Lab RAT - an unorthadox take on a classic

[vigilante397]

I've been going through a tube overdrive fetish lately and have made some pretty cool things that made me pretty excited, but then a friend asked me to build him a Rat, and I was reminded that no matter how much I like my tube pedals the Rat will always be my favorite dirt pedal.

Then I woke up yesterday with this thought in my head: why not take my favorite parts of the Rat and incorporate a tube section? So that's what I did.

I started with the stock Rat gainstage with the LM308, but I changed the Gain pot to an A50k because I almost never go past the halfway point anyway. Then I took out the infamous Filter control and put in a Baxandall/James two-band EQ. I left it with the stock 1N4148 diodes for clipping because I've tried a million different diodes on switches but always end up sticking with the 1N4148's.

Finally I removed the output buffer and stuck in a 12AX7 buffer in its place. I'm running an L7806 regulator with a diode between middle pin and ground to get me damn close to 6.3V DC for the heaters. Then I'm running a nixie power supply I bought as a kit from eBay that takes 9V DC in and spits out 180V DC, which I run to the plates (after some thorough filtering to get rid of the whine). I'm using the first triode for the output buffer, then I have the second triode set up as a switchable boost, because I like having my drive and my boost in the same box.

So anyway, here's the schematic:



It's still just on the breadboard right now, though I do have an unverified layout for it I'm hoping to try out soon. The board is too big for my cheap CNC machine to stay calibrated, so I'll have to see if I can use one at work on my lunch break Monday. It's a little too big for a 1590BB, but I have a nice big slanted enclosure I've been waiting to use for somehing awesome that should fit this nicely

The really big reason I'm excited about this pedal is that I've designed and built a dozen different overdrives and none of them sounded bad, but none of them earned a place on my pedalboard. This one, however, is the best sounding overdrive I have ever heard. It's obviously not going to be the sound everyone is looking for, but it is definitely the sound that I want. I hope this is at the very least educational for someone, and maybe tomorrow when my kids aren't asleep I can get a quick demo recorded.

[rankot]

No need to use LM7806, just use appropriate resistor to drop voltage from 9 to 6.3V. Much simpler and less space!

[Cozybuilder]

Nice one Nate. I knew that 12AX7 PCB conversion board would come in handy!

[Plexi]

I would say this is brilliant.
To me, the Rat is a 'exquisite' distortion (used about 15 years a Deucetone as main distortion).

The Rat's trick is in the Jfet recovery stage: the perfect "tube like" output stage.
Which better now than have a real tube there?

[vigilante397]

No need to use LM7806, just use appropriate resistor to drop voltage from 9 to 6.3V. Much simpler and less space!

A 12AX7 (or anything in the 12A*7) family draws 300mA @ 6.3V, so dropping 3V across a resistor means dissipating about 1W across the resistor, whichis going to make for a HOT resistor. This means I would need at least a 2W resistor, which are pretty big and will still be getting warm. Voltage dividers are great for biasing and things like that where the ratio is more important than the actual voltage, but for tube heaters the actual voltage is a big deal. So the regulator/diode combo is what I want

[cnspedalbuilder]

Awesome idea! Since you have it on breadboard, have you tried it with the diodes lifted and the Ruetz mod?

My Aion Rat is my favorite distortion of all time. It has a CA3130EZ instead of the LM308 and I usually have it with diodes lifted and Sweep pot at 3:00. Somehow that combo made it the roaring voice of hell. I can only imagine how it would sound with a 12AX7!

[vigilante397]

Awesome idea! Since you have it on breadboard, have you tried it with the diodes lifted and the Ruetz mod?

I tried the Ruetz mod and I liked it, but not enough to keep it. Obviously everyone's ears are different, and on the Rat I have on my board with the Ruetz mod wired in I usually just keep it adjusted so it sounds stock anyway

As for the diode lift, it was a boost in gain (obviously) so it let the tube clip a little more, but I love the sound of the 1N4148's for clipping so I put them back

[vigilante397]

Got a quick video thrown together to show some tones. Try not to judge the playing too harshly, we can't all be Rob or Jimi

[Plexi]

Yeahh!!!
Sweet LP Studio there... 

[vigilante397]

Yeahh!!!
Sweet LP Studio there... 

Thanks it's not my most expensive guitar anymore, but I still love it and plan to keep it forever

[PRR]

> a HOT resistor

SAME power (actually a hair more) in the regulator.

And heater voltage is not *that* critical. 12AX7 have been run 5.5V to 7V.

Since he is surely using a wall-wart, and probably one with good regulation, I'd expect a resistor to work fine. Maybe not cheaper. The resistor alone is smaller, but the same-heat means you need about the same air-space around it, so not really smaller.

There is a theory that tube heaters should be soft-started. Actually heater failure is very very rare. Also the UNIVAC techs with 17,000 tubes did have heater trouble, tested slow-start, and did not find much difference. However if you favor slow-start, the resistor is a little slower (due to heater cold-resistance) than a 1A regulator.

[PRR]

9AQ8 is a twin-triode, good Mu, 9V 0.3A heater. Change cathode resistor to 2.2K-3.3K, it will bias OK and give some less gain than 12AX7 (which is way more than you need/want). $6 at major tube depots.

[vigilante397]

9AQ8 is a twin-triode, good Mu, 9V 0.3A heater. Change cathode resistor to 2.2K-3.3K, it will bias OK and give some less gain than 12AX7 (which is way more than you need/want). $6 at major tube depots.

Good to know, thanks I'll have to look into it, that would save a bit of trouble!

[italianguy63]

Yeahh!!!
Sweet LP Studio there... 

Thanks it's not my most expensive guitar anymore, but I still love it and plan to keep it forever

I remember when you built it! 

[Perrow]

However if you favor slow-start, the resistor is a little slower (due to heater cold-resistance) than a 1A regulator.

Just asking out of interest; would an LM317 wired for constant current rather than voltage be "kinder" to the heater?

[diydave]

Is there a mistake in the schematic?
Because V1A is not a buffer, it's a gain-stage.
To use V1A as buffer, connect plate to B+, and put 47k resister between cathode and ground.
Take the signal from the cathode using the 22n blocking-cap.

But a gain stage is probably the better thing to use here. Since you already have rather small signal (+- 0.6 volts max signal-range from the diodes) which is heavily loaded by the baxandall-tonestack.

[vigilante397]

Is there a mistake in the schematic?
Because V1A is not a buffer, it's a gain-stage.

You're correct, I'm not really using it as a buffer, more as a recovery gainstage after the EQ. I tried it just as a buffer and wasn't thrilled with the tone.

[PRR]

> constant current rather than voltage be "kinder" to the heater?

me: "...the UNIVAC techs ..tested slow-start, and did not find much difference."

I don't coddle my tube heaters. Last bad heater I saw had been chucked 10 feet into a dumpster, no wonder.

[vigilante397]

I don't coddle my tube heaters. Last bad heater I saw had been chucked 10 feet into a dumpster, no wonder.

I honestly don't know that I've ever seen a bad heater. Usually either loss of vacuum (bad seal) or an internal short killed all the tubes I've seen.

Just ordered a 9AQ8 to play with by the way, thanks again for the recommendation.

[Perrow]

> constant current rather than voltage be "kinder" to the heater?

me: "...the UNIVAC techs ..tested slow-start, and did not find much difference."

I don't coddle my tube heaters. Last bad heater I saw had been chucked 10 feet into a dumpster, no wonder.

LOL. Ok, fair point. But it would be kinder to the power supply, wouldn't it?