Next up: a Rat build

[ashcat_lt]

Makes absolutely no difference.  The order of components in series is arbitrary.  If there was something else happening in between, it might, but here it's exactly the same either way.

[karbomusic]

Also notice, it's showing as inverting. I thought I remembered that being an old schematic with an error, at least it didn't work when I built it but maybe I made a mistake. I found this one... Which is what I had expected and it worked fine. I think is the corrected version of what you posted:



And the PCB layout I created from ^that schemo which has a ruetz mod trim pot designed in...



It's all coming back to me now, I built the schematic you posted in it's entirety, didn't work, did a small jumper/hack job on the rev1 of my PCB, changing it over to non-inverting which fixed. Then I went back and corrected the schematic and PCB layout after finding the schematic I posted and some post somewhere on the net where Jack mentioned the old erroneous schemo might still be floating around...

[thermionix]

Makes absolutely no difference.

Exactly what I wanted to hear, I had already soldered the 1k in.

Also notice, it's showing as inverting.

Hadn't noticed that.  Pre-corrected just by using the Runt PCB, though.  So far the only area I'm deviating from the Madbean schem is the power supply.  I'm sure it makes no difference, just for mojo or something.

Thanks guys!

[GGBB]

As a secondary guide I've also been referring to this Keen/Orman schematic from the web, I guess it's an actual Rat:

Most likely, errors aside, this is a schematic for the "bud box" RAT, which was essentially the prototype (apparently 12 produced by hand). The notable difference is the JFET input buffer, which was dropped for the production version because, well, what do you need an input buffer for if you are going into an op-amp? Plus it reduces cost. The input impedance of the op-amp was accordingly increased. Never really noticed the probably unintentional humor in the notes about the distortion pot  - 100k would have slightly less gain - like that's going to be a problem.

[thermionix]

I was going to do the input just like the madbean schematic, but I just noticed the differences around the output JFET.  I'm going to use a BF245A and not a 2N5457, do you think it's best to use the "bud bux" setup?  I can mount the 10k across the volume pot since there's no place for it on the Runt PCB (or just use a 10k pot if I have one).  Then just change the source resistor and output cap.  Thoughts?

[GGBB]

do you think it's best to use the "bud bux" setup?

No. Do you want a RAT? The one with the famous sound? Build a RAT - the famous one. The RAT isn't famous for a dozen hand built prototypes, it's famous for the production version (in particular the smallbox whiteface issue from the 80s). The Madbean Runt will sound great.

[thermionix]

That's the thing, I don't know which setup is the "famous" Rat sound.  I thought the 80s Rats used the BF245A, but maybe I have some bad information.  I've never seen an actual ProCo schematic.

Is the Runt the common output circuit, just with a different JFET?

[thermionix]

Okay, I'm looking at these shots of a mid-80s * original, I guess a Rat 2:



Unless I'm mistaken, it looks like the JFETs match the pinout of a 2N5457, the source resistor is 10k, and there is a 10k in parallel with the volume pot (presumably 100k).  Can't tell the value of the output cap.  Maybe there's no audible difference between 1uF and 10uF, I just don't know.

*edit:  it's pot-dated 1990.

[thermionix]

Actually, disregard all the above.  I've looked at more gut shots online and it seems the Runt output is correct for the 80s units.  I'm gonna use the 2N5457 and a 1uF tantalum for the output.  It will sound how it sounds.  I'm no Rat cork-sniffer, I'm not really expecting to like it too much, lol.

[GGBB]

The multi rat schematic is accurate:



There are very few differences between all production RATs. Output FET is 2N5458, if you want to be precise - that's almost the only thing not accurate in the Runt. MPF102 is a great substitute.

[PRR]

> reversed in order

Parts in series, nothing else connected at the joint, can be either order.

[thermionix]

Parts in series, nothing else connected at the joint, can be either order.

Thanks, I knew that was the case for resistors in series, or caps in series, but I didn't know about a resistor and a cap in series.  I tried both ways on a wah input and couldn't hear any difference, so i guess that's why!

