My Rat Project

[XAXAU]

I´m building a RAT for christmas (used to get Lego as a kid but I can´t use it in the studio heh) and here´s my plans so far. I´m basing this on a Musikding kit but plan on modding it heavily. I´ll be building a prototype to see what works for me then V2 will be more refined. I don´t intend on using this on guitar but on hardware/software synths as well as drummachines.

Pots:

1. DISTORTION
2. FILTER
3. VOLUME


Rotaries:

1. CLIP TYPE (Non-shorting rotary, see clipping types)
2. HPF (Shorting rotary 1x12 with capacitor x, wired in series, located in the feedback loop)
3. ASYMMETRY (Shorting rotary 1x12 with diode x with low Vf)


Toggles:

1. CLIP DESTINATION (DPDT changeover to throw the clipping network between the stock position and feedback loop)
2. SLEW (Toggle between stock and "value x" to change the opamp slew rate, will try up to 1pF for more HF)
3. BOOST (Toggle between the "47pF-ish" in the feedback loop and "value x" to change the overall drive, would like it to go from subtle overdrive to decapitation at the flick of a switch)
4. CLIPPING ON/OFF (OPAMP)


Clipping types:

1. 1N914 (RAT V1)
2. 1N4148 (RAT V2)
3. Red Led (TURBO RAT)
4. 1N34A (DIRTY RAT)
5. BS170 (FAT RAT)
6. BC184C (I´m building a Neve summing mixer and have these trannies left over that are supposed to be the bees knees)
7-12. ?


Do you see any problems with this design?

1. I read about diode connected mosfets (G+D - S) on this forum so I thought I´d try that out as well as some other diodes that might sound cool. Recommend me some please!
2. Suggestions for the diodes in the ASYMMETRY rotary?
3. Is it stupid to put capacitors in series for the HPF rotary? Sound degradation? Better solution?


Thank you and merry christmas!

[GibsonGM]

I think you'd get some good answers if you began to draw out your idea, schematic-wise, X.   It might help you to keep switching and stuff straight.

Your plan seems to have potential, for sure.  What level of experience/background do you have?  Just seeing if what you're planning matches that 

One big thing right off the bat that I'd suggest is that you plan to BREADBOARD all this stuff, rather than "prototype" (unless that's what you mean?)  If you do that, you can get a Rat going, then put it on a PCB....then try out filter options at your leisure....

[XAXAU]

I think you'd get some good answers if you began to draw out your idea, schematic-wise, X.   It might help you to keep switching and stuff straight.

Your plan seems to have potential, for sure.  What level of experience/background do you have?  Just seeing if what you're planning matches that 

One big thing right off the bat that I'd suggest is that you plan to BREADBOARD all this stuff, rather than "prototype" (unless that's what you mean?)  If you do that, you can get a Rat going, then put it on a PCB....then try out filter options at your leisure....

Hey Man! My skills definitely does not match what I´m planning to build but I think I can pull it off!  I used to be an electrician 15 years ago and soldered in school. Been reading up on these kind of circuits for the last year and I´ve studied all documents I could find on the Rat! I recently touched up on my soldering "skills" so I know I can put it together.

It´s getting a bit late here but I´m gonna draw up a schematic as well as I can tomorrow!

[XAXAU]

Couldn't sleep so I drew up this:

[ElectricDruid]

You won't hear any difference between 1N914 and 1n4148 diodes in a clipper circuit. Well, in *any* circuit really. They're interchangeable.

The other clipping options will provide more variations, but you'll have widely different levels between them, so you'll get volume jumps when you switch between them, unless you compensate this somehow.

I second the "breadboard it" comment. The Rat is simple enough that that's feasible, and it'll let you try out a ton of ideas and select options which are worth including rather than just chucking everything and the kitchen sink at it.

Tom

[XAXAU]

You won't hear any difference between 1N914 and 1n4148 diodes in a clipper circuit. Well, in *any* circuit really. They're interchangeable.

