Cap in parallel with diodes

[GGBB]

Op-amps clip based on the positive and negative voltage supplies and the bias voltage. In a typical +9V single-sided supply with 4.5V bias (1/2 * 9V) that means up to +/- 4.5V - but most op-amps are limited to a bit less than the power supply rails - you need to check data sheets for actual output voltage.

[antonis]

According to datasheets, it can swing 1 to 2 V lower than its power supply  rails..
(always dependent on output current..)

edit: Didn't see Gord's answer..

[PRR]

what voltage does the opamp clip?

Look on the op-amp datasheet. Tip- they don't want to tell you clearly.


LM308 (a popular RAT part) is very old and has been made several ways. The only data is for the old +/-15V supplies. Obviously it can't pull all the way to either supply (+ or -). "Typical" drop is about 1V but they only promise 2V, all at +/-15V supply. Does it scale to lower supplies? Well most of the loss is diode drops which run 0.6V-0.7V at any sane current. So at zero and +9V supply, we would hope for +1V on the low side and +8V on the upside, but maybe +2 and +7V for a weak chip or heavy load.

[Christoper]

Thanks guys, this information has been very useful and I can't think of too many other places I would be able to aquire it.

One last question before I'm done tweaking my Rat (which could make a pretty colorful euphemism) is about two of the filters. In the feedback loop the 4u7 and 560r HPF caps have a cutoff of around 60 Hz, but right before the clipping diodes there is a high pass filter of 4u7 and 1k with a corner frequency of around 30 Hz. If I were to swap the resistors of these filters, it would have no effect until the instrument reached op amp clipping correct? In which case the frequencies between 30 Hz and 60 Hz would create a fuzzier type of distortion? I reckon it would be fairly subtle

[GGBB]

Not the same. Relative to a guitar signal, 30Hz and 60Hz both pass everything, but you need to consider gain as well. The 560R/4µ7 filter really only has meaning relative to the 47R/2µ2 filter where it passes frequencies lower than that of the 47R/2µ2 filter but at much lower gain, which is variable but synchronized across the two filters. Changing 560R to 1k lowers the gain of 560R/4µ7 relative to 47R/2µ2 and thereby changes the sound. This is the reverse of how my bass mod works so you've experienced this already.

[Christoper]

Right I forgot to factor in gain. I was thinking swapping the resistors would be easier, but that would affect gain. I would have to double the feedback cap to 10uF and halve the diode cap to 2.2uF. But I guess, yeah, that wouldn't have a meaningful effect either because a standard tuned guitar doesn't go down that low.

I guess this circles back to the first couple posts of this thread, the way to do what I want would be to increase the gain of the 560R/4u7 leg or decrease the gain of the 47R/2u2. I definitely like how the gain gets fuzzy with the gain cranked with your bass mod.

[Christoper]

what voltage does the opamp clip?

Look on the op-amp datasheet. Tip- they don't want to tell you clearly.


LM308 (a popular RAT part) is very old and has been made several ways. The only data is for the old +/-15V supplies. Obviously it can't pull all the way to either supply (+ or -). "Typical" drop is about 1V but they only promise 2V, all at +/-15V supply. Does it scale to lower supplies? Well most of the loss is diode drops which run 0.6V-0.7V at any sane current. So at zero and +9V supply, we would hope for +1V on the low side and +8V on the upside, but maybe +2 and +7V for a weak chip or heavy load.

Looking at a waveform it looks to clip at about +/- 4.4v
I can't figure out how to attach the image though

[PRR]

A good simulation was posted here.
https://i.postimg.cc/Kzg3yC0n/Rat-Slew-1.jpg

For the simulated LM308: swing @9V supply is 0.6V to 7.8V and well squared-off.

[Christoper]

So the diodes clip considerably earlier than the op amp.

Mr. GGBB, one big thing I'm noticing with your mod is that the added bass definitely makes the extreme end of the gain knob into more of that classic wooly fuzz sound. How big of a difference is there between the double and triple ratio that you put on the switch?

I really like the thick gnarly fuzz I can get with higher gain settings, but does the 3x ratio setting make it go from OD to fuzz faster enough to where it loses versatility?

[GGBB]

I haven't experimented with that. I don't use extreme gain settings and actually only use a 50k pot. I also use either MOSFET or LED clipping. If you are raising the relative gain of the 560R/4µ7 filter either by lowering the 560R or by raising the 47R (with 100k for the pot as a constant), you will move more into fuzz territory. Minimum gain will always be 1 (pot value zero), but max gain will be greater only if you lower 560R. If instead you raise 47R, max gain stays the same for the 560R/4µ7 filter (lower for the other filter and on average) so therefore "versatility" should be the same.

[Christoper]

Interesting, so you increasing the gain of the 560R/4µ7 filter and lowering the gain control to 50k should be going in the same direction as me lowering the gain of the 47R/2µ2 and keeping the stock pot, right? Granted it won't be exactly the same, but reasonably similar.

What do you like about the MOSFET and LED clipping? The only LEDs I've experimented with are IR in a DS-1 type circuit, and I've never used MOSFETS. Do you just use one MOSFET for both directions of clipping, or do you use two?

Also which setting on the bass switch do you find yourself using more?

[GGBB]

Interesting, so you increasing the gain of the 560R/4µ7 filter and lowering the gain control to 50k should be going in the same direction as me lowering the gain of the 47R/2µ2 and keeping the stock pot, right? Granted it won't be exactly the same, but reasonably similar.

Relative to low frequency gain, I think that's a fair approximation, yes.

What do you like about the MOSFET and LED clipping? The only LEDs I've experimented with are IR in a DS-1 type circuit, and I've never used MOSFETS. Do you just use one MOSFET for both directions of clipping, or do you use two?

I think I mentioned earlier - I generally prefer higher clipping thresholds - they feel slightly more touch sensitive to me. I think my current RAT has BAT48 in series with the MOSFET (not using the MOSFET body diodes) - paired.

Also which setting on the bass switch do you find yourself using more?

560R||270R 4µ7||10µ - always on.

[aron]

>I generally prefer higher clipping thresholds

Me too in general.

[Christoper]

I choose my clipping diodes based off how much gain the circuit has. LEDs are no good if your circuit struggles to reach the voltage threshold. Germanium diodes are no good if you squash your entire signal to a single volume no matter what you set the gain to.

[Christoper]

I know I've said it before, but one last mod on this Rat. I will replace one silicon diode with an IR LED, which is the same mod I did to my best DS-1. Given the Rat has an audio taper gain knob, it should make less of a different to the sweep of the gain control than on the linear taper DS-1 correct? Everything above 60k on the gain knob sounds the same, and I think changing the clipping diodes would be the least drastic decision. If the signal saturates at 1.4/2 = +/- 0.7v with my current circuit at 60k ( 1 + [60k / 100||560] = 700 times gain), changing one of the diodes with an LED would change the clipping voltage to (1.2+0.7)/2 = 0.95v. To reach saturation at 0.95v I would require 700*0.95/0.7 = 950 times gain. Given my max gain is 1 + [100k / 100||560] = 1176, I should be able to reach that same amount of saturation, right? I could in theory put two IR LEDs (1200 times gain) and still be almost able to reach saturation.

Am I out of line here?

[GGBB]

I think in terms of clipping that's a decent estimate. Raising the clipping threshold has a similar effect as lowering gain. What you end up hearing may or may not correlate directly with the math.