Orange Squeezer, silicon diode

[heisenberg]

Is there a way to alter the circuit to use a silicon diode 1N914 instead of a germanium? I can't get germanium at radio shack and don't have access to germanium for a while. I can always switch out the TL072/82 for the 4558.

I saw clones or re-issues on the net that used silicon diodes instead of germanium. And the schem i'm going from is from generalguitargadgets.com. It says to use silicon i'd have to alter the circuit.

[aziltz]

You could lower the value of the resistor (1.5k?) its in series with to compensate for the higher turn on voltage of Si.  Or socket it and just use Si for now...  I have it on breadboard (I've been tweaking) and I've been using Si the whole time, it still works.  Ge will just work closer to the original design.

Small Bear and Effects Connection both carry 1n34a's which are Ge...


Nice name choice btw.

[heisenberg]

oh really? I guess I will try to breadboard it then. I never did cuz on the schem at ggg.com  it says it won't work right. I guess "right" was used a little loosly.

What are you trying to tweak on the orange squeezer?

[aziltz]

oh really? I guess I will try to breadboard it then. I never did cuz on the schem at ggg.com  it says it won't work right. I guess "right" was used a little loosly.

What are you trying to tweak on the orange squeezer?

i've read that the way it clips is kinda nice so I'm trying to develop it into a half-compressor/half-light overdrive by duplicating the op-amp and letting the second one clip.  although I'm not sure I can do it with just one dual op-amp.  right now on the board I'm using two ICs, with buffers inserted between the clipping stages.

I have a "Sustain Control" on my version, where I replaced the resistor in series with the diode with a 10k Pot.  This lets me adjust the compression threshold, so the type of diode doesn't matter as much in my setup.

[cpm]

schottkys (eg 1n5819)
have a similar low drop like Ge's

[Mark Hammer]

You can use a silicon diode withut any modification to the circuit, but it will not work quite as well.  The thing is that the Si diode has a higher forward voltage which means that it will turn on juuuuuust a little slower, since it takes a few moments for the input signal to rise to the critical level to start governing the JFET.  It also means that the control voltage will shut down a little sooner, since once the signal hitting the diode drops below roughly half a volt, the pathway shuts down.

The good news is that yo can offset it a bit by boosting the gain of the circuit a bit.  Either increase the 220k feedback resistor to 270k or reduce the 10k resistor to ground to 8k2.  The other thing you can do to offset the possibility of audible ripple during the decay is to increase the 4u7 cap to ground to 10uf.

[heisenberg]

thanks. U mean the cap by the diode, right? 

I breadboarded it and it works as far as i can tell. Not that noicable effect, and theres a loss on the high end. Maybe its just the tlo72? I used a 680 ohm instead of 1k5 for the diode resistor and 1n914. Ill try these new suggestions now.

[heisenberg]

So mark, reducing the diode resistor is not the way to go?

[heisenberg]

I have a couple 1n5818s. Will it make a difference if i use those? I dont really have a good ear for this. Ive never played with real guita pedals.

[heisenberg]

Is it even worth biulding? Are tgere any sounf clips online? I cabt find any.

[heisenberg]

Nm. I found some. But notnwith the volume high to get the overdrive.

Do you think thaysbhow they did it? With the silicon re issues? Just ibcreaaed or decreased the feedback resistors?

[aziltz]

the diode resistor definite controls how much signal gets feed back to the FET controlled resistance.  you might call it sustain.  make sure you set the bias on the FET to the sweet spot though!

[heisenberg]

What value resistor should i use?? I used 680 ohm.

[Rob Strand]

Use the Schottky.  Nothing else will be closer.

As Mark mentioned the silicons have a turn on voltage than a germanium.     That means the output voltage needs to be higher before that part of the circuit will do anything at all.  You could double the feedback resistor on the opamp which will re-scale the output voltage so the relative threshold is the same (then half the output level to make the output the same again).  It's still not quite the same as a Ge and the extra means the opamp has more chance of clipping during the attack period.

[heisenberg]

Jees,double it?? Mark said frpm 220k to 270k. Ill try it on the breadboard.

So the silicon re-issues were just crap??

[heisenberg]

By half the output level, do you mean turn the volume down?

[Mark Hammer]

The op-amp in the OS provides gain for the envelope/rectifier stage as well as amplifying the audio signal.  So, if you boost the gain of the op-amp to feed the diode a bigger voltage, you also get a boost in output voluem, which you may not want.  Of course, you are always free to turn the OS volume pot down, but if only a small portion of the pot's range is of use to you, that's a nuisance.  It is easily fixed, though, by simply sticking a fixed resistor in series with the input lug of the volume pot.

So, if you increase the feedback resistor from 220k to 270k, and dropped the 10k going to ground to 8k2, that stage would have a gain of around 34x instead of the 23x it has now.  That ought to get you over the "diode hump" easily, without producing any distortion.  Then, you can offset that extra boost by inserting a 3k3 resistor between the junction of the 1k5 resistor and 4u7 cap and the output pot.

[heisenberg]

Yes! Ok, i will try that now too. I just tried it with a 390k feedback resistor and messed aroubd with the schottky. Very bad. Lemme try agains

[heisenberg]

thanks that seems to have done the trick as far as i can tell! Orange squeezer with silicon diode! Bam! I used schottky 1n5818

[heisenberg]

To clarify what i did:

Feedback resistor:220k
10k resistor changed to 8k2
Diode bleed cap: 10uf
Diode: schottky 1n5818
3k3 resistor in series with volume pot.

I kept the 1k5 resistor. I dunno how far to lower it. Ill experiment some more. Its wierd getting the trimmer right, im not sure how much is too much. Ill try it with a big muff.

Any more suggestions before i biuld it?