RAT mod with MOSFETs, LEDs, etc.

[jarick]

Hello, first post so I'll introduce myself.

I love EE and love music, so naturally I want to find out how and why everything works.  I'm also a tone perfectionist, but broke, so modding pedals has been a real interest to me.  I've done a few mods from kits, and now I'm working up to schem reading and planning my first "orginal" mod...

So anyways, I've got a RAT2 coming to me next week.  I was inspired by the whole OCD and RAT controversy, as I think the OCD is probably one of the best distortion pedals out there.  The clarity and response of the pedal really does feel closer to a tube amp than any other distortion I've played.  Naturally, I had to learn everything I could about it, including the de-gooped board picture.

Here's my plan thus far:

- Replace ceramic caps with film and mica for additional clarity.  Tantalum caps to replace electrolytic.  I'll tweak some of these values as need be.

- Possibly add a small cap across the volume pot to tame the highs a bit IF it's too bright.  I want the response of the pedal to be between the LP and HP modes of the OCD if as a reference point.

- 3-way switch a la MultiRat schematic, but with the following options:
   1. Standard 2x1N4148 diode clipping
   2. Guv'nor style 2xLED clipping
   3. OCD style 2xMOSFET clipping

- Add a 1M trim pot resistor to the opamp filter for fuller sound as in a mod I can't remember (Muzique?)

I'm going to play around with asymmetrical clipping to see if it sounds a bit more natural.  Also, looking at the de-gooped OCD, it appears to have 3 MOSFETS, probably connected G+S-D asymmetrically.  I'm going to try both this way and with germanium diodes as in the Shaka circuit.

So what do you think?  Any MOSFETs you'd recommend? I'm going to try IRF520's and 1N34A's as used in the Shaka, but I haven't seen an extensive use of MOSFET clipping, so I don't know the variety in use.

Thanks!

[scaesic]

arn't tantalum caps the worst offender for "cap distortion"?

and the rat uses diodes to ground after an op amp gain stage, how can you add mosfets for the distortion stage? the fet in the rat is just used for true bypass + led using a dpdt?

[Bernardduur]

Using MOSFET's for diodes are great. Check out the Shaka 5 for this!

[wampcat1]

arn't tantalum caps the worst offender for "cap distortion"?

and the rat uses diodes to ground after an op amp gain stage, how can you add mosfets for the distortion stage? the fet in the rat is just used for true bypass + led using a dpdt?

6 to one, half dozen to the other as far as 9v pedals are concerned IMO.
What the poster is talking about is using mosfets as diode clippers. The Fet in the stock rat is just a buffer, older rats used jfet with millenium type switching circuit.

bw

[jarick]

arn't tantalum caps the worst offender for "cap distortion"?

I dunno...I used them in some Monte Allums kits and they sounded decent.  I may just try regular electrolytics and see what the difference is.

and the rat uses diodes to ground after an op amp gain stage, how can you add mosfets for the distortion stage? the fet in the rat is just used for true bypass + led using a dpdt?

I'd wire up the MOSFETs like in the Shaka circuit...trying both with and without additional diodes.  There should be enough voltage after the opamp to clip the MOSFETs...

[jarick]

Ah, he beat me to it!

One more thing: anyone tried the LM308AH metal can versus LM308N DIP for the opamp?  Lots of talk about the "metal can" mojo but I've heard both good and bad.

It's only $4, might as well see what she does.

[JHS]

IMHO the only useful klipping mod for a RAT2 is: 2 diodes, none and 3 in asym. wiring and a 1M pot (trimpot) in series w  the 150k BIAS R for the opamp (forming a simple pre-Bass control). You can replace the 33p ceramic w Silva Mica for a clearer sound (100p are fine too, but this yields in more overall compression).

After the OCD desaster I would not recomment any Mosfets alone as a substitution for klipping diodes (to me the distortion sound is clinical, compressed and very artificial). Interesting that there are 3 in your OCD, I only discovered 2 in sym. wiring on the boards I' ve seen so far.

