RAT Q / Mod idea...

[Brett Sinclair]

I was looking at the Ruetz Rat mod (http://www.diyguitaramp.com/rat.html) and i've got some questions about that gain-setting rc high-pass thing there...

If i understand anything about theory correctly, the 560 ohm + 4.7 µ = max gain of 180 @ 60 Hz and 47 Ohm + 2.2 µ = max gain of 2129 @ 1539 Hz? Right?
Thus taking out the 47 + 2.2 results in less gain + flat feq response (less high freq distortion).

Schematic:


Now i figured to replace the drive pot with 250K and to add a 1K pot to dial between one end 220 Ohm + 2.2 µ and other end 100 Ohm + 1 µ.

Then the freq response and gain would be more or less the same when the wiper is at the end of 100 Ohm resistor.
But as the wiper moves to the 220 Ohm, the hi freq boost would lower but also the overall gain. And at the other end there's more bass rolloff and mid boost...

Think it's worth a try... any ideas?

[StephenGiles]

The word noise comes to mind, but it's certainly worth a try. Does it make your fingers move faster

[Mark Hammer]

The distinctive tone of the Rat comes from that demand for outrageously high gain from the mids and treble from an op-amp that struggles to provide it.  The differential gain for the lows and mids/highs is a result of the dual ground path identified.  So, clearly changing the tonal balance by means of tinkering with how much gain is allotted to each segment of the signal is an idea worth pursuing.

There are two aspects to it.  One is where the split occurs in terms of the two ground paths, and the other is the degree of "advantage" accorded each segment of the signal.  As you note, there is one heckuva lot more gain applied to the top end than the bottom when the Gain control is dimed.  And, as is suggested by the diagram shown, anything that results in more resistance along either of the two ground kegs will necessitate a higher feedback resistance in order to achieve the same amount of gain as always.

I may be presuming too much, but in some respects, you want to make it easier on yourself to restore the stock Rat sound.  As a result, while the suggested mods you show will make for some interesting experimentation, one builds a Rat for the Rat sound and if you can't get a Rat sound, why be constrained by all the other things inherent to a Rat?

So, I would suggest a slightly different tack.  First, stick with the 100k Gain pot and accompanying cap; a known quantity.  Second, while a colour pot is a useful idea, try opting for a value that lets you re-find the Rat sound easily, e.g., at the pot's midpoint.

Note that when the resistance along each ground leg is altered, that changes not only how much gain is applied to the band addressed by that path, but also where the split-point occurs.  Each path specifies a point above which gain will be applied, so the split-point is really a question of applying more gain above a certain frequency. 

Of course, if the lower band (wherever that happens to start) has its own low-frequency rolloff, that also alters the tone.  Stock, the low end applies gain down to around 60hz.  As the 560R resistance gets dropped, there is not only more gain applied to the low end, but the low end gets shrunk a bit.  For example, 330R in place of 560R, raises the maximum gain for the low band from 180 to 304, but it also imposes a rolloff below 100hz.  So, you don't want to tinker with the resistance of the low-end path too much.

T'wer I, I think I'd opt for a 330R fixed resistor instead of the 220R shown, and a 500R pot instead of 1k.  What I would also do is stick a fixed resistor in parallel with the treble side of that colour pot. to limit how much it reduces the "treble advantage".  So, for example, with a  500R colour pot, you can have a 39R fixed resistor instead of the stock 47R, and place a 220R fixed resistor in parallel with the treble side of the pot so that the resistance added on top of the 39R never exceeds 152R (total max resistance along "treble leg" = 191R).  With a 39R/2u2 combination, that will result in a gain of 2565 at 1855hz and above at max gain and the colour pot tuned to min resistance on the treble path.  Using the parallel 220R component, that will result in a max gain of 524 at 378hz and above.

That strikes me as about as much variation in colour as you might want.  YMMV.

[soulsonic]

This is a mod I did that sounds good to my ears:
http://www.aronnelson.com/gallery/v/soulsonic-fx/ProCo+Rat2+Modded+-+resized.jpg.html

[Brett Sinclair]

@ soulsonic: cool i hadn't seen that one before!

@ Mark:
Thanks for the detailed reply, confirms i got the math right. The values you suggest indeed seem wisely choosen and it's certainly something i'll persue further... (Seems to me such a "colour" control is interesting in a tubescreamer-type circuit aswell. Makes it more versatile than having merely a lowpass filter at the end.)

Back to my spreadsheet and veroboard. 

[slacker]

If you haven't seen it already check out the Beavisaudio FKR (four knob rat) it's a highly modded rat that includes the Ruetz mod using a 1k pot instead of the 47 ohm resistor.

I've just build a rat style thing, using Joe Davisson's discrete opamp with the same mod and it's a really nice control. Basically it's sort of like a presence control, increasing the resistance tames the top end and gives a less aggressive distortion, this is especially nice at lower gains or with diodes in the feedback loop for a more overdrive like tone.
If your idea works out it would be an even more versatile control.