The other clipping options will provide more variations, but you'll have widely different levels between them, so you'll get volume jumps when you switch between them, unless you compensate this somehow.

I second the "breadboard it" comment. The Rat is simple enough that that's feasible, and it'll let you try out a ton of ideas and select options which are worth including rather than just chucking everything and the kitchen sink at it.

Tom

Thanks Tom!

Well regarding the similar sounding diodes I didn´t know that they were interchangeable. If you have any more suggestions for interesting clipping sounds other that what I listed please feel free to let me know.

I might go the breadboard route. It seems to be the sensible way to go! 

I´m not that worried enough about the different volumes to do anything about it. If I designed a new PCB then sure I´d make sure everything was nice and even but this is just gonna be a toy to molest my studio gear with! 

[GibsonGM]

It's not a bad idea or anything.  Another thing you could try is to make a basic opamp gain stage, just copy something like the Distortion Plus using a TL071, say.   Then you can play with clipping elements (diodes...) all day and night, and NOT be overdriving the OPAMP, which I believe the Rat does do! 

The sound of opamp clipping will affect what you're getting with the diodes, plus the filtering.    No reason you can't do the rat front end 'offboard' like I mentioned, and your other things on breadboard, and then plug in just a gain stage after you try the Rat!!   Options...


I did see your string of caps at top....why?  I'd rather see you switch in/out resistors.  Easier to calculate what's going on....the value of 5 caps in series requires math:   1/C1 + 1/C2 +....1/Cn     OR, you could switch in individual caps with a rotary...just thinking, it all works if done right.

For your HPF, you can just use F = 1/2pi R*C, which, to find what cap to use for your desired cutoff frequency, will rearrange to     

C = 1/2pi F * R    C in uF, R in megohms, F in megahertz

Ex: if you want 1kHz cutoff ,cap = 1/6.28(1k)(.001) = .16uF  (150n would be close enough)

Or go here and use the calculator     http://www.muzique.com/schem/filter.htm

[XAXAU]

It's not a bad idea or anything.  Another thing you could try is to make a basic opamp gain stage, just copy something like the Distortion Plus using a TL071, say.   Then you can play with clipping elements (diodes...) all day and night, and NOT be overdriving the OPAMP, which I believe the Rat does do! 

The sound of opamp clipping will affect what you're getting with the diodes, plus the filtering.    No reason you can't do the rat front end 'offboard' like I mentioned, and your other things on breadboard, and then plug in just a gain stage after you try the Rat!!   Options...


I did see your string of caps at top....why?  I'd rather see you switch in/out resistors.  Easier to calculate what's going on....the value of 5 caps in series requires math:   1/C1 + 1/C2 +....1/Cn     OR, you could switch in individual caps with a rotary...just thinking, it all works if done right.

For your HPF, you can just use F = 1/2pi R*C, which, to find what cap to use for your desired cutoff frequency, will rearrange to     

C = 1/2pi F * R    C in uF, R in megohms, F in megahertz

Ex: if you want 1kHz cutoff ,cap = 1/6.28(1k)(.001) = .16uF  (150n would be close enough)

Or go here and use the calculator     http://www.muzique.com/schem/filter.htm

I´ve also heard that the Rat overdrives the opamp and that´s why there´s a mod called Ruetz which lowers the drive. So basically I´m just gonna test different resistors until I find something that sounds like a mild overdrive with the "distortion" knob counter clockwise and switch between that and stock value.

If I switch resistors with the rotary then I mess with the drive settings. But I see now after using an online calculator for capacitance that it´s not gonna work the way I want it to if the capacitors are in series and I use the same value for each step on the rotary. It could work if I do some math. Or switch in individual caps like you suggested would be easier I guess! 

[antonis]

Although is a matter of taste, you better limit your clipping options to a maximum of 3 with "enough" forward voltage drop difference between them..
(i.e. Ge or Schottky - 2 Si in series - Red Led..)

