input and the envelope follower in order to dial in just about any instrument regardless of Q or IC - works pretty good.

[hilbi]

maybe lowering te 3,3M resistor to stock dr Q value (2,2M) would help prevent the popping and distortion?

[R.G.]

Dr Q, Nurse Quacky, the Seamoon Funk Machine, the Pulsar tremolo, the EA tremolo, the Orange Squeezer and most of the other pedals I can think of that do any kind of sidechain processing in a simple circuit are all touchy and component sensitive.

That is because, whether through clever design or happy accident, the processing and/or sidechain relies on the quirks and foibles of one or more special components that happens to have a property that does what that particular circuit needs.

From a designer's viewpoint, there are two competing ways to get a special circuit function. You can either do the brute force, nail-it-down kind of design where you use lots of parts and synthesize each part of the circuit to reliably do that particular function even in the face of component variation... or... you can pick up a single device whose eccentricities happen to be just the set of eccentricities you need.  The eccentric-component  approach could be looked at as very, very clever design where you have found a custom piece of silicon (or germanium!) that does exactly what you need in only one chip, a kind of custom IC if you will, that does whta you need. Well, dumb luck also gets in there, but if you have to choose skill or luck, choose luck...

That's why you have simple compressors that work OK when you get just the right parts and are difficult to sub into, why you get envelope followers that only work if you get a real, no-fooling LM1458 opamp, selected JFET and so on.

Or you can start with the simple design and design the dependencies out of the circuit: find what special quirks of the opamp are needed, and why and put in more parts to make an ordinary opamp work; find out why that JFET is special and put in a trimmer to let you adjust in others; find out why that NPN bipolar works and put in diode clamps and trimmers to make more ordinary NPNs work. Find out why low gain devices only work and figure out how to piggyback another device to lower the gain, and on and on and on.

Commercial stuff (other than bottom dollar effects) will always go for the complex but reliable route with synthesized functions that are slam-dunk solutions, or special-purpose ICs that do the same thing, but first time, every time.

[Paul Perry (Frostwave)]

Boscorelli's "Stomp Box Cookbook" designs are prime examples of what RG calls "nail it down design".

[Brian Marshall]

when i was working on the money shot, i breadbaorded a dr quack just out of curiosity.  4558/4559 had very low volume at higher end of the filter effect.  probably due to input impedance.  TL072 was better, but didnt sound great.  i tried a TL022 which sounded better, but still didnt sound right.

i built a couple dod 440's with TL022's i actaully didnt like them at all.

[hilbi]

yesterday i finally put a 1458 in my quaky, but still i get same pops and noise.

anyone have an idea what else can be wrong?

[Mark Hammer]

Okay.  Rereading this, there are a few things that have not been verified.  Or at least no mention is made of them having been verified.

The "compensation circuit" that Jack Orman came up with for the Dr. Quack, and which is used on the Nurse Quacky uses a pair of LEDs.  If the pedal is wired correctly, what you should see is that one of the LEDs is lit up constantly, when power is applied.  The other should flash briefly when you strum.  You should probably be able to measure somewhere in the vicinity of 500-1000mv of AC signal on the transistor side of the second LED. If you do not observe any of this, then there is a good chance your LED is either oriented wrong, or burnt out.

The long and the short of it is that we have no confirmation that the envelope follower stage is appropriately biased at this point, or that audio signal going in is producing envelope signal coming out.  Once that is addressed, THEN we'll know what to make of op-amp choice.

[hilbi]

Hi Mark, Thanks for your help.

The leds are working fine, one lit up constantly and one depending on the strumming.

The enveloppe signal is also working fine, I do get an auto wah sound, all the controls work also.

But I still get a bit distortion and popping sounds especially when i minimise (turn left) the attack (fastest attack) and sensivity pots (maximum sensivity).

[Mark Hammer]

Well, then, we may well share the same problem.  I'm currently wrestling with a version of the Quack circuit to try and get a more vowel-like sound, using dual-filters with slightly different Q, attack/decay parameters, and different filter range.  My LEDs light up fine too, with a raspy gargling sound from the filter sections themselves, none of which responds to trimpot tweaks, or change of chip.  I'm beginning to wonder if the biasing of the op-amp "+" input may have something to do with it.

[hilbi]

Yesterday I replaced the 3M3 resistor with a 2M2, this didn't really change the popping and noise.

I can dial out the popping, just raising the attack pot a bit. But the background distortion with sustaining notes cannot be dialed out.