On the Shaka the diodes in series w the Mosfets seems to have the most influence on the clipping and the distortion structure, the Mosfets only kill some harshness at higher distortion.

BTW:
I prefer high end plain foil electrolytics, but IMHO tantalum caps sound way better than the cheap rough foil electrolytic caps.

The metal can chip is the older stock stock chip w outside metal shielding, the newer chips are all inside shielded.
The chip is OK when it's the discontinued Motorola, 308N from other manufactures will not sound that good in a RAT.

JHS

[Ed G.]

I don't know what the circuit of the OCD is, but I don't think a RAT can be modded to be like it. The OCD has none of the compression or sustain of the RAT. Even the turbo rat, which has less compression, more volume, has lots more sustain. Not necessarily a bad thing, but the OCD is more 'overdrive' than 'distortion' to me. Personally, I don't care for the OCD, but I didn't have much time to play with it.

[jarick]

IMHO the only useful klipping mod for a RAT2 is: 2 diodes, none and 3 in asym. wiring and a 1M pot (trimpot) in series w  the 150k BIAS R for the opamp (forming a simple pre-Bass control). You can replace the 33p ceramic w Silva Mica for a clearer sound (100p are fine too, but this yields in more overall compression).

That's actually similar to what I was going to do...I'm just not familiar with the terminology.  "Bias resistor"...I'll have to remember that one!

I am going to experiment with a lot of different clipping schemes...I'll have a separate perfboard so I can swap out circuits easily...should be fun at least.

Thanks for the tip on the 33p cap...I was going to do a silver mica there, but I'll keep in mind the cap value change.

After the OCD desaster I would not recomment any Mosfets alone as a substitution for klipping diodes (to me the distortion sound is clinical, compressed and very artificial). Interesting that there are 3 in your OCD, I only discovered 2 in sym. wiring on the boards I' ve seen so far.

Do you mean OCD disaster as in the pedal itself or just the fiasco surrounding it?

I do like the OCD and will pick one up someday.  But as I said, it's kind of an experiment and learning for me as it's the first non-kit mod that I'll be doing.  So baby steps on the way towards pedal design I suppose.

Here is the picture I'm referring to:



On the bottom left portion of the circuit board, there appears to be three FET's.  I know there's a pair for clipping.  But there might be a third one (just to the top left of the bottom pot).  I don't know if that's a FET for sure, but it sure looks like one.  I don't see any diodes, except for possibly one just to the left of that bottom pot.  It could be a resistor though, it's hard to say.

One thing's for sure, the OCD uses MOSFET's for it's main clipping.  And there's a lot of love for the Shaka design, so I will probably end up with FET/Germanium clipping.  Like I said, it's an experiment.

On the Shaka the diodes in series w the Mosfets seems to have the most influence on the clipping and the distortion structure, the Mosfets only kill some harshness at higher distortion.

I think the "rounding" of the MOSFET's definitely would help.  Should be a good sound.

BTW:
I prefer high end plain foil electrolytics, but IMHO tantalum caps sound way better than the cheap rough foil electrolytic caps.

The metal can chip is the older stock stock chip w outside metal shielding, the newer chips are all inside shielded.
The chip is OK when it's the discontinued Motorola, 308N from other manufactures will not sound that good in a RAT.

JHS

I think I'll go with the electrolytics from Small Bear.  Should be good quality.  As for the chip, I'll hold off until I can find a "real" 308N.

Thanks for the insight!

[jarick]

I don't know what the circuit of the OCD is, but I don't think a RAT can be modded to be like it. The OCD has none of the compression or sustain of the RAT. Even the turbo rat, which has less compression, more volume, has lots more sustain. Not necessarily a bad thing, but the OCD is more 'overdrive' than 'distortion' to me. Personally, I don't care for the OCD, but I didn't have much time to play with it.