You can leave Leds (or whatever pair have the highest voltage drop) permanently connected and select between the others with a SPDT switch but, as ElectricDruid told you, you'll have to compensate between different output levels - if you have to make changes during playing.. 

[GGBB]

I´ve also heard that the Rat overdrives the opamp and that´s why there´s a mod called Ruetz which lowers the drive. So basically I´m just gonna test different resistors until I find something that sounds like a mild overdrive with the "distortion" knob counter clockwise and switch between that and stock value.

If I switch resistors with the rotary then I mess with the drive settings. But I see now after using an online calculator for capacitance that it´s not gonna work the way I want it to if the capacitors are in series and I use the same value for each step on the rotary. It could work if I do some math. Or switch in individual caps like you suggested would be easier I guess! 

The characteristic sound of the RAT comes from overdriving the op-amp along with the high pass filtering. Mess with those things too much, and you end up with something that doesn't really sound like a RAT. If that's your goal then bob's your uncle. Otherwise, the point of the reutz mod was to give the RAT more bass at low gain settings. But it messes with the sound a lot, so not everyone likes it. To get more bass without affecting the tonal character, you need to change the high pass filter so that gain is increased for low frequencies (this is done by reducing the 560R resistor) or decreased for higher frequencies (reutz). With reutz, raising 47R gives "more bass" by lowering gain for the high frequencies. But that changes character a lot because change the R affects the knee frequency of the filter. And that's the problem - when you change either the cap or the resistor in an RC filter, you change the knee frequency. The solution is to change both. So raise 47R and reduce 2u2, or lower 560R and raise 4u7. Option 1 means both resistors and caps in series, option 2 means parallel, which means you'll need double pole switches. I like changing the 560R /4u7 combo and I came up with a simple mod to do this. It gives two "bass boost" settings using a center-off DPDT toggle. Someone name Marcus did a nice blog on it here: https://marcuseffects.wordpress.com/2016/06/20/proco-rat-bass-mod/. My original schematic with three variations is below (click to enlarge). A couple of notes - you won't be able get too much variation in bass - two settings is about all there is as finer increments aren't that discernible. Also note that as you turn the gain up, bass response goes up naturally, so the BB mod only works at lower gain settings.

[PRR]

Strictly: 1N914 is an old spec with very wide limits, 1N4148 is a newer tighter spec. Neither gives a meaningful spec for Forward Voltage, which is key for a distortion. (The 1V spec is clearly a reject spec, and I would not expect to find one in 1,000 anywhere near there.)

If you do a very critical detector which specifies 1N4148, you may have trouble with the 1N914 you bought back in 1964. Today a "1N914" is a 1N4148 re-marked for customers who think it matters (1N4148 will work in ALL 1N914 jobs).

If you are just distorting, use "any" diode, of same or different material. Ge will clip lower than Si. Any transistor, even half-dead, contains a diode. Transistor diode is doped somewhat different than diode diode, and while I doubt this matters for clipping it is worth a try.

[XAXAU]

The characteristic sound of the RAT comes from overdriving the op-amp along with the high pass filtering. Mess with those things too much, and you end up with something that doesn't really sound like a RAT. If that's your goal then bob's your uncle. Otherwise, the point of the reutz mod was to give the RAT more bass at low gain settings. But it messes with the sound a lot, so not everyone likes it. To get more bass without affecting the tonal character, you need to change the high pass filter so that gain is increased for low frequencies (this is done by reducing the 560R resistor) or decreased for higher frequencies (reutz). With reutz, raising 47R gives "more bass" by lowering gain for the high frequencies. But that changes character a lot because change the R affects the knee frequency of the filter. And that's the problem - when you change either the cap or the resistor in an RC filter, you change the knee frequency. The solution is to change both. So raise 47R and reduce 2u2, or lower 560R and raise 4u7. Option 1 means both resistors and caps in series, option 2 means parallel, which means you'll need double pole switches. I like changing the 560R /4u7 combo and I came up with a simple mod to do this. It gives two "bass boost" settings using a center-off DPDT toggle. Someone name Marcus did a nice blog on it here: https://marcuseffects.wordpress.com/2016/06/20/proco-rat-bass-mod/. My original schematic with three variations is below (click to enlarge). A couple of notes - you won't be able get too much variation in bass - two settings is about all there is as finer increments aren't that discernible. Also note that as you turn the gain up, bass response goes up naturally, so the BB mod only works at lower gain settings.

