New at runoffgroove.com: Eighteen

[Alpha579]

Hey,
I want to to a fet emulation of a Ampeg J-12, and ive got the schematic:
http://www1.korksoft.com/~schem/ampegamps/j12b_jet.pdf
Couple of questions:
1)How do you mix the output of the anodes of the 6v6 tubes?
2)Is there any dc on the anodes of the 6v6 tubes?
3)would you make the 120K resistors from the plates of both halves of V2 pots for tuning bias?
4)can i just completely remove the tremolo? and if i wanted to, could i emulate the tremolo section with a mpf120?

sorry for sounding like a newb, but im not great with push/pull stages and tremolo sections...

Alex

[Gary]

Hey,
I want to to a fet emulation of a Ampeg J-12, and ive got the schematic:

Ok, here's what you do:
Choose one signal path (channel) and trace it out.  Let's do the top one.  The first stage is pretty straightforward, so just use the values as you see there, but use a trimmer for the drain of the fet.  Use the exact same tone and volume values, too.  On the second stage's source, you'll have the 220 Ohm in series with the 1k/100uF pair.  Again, a trimmer on the drain.  From the drain, you'll see the 470k and associated path.  This appears to be some sort of local feedback path to the other half of the 6SL7.  Quite honestly, I'd say you can omit that, but you could also try going from the second stage drain with a 470k, through the 0.02 cap, on to the 300p/470k pair to ground.  Now, back to the main signal path.  You'll have an 0.02 cap from the second drain going to a 1k, then on to the last stage's gate.  You'll want to stick the 270k biasing resistor from between the 0.02 cap and the 1k and tie the other end of the 270k to ground.  The last fet's source will have a 250 Ohm/20uF pair to ground.  What type of speaker was in this amp???  Find a response graph and setup your coupling cap from the drain to act as a high pass filter to help set the bass range of your speaker.  Add a low pass filter after this to set the treble response for the speaker.  You'll find one isn't enough to roll off the treble at the proper rate.  In most cases, it gets closer to double the LPF.

Lastly, let's not forget that feedback path from the output transformer to pin three of the second stage.  Note how the resistor close to the tranny say 5-10k.  Pick a value.  OK, but there is one more problem.  You don't want to feed DC back to the second source, so you'll probably want a blocking cap.  Start with a cap value of your choice and tinker until you find a value that sounds best.  You will find some values cause resonant sounds that can be very cool.  Ofcourse, you'll want to decouple the power supply with something generic like a 100uF cap to stop possible oscillations.

That's pretty much how you do it.  It's ham-fisted, but it works in most cases.  If you want tremolo, why anchor this circuit with a built in trem?  Just use a separate EA trem before or after.

BTW, I did try a 5150 (minus reverb), but it wasn't worth the 9 or so fets.  Now that all of you have a rough idea how we have been doing this, feel free to try some amps of your own.

Best of luck!

[Lonestarjohnny]

Gary, would it be possible to do a 440 Magnatone with varistors on the tail end of the circuit for the Magnatone Vibrato ? the 440 is one incredible sounding amp but also very hard on the saving's account,
JD

[Doug H]

 So there you have it, I reported wide variances in pedal performances testing them in different amps. I never push them on amps that don't like it...I just try some settings and if it really just sounds like crud o' I switch them off, or pull and bag them...

I have found a graphic EQ can do wonders for helping a pedal get along with a particular amp. For example, if I want to set up my Fatness clean and bright and switch a fuzz on and off, the fuzz sound can get kind of shrill. Using the EQ to roll off some of the highs really smooths it out and makes it work well. Problem is, then I have to switch 2 pedals on/off and the EQ is a tone-sucker in bypass (is that stupid or what??). So I need to build some bypass loops, but you get the idea....

Doug

[javacody]

Thanks for that walk through Gary. That was very educational. So the 5150 didn't sound too good?  

I would think that the simpler amps sound better with this setup?

[B Tremblay]

I would think that the simpler amps sound better with this setup?

Precisely.  We've found that once you're using more than four FETs, things begin to unravel a bit.

[BDuguay]

Mr. Tremblay

I wondered if you would expand on the idea of these emulation citcuits and the amps they're played through.
For example, I have a J.M.I. AC30 and I'm interested in the Matchless EF86 emulator version. I suspect this circuit would sound more authentic through the 30 than, say, a Twin or other class B amps.  Do know what I mean?
Just curious.
B.

[B Tremblay]

I wondered if you would expand on the idea of these emulation citcuits and the amps they're played through.

