Help me understand why my O/D gain pot and diodes aren't doing anything

[wayfaerer]

Hi, been reading this forum for a few months but this is my first time posting.

I tried to implement a schematic I found online for an overdrive, but found that the gain control and the diodes do nothing; the gain does not seem to change when adjusting the pot, and there is no clipping. I then made the circuit in a simulator and found the exact same thing. I must be overlooking something and would appreciate some help figuring this out.

Here's my schematic


and the simulation in Falstad:
http://tinyurl.com/yaeajy5o

I adapted this circuit from the circuit at the bottom of this page: http://www.generalguitargadgets.com/how-to-build-it/technical-help/articles/design-distortion/. From that circuit I removed the input filters, the output clipping diodes, and the tone control.


Thanks!

[ElectricDruid]

I don't know about the real-life circuit, but I found a few things wrong with the simulation, although it doesn't do exactly what I expected either.

The op-amp is set up with +/-15V output. This is what you'd expect with the standard bipolar +/-15V supply for audio, but it's not a stompbox situation. Try changing that for max 9V and min 0V.

You've got no input capacitor, so the offset on the sine generator is fighting with the incoming bias voltage.

The diodes are set up with an 800mV forward voltage. While there *are* diodes that do that, it's a bit high for typical silicon. For experiments, you could drop it down to 500mV.

First, though, I'd remove the two wires leading to the diodes and make sure that you've got an op-amp set up that produces the gain that you expect. With a 2V sine input and x50 gain (your 50K pot) you should be able to get clipping from the *op-amp* without any diodes at all.

http://tinyurl.com/yc9ezmqe

But I still can't get the bias to give me a simple 4.5V for 4.5V input. Don't know Falstad well enough to know what's going on there.

[GGBB]

ElectricDruid beat me to it, but yeah - there are several problems with the simulation. The big one is the 2V input signal which even at minimum gain (unity) is higher than the diode drop. If you remove the diodes from the circuit and turn up the gain (left), you will see that the diodes were actually doing something. Also, the 100Hz test input signal frequency is a problem - at minimum gain, that frequency is attenuated considerably by the 1k-50nF (~3kHz) filter anyway so the diodes will have little effect at that frequency until gain is raised thus lowering the filter knee. The overall effect is that your test frequency is always attenuated either by the filter or the diodes, so isolating the effect of the diodes alone is difficult in the simulation, although you should be able to hear the difference in the actual circuit.

[pinkjimiphoton]

use a 1k input signal, thats right in the middle of our audio spectrum.

try running 9v to v+ and 0v to v- on the chip, and like the other said, ditch the diodes temporarily.

you want an input cap on there too, 10-100n is probably good, tho you can go higher... the bigger the cap, the more distortion usually, up to the point of subsonic noise. keep your input cap under 10u or so

your biasing is gonna be weird. you can run the half voltage into the + side,  but first ya gotta get it. put a pair of 10k-470k resistors in series between + and - and take your half voltage from the node in the center.

you may like diode clipping better AFTER the feedback loop, an antiparallel pair to ground just after the output cap should do the trick.

its almost there. if you're on a breadboard, play with the resistor values til ya find what sounds best.
for caps, i'd go smaller on input and output, really, say 100n input and maybe 220n output. if too bassy, use smaller caps.

good luck,  and welcome to the forum!

[stringsthings]

Nothing new to add but if you haven't already, get yourself a breadboard and go to town.
Simulations are good, but messin' around with components in real-time is more fun, IMO.
And these parts is cheep, so even if you smoke an IC, just use one of the other 3/4/10 that you bought.

( in all seriousness, check out mouser for very reasonable prices on IC's ... they can have good discounts if you get 10
.. and having extra TL07x's will always come in handy ...or 741's ... also tayda is cheap if you can wait for the longer shipping time )

[wayfaerer]

Thank you everyone for your input!

I seem to have gotten it working. I'm not sure if this is what fixed it or not (please let me know if you know), but I adjusted the biasing. In the circuit in the first post, I realized that although the voltages on the op amp pins was biased at 4.5V, the 1K resistor and 50nF cap were dumping to ground, instead of to the 4.5V virtual ground. I have no idea if that matters or not, but my current iteration, which works, is dumping to virtual ground there instead of ground.

Other than adding a coupling cap, the only other thing I did was remove the 47pF and 50nF caps altogether, I tried some different values, but I couldn't find a configuration that didn't sound awful. Clearly I have more experimentation to do, but for now, without the caps, it at least sounds not *too* bad. Now I'm off to see if I can figure out how to make it less crackly and more smooth/creamy.

Thanks again!

[ElectricDruid]

I seem to have gotten it working. I'm not sure if this is what fixed it or not (please let me know if you know), but I adjusted the biasing. In the circuit in the first post, I realized that although the voltages on the op amp pins was biased at 4.5V, the 1K resistor and 50nF cap were dumping to ground, instead of to the 4.5V virtual ground. I have no idea if that matters or not, but my current iteration, which works, is dumping to virtual ground there instead of ground.

Since the cap blocks DC it makes no difference what voltage it goes to.

Other than adding a coupling cap..

This will have helped!

..the only other thing I did was remove the 47pF and 50nF caps altogether, I tried some different values, but I couldn't find a configuration that didn't sound awful.

The 50nF cap was too small as GGBB and PinkJimi pointed out, so removing the 50nF cap will help give you more boost. And if you remove it, then it *does* matter what voltage the bottom of the 1K goes to, and it needs to be Vbias.

The 47p should give you some roll-off of the high frequencies, at a frequency which depends on the gain. This is generally a good thing as it tends to smooth more as the gain goes up.

Clearly I have more experimentation to do, but for now, without the caps, it at least sounds not *too* bad. Now I'm off to see if I can figure out how to make it less crackly and more smooth/creamy.

I'd put the 47p back in, and I'd add a resistor in series with your gain control, so that at minimum gain the resistance isn't zero. Perhaps 4K7 or 10K.

You can see what the roll-off frequency will be with a calculator:

https://electricdruid.net/rc-filter-calc/

You'll have (say) 4K7 with minimum gain. That gives 750KHz, so doing absolutely nothing. At maximum gain you'll have 50+4.7K = 54.7K, which gives you 66KHz. This tells me that the 47p isn't doing anything and you need to increase it markedly, by a factor or at least ten, so maybe try 470p or 1n. That should tame some high end and get you closer to creamy.

Thanks again!

You're welcome.

[pinkjimiphoton]

+1 on tom's post.

try using a couple caps and a pot to fake a variable cap, tie say a 47p cap to pin one of a 100k-250k pot, and a 4.7n to pin 3. tie the free ends of the caps together, and use the 2nd pin of the pot as an output. tie the joined ends of the caps to one side of the feedback loop, and the wiper <pin2 > to the other.
now ya should be able to not only tame the clipping, but have a bit of tone control available. even if you don't have a meter that reads capacitance, you can "ballpark" what sized cap will work best... turn the knob until it sounds good, then take note of how much of the pot you're using to guesstimate to the next closest value the optimal size of the cap.

do a search for "pinks clipper" for a really easy way to do a simple opamp od, or any of the stupid simple overdrivers. if the graphics are missing <probably, thanks to botosucket> i will reup them for ya

[stringsthings]

botosucket