Hey GGBB thanks for the input however I´m more lost that ever now to be honest! 

First off, what makes the tonal character of the Rat? Is it the amount of drive set by the two R´s in parallel in the FB loop? 47R & 560R in parallel = 43R or so. From reading this: http://www.electrosmash.com/proco-rat I learned that there are 2x highpass filters (47R & 2.2uF = 1540Hz and 560R & 4.7pF = 60.5Hz) and together with the opamps 30pF (pin 1 & they constitute the Rats tone.

Now forgive my noobness but I was under the impression that a filter made up of 560R/4.7pF will sound exactly the same as 5600R/0.47pF?

This is how I came up with being able to switch between HI GAIN/LO GAIN and BASSY/STOCK HPF with a SPDT & a DPDT:

Strictly: 1N914 is an old spec with very wide limits, 1N4148 is a newer tighter spec. Neither gives a meaningful spec for Forward Voltage, which is key for a distortion. (The 1V spec is clearly a reject spec, and I would not expect to find one in 1,000 anywhere near there.)

If you do a very critical detector which specifies 1N4148, you may have trouble with the 1N914 you bought back in 1964. Today a "1N914" is a 1N4148 re-marked for customers who think it matters (1N4148 will work in ALL 1N914 jobs).

If you are just distorting, use "any" diode, of same or different material. Ge will clip lower than Si. Any transistor, even half-dead, contains a diode. Transistor diode is doped somewhat different than diode diode, and while I doubt this matters for clipping it is worth a try.

Thanks PRR! I´m just a noob trying to build a box with nuances of clipping. I do lots of clipping ITB and play around with curves and soft clipping and it´s really useful (for me at least). I watched some videos with Brian Wampler breadboarding and testing lots of different clipping variations and it feels like it´s got to be something there and not just smoke and mirrors? Or do they have their own character but are too similar to justify having more of them to switch around between?

What about diode connected mosfets then? Aren´t they supposed to have a very wide knee resulting in relatively soft clipping?

[ashcat_lt]

An "assymetrical" set of diodes won't actually clip assymetrically for any significant period of time because the caps around the diodes will float toward the middle of the diode drops and take the audio "ground" with it.  There might be some asymmetry on the leading edge of the first note you play, but then it pretty quickly settles back to plain symmetrical clipping.  There isn't a particularly easy way to "fix" that in a Rat.  You may hear a difference due to the difference in overall output, but it won't actually be assymetrical.

Also, an important part of the Rat is that it has a whole buttload of gain in the passband.  If you want to hit the diodes as hard as a Rat does, you can't help but clip the opamp no matter what opamp you use.  You would need very very high voltage power supply to have a chance, and you'll never find an opamp that can handle that.

[GGBB]

First off, what makes the tonal character of the Rat? Is it the amount of drive set by the two R´s in parallel in the FB loop? 47R & 560R in parallel = 43R or so. From reading this: http://www.electrosmash.com/proco-rat I learned that there are 2x highpass filters (47R & 2.2uF = 1540Hz and 560R & 4.7pF = 60.5Hz) and together with the opamps 30pF (pin 1 & they constitute the Rats tone.

