Crazy "Amp in a box" idea: need help...

[greatmagnet]

M'kay I'll measure again the proper way.

[The Scum]

Don't forget to unplug the box from everything before measuring, too. Otherwise, you're measuring what it's connected to, in addition.

The easy way:
-plug into the input, measure resistance between tip and sleeve. That's the overall pot value. Call it R1.
-plug into the output, measure resistance between tip and sleeve. That's the proportion of the pot that's left at the end. Call it R2.

So the overall attenuation factor is the ratio of the two: atten = r2/r1

Think about it a little deeper - with the pot all the way up, R2 = R1. Thus the factor is 1. output equals input.

When it's all the way down, in your case, it's something like 4/95 = .042. Multiply input volts by .042 to get output volts.

Then, to put it into units you're probably a little more used to dealing with: Do 20*log(atten) to get dB voltage attenuation: -27 dB.

[ashcat_lt]

I mentioned out-Z in my last post. Now that I think about it, the pedals you're plugging this into are probably 1M in-Z. Even with all of the 100K in series, you're still obeying the 10:1 rule of thumb.

If you could come up with a good ratio and use that to find smaller resistances it certainly wouldn't hurt. The fixed divider, though, will work for this particular application with the settings you've got now. Change anything and it might not be exactly ideal. Better, I think, would be to use a smaller pot, which would make it more useful for more applications.

Also, don't forget my idea about using fixed resistors to set a max and min value, to get more useable resolution when turning the pot.

[The Scum]

My original thought was regarding hanging a pot on the speaker output of an amp, in parallel with the speaker. My concern was making the load as small as reasonably possible, so it was effectively invisible to the amp. 100K in parallel with 8 ohms would do that.

But it also happens to be a reasonable value in this circuit - a 1M guitar input won't appreciably load it.

As far as using it as a generic pad for universal application, it's a bit of a compromise. A 10K input would significantly load it. So we fall back to 10:1, and reach 1K...which the good ol' tl072 won't be very happy driving...so we bump up to 5K and hope for the best.

And to answer a question we ignored: pot quality:

2.) If I DID want to make another varaible resistance box that is more accurate (goes all the way down to a full 100k of resistance), where do I find a more reliable pot? Are pots sometimes sold with a stated degree of tolerance? Much in the same way that a metal-film resistor is stated to be accurate within 1% whereas most other are 5% or more?

The tolerance usually indicates the overall resistance - a 100K pot with 10% tolerance could be between 90K and 110K. But the wiper should still travel between one extreme and the other. If it can't hit both 0% and 100%, it's likely a mechanical problem with the pot itself, which isn't reflected in the specs.

For the purposes here, the tolerance won't make much difference...for the same rotation, the 110K and 90K pots should give the same amount of attenuation. But the overall quality of the pot has an impact. I'll say that I've used an absurd number of the Mouser/Alpha 24mm pots through the years, and never had any serious complaints. I've got a sneaking suspicion that Bourns new line of affordable "Guitar Pots" are make in the same factory as the Alphas...they look very similar, and Bourns probably realized that the 51 series had become out of reach for a lot of purposes.