723 voltage regulator IC makes a distortion.

[anotherjim]

The uA723 or LM723 voltage regulator is ancient. Born in 1967, it should be dripping in mojo. It is still in production.
http://www.ti.com/product/LM723/technicaldocuments

But it's a voltage regulator, and one that needs extra bits and pieces. What use in a guitar effect?

Talk of ancient op-amps got me thinking - hang on, there's one in the 723 & that is ancient too. How could that sound?


The internal arrangement. Note those pin numbers are for the 10pin metal can package, not the 14pin DIP LM723C that I'm using.

The equivalent circuit it also useful to see. It shows the output has a bit more to it. Pin numbers match the 14pin.


Actually, I don't know if the Error Amplifier qualifies as an op-amp. Maybe it's just a differential amp, but it seems to work if treated as an op-amp.
There are some transistors hanging on the output. What if their PN junctions could be used as feedback clip diodes, Tube Screamer fashion?

That's what the circuit on my breadboard at the moment basically is. Either a base-emitter or base-collector junction of the transistors is used for the clip diodes. Input bias comes from the chips Vref pin for no better reason than because it's there.

I don't think it's quite finished. I haven't built it up or recorded any demo tracks yet.

Capacitors labelled TBC are "to be confirmed". They are not currently present, but one or both will be handy if its thought that some treble needs cutting back.
C5 is currently 10uF.

Build it. Feel the hidden mojo!

[Marcos - Munky]

Very interesting! Waiting to hear a few soundclips when you record them.

[PRR]

> if the Error Amplifier qualifies as an op-amp

Objection that an "operational amplifier" is usually understood to have an output range both-ways from ground. Or some half-voltage. Here it can only (cleanly) come down to the pin 4 input voltage (a wee bit lower (lower yet using the Zener)).

But yeah, for audio purposes it is a lame op-amp. You even get use of both inputs. You will have to think how to offset your DC feedback. You will want Compensation cap. I dunno how happy it is at 9V, but this isn't about the chip's happiness, is it?

[anotherjim]

As it is, it can sound like a cranked amp channel.

There does not appear to be any instability, although I have added a 100p feedback cap at C8.
Output wave reaches a hard clipped flat bottom, but positive peak remains rounded when limiting. Might explain the "amp-ish" sound.

Gain pot is scratchy. DC offset current is causing that. Think I'll try alternative gain setting of a fixed feedback R and variable input R, since input bias current can't go that way.

Should add that the "Diodes" switch is currently just a breadboard option and really a link that I'm moving. I'm thinking I prefer the pin11 option, but maybe because it seems a bit louder.

Not sure if I should believe it, but current consumption is just under 1mA. I don't think I've ever made a pedal circuit that frugal.

[pinkjimiphoton]

that looks really cool jim, can't wait to hear it!

[anotherjim]

Worked up a very simple version...

Cut down to minimum/basic/common parts.

Turns out that the amp is pretty basic and can't swing very far towards ground (as Paul points out above). Maybe just a transistor BE drop below reference.
I suspected the clamp using pin 2 wasn't doing anything - it wasn't. A diode with a lower Vf can be seen to clamp the gain, but I don't feel it's worth it.
The "simple" above uses the 7v chip reference to directly bias the input and offers a switched choice. A diode clamp for a more compressed effect, or no diode for basic overdrive clipping of the amp and/or whatever the thing is driving.
Sound of this one is very much in hard driven amp rather than hi-gain territory.

Edit to mention something I noticed...
If the feedback R10 is removed while the clip diode is selected (pin11 connected to pin 4), then you get a weird broken-amplifier/Sitar sound that someone out there might like. It won't work without the clip diode connected.

[anotherjim]

Quite a while later, after much cork sniffing, a build has been settled on and done. Speed of development not helped by the fact that I enjoyed the sound so much, I spent more time playing through it than I did experimenting with the circuit.
Here is the scheme...

Included a mill' bypass, because after finally purchasing some 2n7000, I'm quite taken with this idea. Thanks RG!

Demo tracks will arrive shortly now that I've got the real deal to play with.

A bad pot was troubling the original breadboard version. That realised, I've gone back to variable feedback. I like this better because as the R is increased, treble cut from the feedback cap C6 also increases.

Unlike a proper op-amp, there isn't that much gain. With a 470k pot (log taper is best) the higher 20% of travel doesn't make it much hotter, but the treble cut continues to deepen - so it's a combined drive/tone control in a way.

The "Clip" is subtle, adding some fuzziness. It only affects the positive going signal. Negative going clipping happens anyway with high gain/strong playing.
The error amplifier is not an op-amp as we know it.

Circuit is small enough to make the board a "pot hanger". My usual perfboard method of putting the chip in the middle and soldering stuff in where I can in accordance with the schematic. May get a photo or two up, but I don't have a layout graphic.

Until recordings enlighten, it can be a boost or a crunch overdrive. Clip adds some fuzz onto the crunch.

[pinkjimiphoton]

making popcorn