The multi rat schematic is accurate:

Dang, I forgot about that multi rat schematic.  And it was pointed out on page 1, duh!  Going by that I'm using 1n for C10, 47u for C2, 1N4001 for D1, and 33K for R2 and R3.  I doubt any of that makes much difference.  Maybe C2 a little, but I don't know.

Well, it works!  I don't have the box drilled yet, but I hooked it up with my test harness (jacks and battery clip).  Sounds pretty good!  Much smoother than I remember any Rat sounding, but I haven't played one in years.  Decades probably.  Definitely no shortage of gain, lol!

[thermionix]

Oh, also I don't know what type of clipping diodes I used.  It's an unmarked pair that came from a reissue TS808.  The short-body ones with a yellow stripe.  Maybe those are 1N914s, but I really have no idea.

[GGBB]

Nice work. None of your/madbean's component changes are significant including C2. The diodes are probably silicon, so once again not likely to matter.

[ashcat_lt]

Wow I didn't notice in that first RG scheme about the in buffer or "inverting" opamp stage.  I just looked at the C and R that were questioned.  What I'm seeing in that scheme, though, is that both inputs on the opamp are marked "-", though the pin numbers do suggest connecting the input to 2, which would be inverting, but then you'd have all positive feedback.  I'm not completely sure what that'll do, but if I was to guess I'd think it would just go to one of the rails and stay there. 

I just remembered that my Rat has like 180K when the gain pot is all the way up.  I cut the track at the high end (like a no-load pot) and then put a 180K resistor between lugs 1 and 3.  For most of the sweep it's the same as a 100K pot with a sort of boost switch action at the very top.  The difference is surprisingly subtle since there's already so much gain to begin with.

  It also has two "assymetric" settings - one is a single Si one way and two the other, and the other is one Si and mother on the other side - which is how I initially figured out that it can't actually do assymentry.  Even that "way assymetrical" setting just isn't.

[thermionix]

The diodes are probably silicon

Oh yeah, definitely some common small signal Si.  They look like these:



No idea what specific type, but whatever Ibanez was putting in 808 reissues five or ten years ago.  Seem to sound fine, so I'm leaving them in for now.

[karbomusic]

I'm not completely sure what that'll do, but if I was to guess I'd think it would just go to one of the rails and stay there. 

I can attest it didn't work that's for sure. Since it had RG and Jack's name on it, I didn't breadboard it first like I usually do. It may have done what you suggest but I don't remember the exact symptom at the time. Took me a bit post build to find the problem but was easy enough to hack and reverse pins 2 and 3 on the PCB. I only posted because it bit me and that wrong schematic seems to always come up first based on how you search for "RAT schematic".

I have to wonder how many have built it, failed, and never knew it was actually erroneous. Either way, the correct schematic sounds just grand, it's a mainstay on my pedal board.

[thermionix]

Well, it's boxed up and rockin'.



The PCB isn't even mounted.  It's so small and light I just let it hang on the pot leads, and it's held down on the bottom by the one of the LED legs.  Doesn't flop around, easy to remove for mods or repairs.

I have no idea if my JFET is biased properly and I don't know how to check.  I only had one 5457 and I soldered it in.  I think it's running fine.

Well, I gotta say the biggest surprise is the clean tone.  It's a monster!  With the volume dimed, and the gain just barely above zero, overall volume is about unity with bypass.  I'm going into my blackface type amp, low volume and clean, turn on the Rat, and it sounds more like an old Marshall.  Adds a bit of pleasing compression.  Apartment living prevents me from doing a cranked check, but I bet it would be quite JTM45-ish.

I'm digging it.  Back to playing it now.

Thanks for the help folks!

[thermionix]

Gotta get knobs for it.  The box is Kentucky blue, so I'm thinking white knobs.  Seems the Alpha pots I got from Tayda have 6mm shafts, instead of 1/4" - 6.3mm whatever.  Most of the knobs I have fit loosely and tilt some when the sets screws are tightened.  Can be shimmed but I'll try to find 6mm knobs that I like.

I inadvertently wired the Filter as a Tone so it gets brighter clockwise.  That's fine with me, but the taper is now opposite what it should be.  No big deal really, but I will probably correct it, just swap two wires.

[bluebunny]

Can be shimmed

Heatshrink.