As with probably nearly all distortion/overdrive circuits, a great deal of the circuit's tonal character comes from tone shaping. Tone shaping is done in the RAT with those two high-pass filters as well as the low pass filter formed with the 100pF cap that parallels the gain pot. The 30pF cap primarily affects distortion by changing the slew rate of the op-amp which affects bandwidth leading to distortion at the extremes. But it's bandwidth so that technically is tone shaping in away, but I'm not sure if the bandwidth limit creeps down into audio range - I doubt it. The most bang for the buck by far are the high pass filters. The 100pF cap would have to be increased somewhat to be audible - this softens the RAT sound - a really great way to take the edge off. The 30pF cap isn't one I've bothered to play with, but I've gathered from other sources that it affects mushiness a bit - if that makes any sense.

The other part of the tonal character equation is the distortion itself. The RAT is very simple diode hard clipping - used in about a billion different dirt pedals. But as legend has it, the LM308 along with it's low slew rate contributes to the character by way of being heavily overdriven by the RATs ridiculously high gain amount. However, mots people don't run their RATs at max distortion (far less in fact - even at 8/10 with the audio taper pot the gain is well below half of max - 5/10 is somewhere around 10-20%), so this factor may be over stated a little, although the gain is still high even at 5/10.

Now forgive my noobness but I was under the impression that a filter made up of 560R/4.7pF will sound exactly the same as 5600R/0.47pF?

No forgiveness is required - questions are completely allowed!

They will have the same knee frequency (when in isolation). Whether or not they "sound the same" depends on what else is involved in the rest of the circuit, which in this case is the other 'leg' of the filter as well as the gain pot. If your objective is to have switchable gain settings, just switch in a resistor in parallel with the gain pot to drop the gain (experiment with value - I'd start with 47k). That is exactly how the gain control works anyway, so this is going to have zero impact on the tonal character of the RAT - it will sound just like adjusting the gain pot. Then use the BB mod just for bass boost.

[XAXAU]

Thanks everyone!

If the asymmetrical clipping is in the feedback loop of the Rat, will it work then?

Maybe I'm better off using s screamer instead if I wanna do asymmetrical clipping?

[ashcat_lt]

If the asymmetrical clipping is in the feedback loop of the Rat, will it work then?

Maybe I'm better off using s screamer instead if I wanna do asymmetrical clipping?

I think it can work in the feedback loop.  Whether you're better off...  Well it's a very different sound.  It's very important to remember that the feedback "soft clipping" arrangement is very much like mixing the hard clipped signal with the clean signal, and the proportion of that mix matters, and the gain knob changes that proportion.  In fact, very large amounts of gain can tend to cause a TS-style distortion to start to "clean up" as the wave blows through the diode drop really quickly near the zero crossings and then spends most of its time at unity.  We've already said that this thing has a lot of gain.  It's a lot more than a TS gives, and I think there's an important reason for that.  It's worth a try, but just don't be too surprised.

A couple things to expand upon GGBB's post above -

The filters in the feedback loop are shelving filters.  The R value determines the "shelf level" along with the cutoff frequency.  Well, it's bizarro world, so I guess we should say those Rs set the gain in the passband.  Changing the R and C in a complimentary fashion will keep the corner frequency the same, but change the overall gain.

Slew rate is not just about bandwidth, it's about Gain Bandwidth Product.  It basically means that it will "clip" higher frequencies to a lower ceiling.  And yes, at high settings of the gain knob, it actually affects most of the audio band.  It's very much like an LPF with a cutoff down below 100Hz, only with a bit of extra distortion thrown in.

[XAXAU]

If the asymmetrical clipping is in the feedback loop of the Rat, will it work then?

Maybe I'm better off using s screamer instead if I wanna do asymmetrical clipping?

I think it can work in the feedback loop.  Whether you're better off...  Well it's a very different sound.  It's very important to remember that the feedback "soft clipping" arrangement is very much like mixing the hard clipped signal with the clean signal, and the proportion of that mix matters, and the gain knob changes that proportion.  In fact, very large amounts of gain can tend to cause a TS-style distortion to start to "clean up" as the wave blows through the diode drop really quickly near the zero crossings and then spends most of its time at unity.  We've already said that this thing has a lot of gain.  It's a lot more than a TS gives, and I think there's an important reason for that.  It's worth a try, but just don't be too surprised.