You could be very right.  But rumor has it (since nobody has traced the schem) that it's a RAT with MOSFET clipping.  I wanted a distortion pedal with post-opamp clipping that would be a full-range distortion that's fairly transparent when cleaned up.  I found a RAT2 with true bypass for $50 shipped, so I figured that's a good jumping-off point.

My theory on the OCD is that it's a distortion pedal but with high quality components.  And if that small value cap between pins 1 and 8 on the opamp controls the compression of the pedal, reducing the value could open it up.

[WGTP]

Remember, there are 2 ways to use Mosfets.  The actual Mosfet and the body diode.  There are many different Mosfets to try.   

http://www.diystompboxes.com/smfforum/index.php?topic=38581.0

[kusi]

OCD uses only two mosfets for clipping - the third transistor is used with a diode as a current-source, for the "non-popping-led-switching" 

[dano12]

I spent most of today working on my modded rat clone. Here are some of the more interesting things that I found:

- Adding an asymmetric clipping option with one side being a single 1N914 and the other being two germaniums in series does add a nice subtle change.

- Adding an option for Red LED clipping made a *really* big difference. Takes the edge off and sounds more like an overdrive than a distortion.

- Remove the 47 ohm resistor that goes from LM308 pin 2 to the 2.2uf to ground. Replace it with a B1K pot. This has a really subtle softening effect. But here's the kicker: when you switch to the LED clipping, this new pot affects the overall tone in a very cool way. It turns the RAT into one of the best overdrives I've heard. Yep, pretty bold statement, but it is really worth a try!

- I tried a TL741 as an alternate opamp. Not very good--adds a lot of high frequency buzziness and pretty much ruins the Rat's charms.

- I added the 1K trim pot before the biasing diodes. It allows a bit more control over clipping, but I'm not sure that it warrants the extra effort for my build.

- I tried lots of options with MOSFET clipping but could not get it to work. All I had were BS170s so that may have been the problem.

Hope this helps, and can't wait to hear about your build.

[jarick]

I spent most of today working on my modded rat clone. Here are some of the more interesting things that I found:

- Adding an asymmetric clipping option with one side being a single 1N914 and the other being two germaniums in series does add a nice subtle change.

Sounds interesting.  I'll give it a shot and see if it's better than the stock clipping for that "mode".

- Adding an option for Red LED clipping made a *really* big difference. Takes the edge off and sounds more like an overdrive than a distortion.

Was it just LED's?  Symmetrical or asymmetrical?  Sounds like fun though!

- Remove the 47 ohm resistor that goes from LM308 pin 2 to the 2.2uf to ground. Replace it with a B1K pot. This has a really subtle softening effect. But here's the kicker: when you switch to the LED clipping, this new pot affects the overall tone in a very cool way. It turns the RAT into one of the best overdrives I've heard. Yep, pretty bold statement, but it is really worth a try!

Yep, that's the bias resistor trim pot.  I'm definitely doing this one.

Here it is:

http://www.diyguitaramp.com/rat.html

The new Rat mod was a huge success. I took my time and did it right. Took me longer to get the thing apart than it did to do the mod!! Certainly now with a fourth parameter there's plenty of tweaking to be done but at first listen it was definitely worthwhile. With the pot wide open it sounds slightly hairier and warmer than with the resistor clipped. As you dial in the pot you get closer to stock Rat and can in fact REALLY hear when you get to the stock value. Almost like the pedal wakes up and is reborn to what it was (not necessarily a good thing) but just before that is the TONE!! As I said. There's plenty to learn yet. The filter control and the trim pot are highly interactive so there are lot of tones to be discovered! The best part is that the trim pot is conveniently accessible just by opening the battery compartment door!! A 500-ohm pot would have been ideal but this 1K from Radio Shack works fine.

- I tried a TL741 as an alternate opamp. Not very good--adds a lot of high frequency buzziness and pretty much ruins the Rat's charms.

Hm.  I was thinking about trying a JRC4558 in there, but I know very little about IC chips and their differences, so I'll probably just leave it alone.

