MAXIMUM gain from a TS circuit?

[ethanw]

I was thinking about modding an 808 circuit for absolute maximun gain. I built one using "Grand Laff" specs, where the gain pot is changed to a 1 meg and the 4.7k gain resistor goes to 1k and the 51K goes to 10K, and also experimented w/the clipping diodes.   Is this all the gain the circuit can offer or are there other tricks?

Reading thru the GEO article on the TS, it looks like dropping the 1k even further should increase the gain. If I calculated things correctly, the Grandlaff values in the clipping stage should have a gain of 1010 and 20 compared to 107 and 12 in a stock TS as per the Zf formulas in the Geo article.  Furthermore, it looks like decreasing the 1k to 500 ohms should double the gain! Anyone tried this?  Of course, all this putzing around with the gain structure will alter the freq. response and require different cap values.

Ethan

[johngreene]

In theory, yes. The gain goes up as you describe. However, you need to factor in things like the gain-bandwidth product of the opamp. You also need to be sure 'gain' is what you are going after. Once the diodes conduct, the 'gain' is near unity. The pedal can sound more 'gainy' just by using higher threshold devices. They allow more signal to be affected by the gain before they conduct. If you really want a pedal with more gain, add a second stage.

--john

[ethanw]

Thanks for the reply, good info. I would probably use LED's for clipping and a standard 4558. This circuit w/the mods described is pretty gainy or capable of a fair amount of distortion. I'm looking for MORE of the same. What do you think will happen with the 500 ohm change? I'd put one in and find out but I'm out of town.

Ethan

[johngreene]

Thanks for the reply, good info. I would probably use LED's for clipping and a standard 4558. This circuit w/the mods described is pretty gainy or capable of a fair amount of distortion. I'm looking for MORE of the same. What do you think will happen with the 500 ohm change? I'd put one in and find out but I'm out of town.

Ethan

At some point you will for all intent and purposes have removed all feedback and the opamp is going to be a comparator. It will get really noisy with no signal input as it will be slamming rail-to-rail with small amounts of noise on the non-inverting input. If you look at the data sheet for the 4558, you will see that the maximum available gain for 10kHz bandwidth is 50 dB. Narrow it down to 1 kHz and, ideally, you have 70 dB available. So, I don't think you are going to get much more gain out of the device if you change that resistor to 500 ohms.

--john

[Mark Hammer]

And keep in mind that those gain factors make no presumptions about input signal.  Feed it an input signal with a 100mv amplitude and just how many times can you amplify it by before that puny little 9v battery can't make the signal any bigger?

Of course, when we say 100mv input signal, one should keep in mind that many portions of that overall signal are much less than that and can be amplified more than 9v/100mv = x90.  Still, the limits to "how much gain" are given both by what the op-amp can do, and what it can do with input signal X and supply voltage Y.

[d95err]

Squeezing too much distortion out of a single stage is usually not a good idea. Better to produce distortion in several stages. The first thing I'd try is a simple booster infront of it. Preferably a somewhat dirty booster. You could go fancy and add some clipping diodes to the booster to get the signal level down a bit before the TS circuit.

Otherwise, I'd look at another circuit design to begin with. Perhaps the Big Muff or one of those JFET "amp sims".

[R.G.]

This circuit w/the mods described is pretty gainy or capable of a fair amount of distortion.

OK, time for this quarter's nomenclature post.

Gain is NOT equal to distortion, no matter how many times one sees them used interchangeably in advertisements. They are quite different things.

Want gain? Just take that feedback resistor off your opamp. The open loop gain of the opamp is between 100,000 and several million. It's not controllable, does you no good, and is highly frequency dependent, but it's gain.

"Gain" is an electronic term for how much bigger the output is than the input. "Distortion" is how much DIFFERENT the output is from the input. You can get lots of distortion at unity gain or less.

[ethanw]

I ran into my old post and thought I'd bring it to the top, interesting stuff.  Thanks for all of the replies.  Reading thru the responses, it sounds like the "Grand Laff" 808 circuit is pretty much at maximum gain for a decent audio circuit, true? And if I wanted additional "distortion"  the best way to do it would be to add an additional stage.

Ethan

[GibsonGM]

You could change the gain of the recovery stage after the tone section to provide more output, and add additional clipping diodes to ground before the output.  Check out AMZ's site, there is some good data on how that would be done in a variable way...call it a "post gain gain control"  ;o)  It could be as fancy or basic as you like; although it would no longer really be a Tube Screamer, but more of a "Dist+ Screamer"...definitely interesting!

[John Lyons]

Good ideas gibson

Also check out the GMarts overdrive. It uses hard and soft clipping. Sort of an all in one pedal with two type of clipping that you can switch in and out.
Lots of ideas over at AMZ for sure!

John

[Jobet]

Try this :

Replace the clip diodes with LED's just as you said.

Add a resistor in series with the "Gain" pot. Oh I dunno...start with 10k, 33, 50 etc. At some point you will slam into the forward voltage of the LED's and once they start conducting, no more gain is possible. If you want more, remove them altogether and continue increasing the value of the resistor you're putting in series with the gain potentiometer.

In theory, an OP-AMP has a gain of about 200,000 (if my memory serves me right). That means, 1 volt can "theoretically" squeeze out 200,000 volts. But of course you can see why that's not gonna happen because you'll encounter the rail-to-rail voltages first.

Yup, slamming the gain on one stage is not a good idea IMHO of course. But if you needed "metal" like gains and a TS is all you have, you can do the procedure above. Yup done it before, even added a switch so my TS7 will go from TS mode to RAT mode (soft clip vs hard clip)

[R.G.]

In theory, an OP-AMP has a gain of about 200,000 (if my memory serves me right).

Your memory is not being helpful. 200k is quite small for the open loop gain of most opamps. Many have gains in excess of 1M, even more.

At least at DC.

[Jobet]

In theory, an OP-AMP has a gain of about 200,000 (if my memory serves me right).

Your memory is not being helpful. 200k is quite small for the open loop gain of most opamps. Many have gains in excess of 1M, even more.

At least at DC.

Thanks. I must have been thinking of the uber-generic LM741. My memory did serve me right though, looking at the datasheets, it says 200V/mV typical, which is, in unitless terms, 200,000

[R.G.]

I did some more looking.

It turns out that many opamps do have a large signal voltage gain of 200K. That's about 120db, and is what many garden variety opamps use.

However, it's common to have exceptions. The NE5535 is 500K; the OP-27 is 1M; the AD711 is 400K.