A couple things to expand upon GGBB's post above -

The filters in the feedback loop are shelving filters.  The R value determines the "shelf level" along with the cutoff frequency.  Well, it's bizarro world, so I guess we should say those Rs set the gain in the passband.  Changing the R and C in a complimentary fashion will keep the corner frequency the same, but change the overall gain.

Slew rate is not just about bandwidth, it's about Gain Bandwidth Product.  It basically means that it will "clip" higher frequencies to a lower ceiling.  And yes, at high settings of the gain knob, it actually affects most of the audio band.  It's very much like an LPF with a cutoff down below 100Hz, only with a bit of extra distortion thrown in.

Thanks man that cleared up a few things for me. When you say shelving filter, do you mean that it has a s-curve to it if you look at it on a frequency plot?

The reason I´m going on and on about being able to switch to a lower gain setting is that I´m gonna use the pedal on synths and although I use distortion in one form or another on just about anything I produce and I´m not sure I´d like it to maul everything. Get another pedal for mild overdrive you say? Well I could get a TS for but it´s also about experimenting and putting a twist on it. Yes I can cook a traditional italian bolognese, but have you ever tried one with caramelised red onions, black garlic, paté and chewing tobacco extract?  That´s the kind of chef I am!

So if I lower the gain in the Rat, could it sound like an overdrive pedal? It´ll lose the Rats character yes but I feel it would be more versatile. Again it´s just for coloring sounds in the studio so I won´t set it for a type of sound and glue the pots stuck! 

[ashcat_lt]

When you say shelving filter, do you mean that it has a s-curve to it if you look at it on a frequency plot?

Yes?  The main point really being that changing the gain changes the filter and changing the filter changes the gain.  I think that if it was me, the "low gain" switch would just drop a resistor parallel to the gain pot.  It'll change the taper a bit, and yes it'll change the overall color, but in the same way that just turning the pot down.  Basically, it gives you more control over the lower range of the existing gain control...sort of...

I suppose it's worth mentioning that big inputs into a TS-style clipper as tend to "clean up" the same way that big gain does.  Really it's gain times input that's the issue.  This matters since you're talking about using this for things with potentially much hotter signals than a typical guitar. 

Most of those things will probably have volume controls of their own also, and no matter what configuration you have the pedal in, those knobs will make a real difference in what comes out.  With a big enough input, you can use less gain at the pedal, which means less relative bass cut.  With smaller inputs, you can use more gain (more bass cut) without smashing it quite so hard.


Now all that said, I personally am of the opinion that the differences in the curve of the knee with different diodes is far less important than the different Vf compared to the input.  That is, all things being equal, a 2V signal clipped by silicon diodes is about the same thing as a 1V signal clipped by age/Shottky and a 4V signal clipped by LEDs.  It's about how much gets clipped off more than how rounded (or not) the corners are.  If you want the most versatile clipper you can get, you'd just use one set of clipping diodes, DC couple them to the gain stage (they can't go to ground anymore ), and vary gain and DC offset going into the diodes.  This sort of requires divorcing the filter from the gain stage, and really wants to be DC coupled to the output buffer.  It makes the relatively simple Rat circuit quite a bit more complicated - in fact I don't think it could really be called a Rat anymore - but it gives you all of the important parts of all the diode options along with "in-between" settings that you can't get other ways and reliable assumetry and it will have a more consistent output level since the Vf limit doesn't actually change.  I've been thinking about it for a while, and have mentioned it before, but haven't really bothered to design or build it yet.

[GGBB]

So if I lower the gain in the Rat, could it sound like an overdrive pedal?

Yes - that's what the distortion knob is for . The RAT is often used at low gain settings for bluesy overdriven sounds. I think it actually sounds great that way.