- I added the 1K trim pot before the biasing diodes. It allows a bit more control over clipping, but I'm not sure that it warrants the extra effort for my build.

Also good to know.  I wonder if this is the "saturation" control that's on the Fulltone Distortion Pro.  But yeah I think that'd be more helpful just to identify the final resistor value.

- I tried lots of options with MOSFET clipping but could not get it to work. All I had were BS170s so that may have been the problem.

Yeah...I'm just going to try the MOSFET stage from the Shaka.  That uses IRC520's.

Hope this helps, and can't wait to hear about your build.

Sure does, and it's funny that we're doing pretty close to the same thing at the same time.

How'd you like the pedal overall after the mods?  How would you describe the sound compared to the stock RAT or an OD pedal like the Tube Screamer?

Good luck, and I'll let you know how the pedal build goes next week (or whenever I get all my parts in)!

[dano12]

Was it just LED's?  Symmetrical or asymmetrical?  Sounds like fun though!

I tried both blue and red LEDs in standard back-to-back formation. The blues worked but the sound was pretty muddy. The reds are awesome

Yep, that's the bias resistor trim pot.  I'm definitely doing this one.

You know I'm wondering...how useful would it be to add another pot to the other leg of the feedback/bias loop? I.e. the one above 4.7 uf cap also? That way you could control the feedback loop at two seperate frequencies...

Hm.  I was thinking about trying a JRC4558 in there, but I know very little about IC chips and their differences, so I'll probably just leave it alone.

Anything that is pin-compatible would be worth a shot. If you socket it, you can swap different values in and out. Jack Orman wrote a great article about the RAT and in it he talked about the "abysmal slew rate performance" of the LM308. But that negative is part of what gives the Rat its sound.

Yeah...I'm just going to try the MOSFET stage from the Shaka.  That uses IRC520's.

Let me know how that goes. I'd love to try it.

How'd you like the pedal overall after the mods?  How would you describe the sound compared to the stock RAT or an OD pedal like the Tube Screamer?

Well, its still the honeymoon period and all, but I couldn't stop playing it last night. With the mods it is so tweakable and it is just about impossible to get a bad sound out of it. This will definately be a keeper. I haven't had a stock Rat for quite a while so its hard to compare. As for the TS comparison, the Rat will definately move you well into distortion territory unlike the TS. Plus there is no mid-hump quack like in the TS. Finally, with the mods, the Rat makes an awesome overdrive, so in my lineup it would replace my TS.

Thanks,
-dano

[nelson]

Dont use a 4558.

The LM308 is a single opamp and has standard single opamp pinout whereas the 4558 is a dual opamp with dual opamp pinout.

Replace the LM308 with single opamps


LM741
TL071
CA3140       


etc etc

[jarick]

I tried both blue and red LEDs in standard back-to-back formation. The blues worked but the sound was pretty muddy. The reds are awesome

You know I'm wondering...how useful would it be to add another pot to the other leg of the feedback/bias loop? I.e. the one above 4.7 uf cap also? That way you could control the feedback loop at two seperate frequencies...

Now my understanding might be way off here, but if the R-C network in the bias resistor attenuates lower frequencies, then the standard rat circuit has 6dB rolloff at 60.5 Hz and 6dB rolloff at 1540 Hz.  The reason so many people would recommend clipping that 47 ohm resistor is that 1540 is a pretty high frequency to roll off the lows (if that's what it's doing).  Alternately, I've heard about increasing the 2.2uf cap to 10uf or higher, which would reduce the frequency to 338.8 Hz.

Now I totally understand the value of the trimpot on the 2.2uf cap, as that's a pretty high frequency and controls the amount of volume boost above a certain frequency (again, I THINK...need confirmation/correction here).  But at 60.5 Hz, I couldn't imagine wanting to roll off any higher than that.  It seems like a pretty good frequency.

IF the RAT is to be used as a bass overdrive, THEN I'd put a trimpot on that resistor as well.  But I don't want 60 Hz on my guitar, so I'm fine

Anything that is pin-compatible would be worth a shot. If you socket it, you can swap different values in and out. Jack Orman wrote a great article about the RAT and in it he talked about the "abysmal slew rate performance" of the LM308. But that negative is part of what gives the Rat its sound.

I'll try and dig that up.  Since I'm not really going for a standard RAT sound, finding subs might be cool.  Basically, I'm using the RAT as a basic schematic/springboard.  You said you didn't like the LM741 (from the Distortion+)...so maybe without completely gutting the RAT, the IC should be left alone.

Well, its still the honeymoon period and all, but I couldn't stop playing it last night. With the mods it is so tweakable and it is just about impossible to get a bad sound out of it. This will definately be a keeper. I haven't had a stock Rat for quite a while so its hard to compare. As for the TS comparison, the Rat will definately move you well into distortion territory unlike the TS. Plus there is no mid-hump quack like in the TS. Finally, with the mods, the Rat makes an awesome overdrive, so in my lineup it would replace my TS.

Thanks,
-dano

That sounds like a blast!  And it'll hit some people upside the head who don't really consider the RAT anything but distortion (or the OCD as anything but overdrive, but hey).

For me, I'd still like to find a pedal that does a nice borderline crunch/borderline distortion.  I think one key to that is achieving a flatter response (BR trimpot, better caps), another key is tweaking the clipping section to find the optimal clipping, and another might be increasing the 30pf cap for more sustain or decreasing it for less.  Not many OD pedals give a lot of sustain.

Anyways, thanks for sharing the insight, and hopefully between the two of us, we'll figure out something different

[dano12]

Here's the prototype so far. Quite a mess

[wampcat1]

By just changing the type of caps you more than likely won't get any flatter eq response. You need to either run a low or high pass filter (or both) or gyrator (see rg keen's geofex page) to get rid of that exact frequency. You may need to boost the signal again after the filters. Here's a soundclip of my Rat with these type of mods:

http://www.indyguitarist.com/soundclips2/ratmod-706.mp3

bw

For me, I'd still like to find a pedal that does a nice borderline crunch/borderline distortion.  I think one key to that is achieving a flatter response (BR trimpot, better caps), another key is tweaking the clipping section to find the optimal clipping, and another might be increasing the 30pf cap for more sustain or decreasing it for less.  Not many OD pedals give a lot of sustain.

[PaulC]

Now my understanding might be way off here, but if the R-C network in the bias resistor attenuates lower frequencies, then the standard rat circuit has 6dB rolloff at 60.5 Hz and 6dB rolloff at 1540 Hz.  The reason so many people would recommend clipping that 47 ohm resistor is that 1540 is a pretty high frequency to roll off the lows (if that's what it's doing).  Alternately, I've heard about increasing the 2.2uf cap to 10uf or higher, which would reduce the frequency to 338.8 Hz.

Think of it more like the circuit boosts the freqs instead of cutting them.  The 560 ohm resistor when looked at with the 150k gain pot (going by the schematic at this site) will give you a gain of 267 or about 48db with the -3db point being 60hz.  The 47 ohm/2.2uf setup will give you a gain of 3191 (70db) with -3db being 1543hz.


So while you saw the two corner freqs at 60HZ and 1543HZ the real story is the difference in db between the two networks.  You could have the same corner freqs with different ranges of gain.  Say for example you made the 47 ohm network 220ohm/.47uf.  you'd still have the same freq point, but the gain for that network would be 681 now.

There's a mod I've done a bunch of times where I just swap the resistors.  The 560/2.2 setup gives you the same low freq gain, but the -3db is at 129hz.  This tightens up the bass a little bit.  the 47ohm/4.7uf will give you the same gain for that network, but the -3db will be 722hz.  This brings in more mids giving it a thicker sound.  Alot of times I'll then add a pair of LED's across the feedback loop while keeping the diodes to ground.  This  adds alot of crunch to the sound while taking out some of the fuzzy mush. 

Later, PaulC