I've been using a Ampeg Rocket II, set for clean with moderate bass and treble, for the majority of development and testing of the emulators.  I've tried them with my Musicmaster Bass Amp as well, but it has a narrower response than the Ampeg.  I've had good results using the circuits with a cab sim when recording with the computer.

I believe Gary has been using a solid-state Randall, set for clean, with the tone controls at maximum.

Does that help?

[Gary]

I believe Gary has been using a solid-state Randall, set for clean, with the tone controls at maximum.

That's true.  The Randall is set absolutely clean.  You can hear it at the beginning of the clips I did.

To answer an earlier question, I did try running things like a Fuzz Face before some of the circuits.  It does behave much like you'd think it should.  A treble booster driving one of these sounds pretty good, too.

[Doug H]

Precisely.  We've found that once you're using more than four FETs, things begin to unravel a bit.

Once you start stringing too many JFETs together at 9v you're going to start running into serious headroom problems.

Another way of looking at it:

For the most part, the point of this exercise is to cop some of the character of these vintage (or vintage-style) amps.  These amps are pretty simple, but their sound depends on the speaker, neg. feedback (or not), output stage, phase inverter, etc- as well as the typical gain stages and tone shaping in the preamp.  To me, that's the advantage of this approach. You can add these other circuit elements into the mix to get closer to "the sound" you are trying to cop.

Most of the modern hi-gain amps such as 5150, SLO, etc are preamp distorters. That's not to say the output sections and NFB don't influence the sound, but for the most part, the preamp is where the action is with these kinds of amps. I think you reach a point of diminishing returns trying to model output stages, NFB, etc of an amp like this with JFETs. I experimented with a JFET model of an SLO on the breadboard and it sounded pretty good wiith just the preamp and "speaker sim".

None of this stuff will ever pass a double-blind A/B test with the real amp. The point is just to capture some of the general "vibe". I think you will find with the modern hi-gain stuff, that modeling the preamp/tone-stack/speaker alone are "good enough".

Doug

[spongebob]

First, a big "Thank you!" to the runoffgroove crew!  

And now my question, in the bottom-left corner of the schematic it says:

"bias Q2 drain to 7.0V"

If I would use the circuit with a 15V power supply, how would I have to bias Q2? Something around 11V? Why is this one not V/2 as usual?

I'm just asking because I'm considering giving this circuit a try as a preamp for my DIY solid-state amp (TDA2050 chip amp), which has the preamp section running at +/- 15V (regulated).

[petemoore]

Just follow the ratio...I don't have my calculator handy..but
 7/9  + ?/15ths...would be the equation
 My calculations say ~12.5v of 15 = 7/9ths

[Gary]

I think you will find with the modern hi-gain stuff, that modeling the preamp/tone-stack/speaker alone are "good enough".

You're right, Doug.

[Gary]

Why is this one not V/2 as usual?

We tinkered around with scaling the voltages supplied to the tubes to see if it would translate to jfets.  In this case and a few others, it did.  You can always tune to 1/2 supply V.  It may change the sound.

You could always step your 15V down to 9V for the preamp section.  You'll find raising the supply voltage will change the sound, just as in the case above.  You may like the results better than the stock circuit, but please try it before building.

[Ben N]

Problem is, then I have to switch 2 pedals on/off and the EQ is a tone-sucker in bypass (is that stupid or what??).
Doug

The answer is pretty simple, really.  Just like the old Marshall Guv'nor, or some of Joe Gagan's pedals:  build a loop into the pedal with shorting jacks, so that the eq is switched together with the effect.  Should be easy enough to do in a BB or bigger size box.  As an added benefit, by using the send and not the receive, your pedal can also switch between amp channels or what have you.
Or, like David Gilmour, have Cornish build you a system that switches a dedicated eq for each distortion  :lol:
Ben

[javacody]

Hey Brian or Gary, the only problem with these designs are the number of friggen trim pots I have to smoke... er use.  Can I combine the trimpot of q1 and q3 to get 4.5 for both of them, or is it dependent upon each JFET? Furthermore, if I can get 4.5v off of one trimpot, is there someway I can also get 7v off the same trimpot?

[Alpha579]

Thanx Mr Tremblay and Mr Gary. Great explanation!
Just one quick question,
For other amps, what would you connect the Screen Grid and suppressor grid to on the fet?
Thanx for all your help,
Alex

[Phorhas]

Gary...B... guy, I think U need help! what about your families, your health? tkae a break!

Well, you know I'm kidding, the jod you're doing is amazing!

Thnx...

[Alpha579]

Sorry, just to add to my last post,
http://www1.korksoft.com/~schem/ampegamps/j12b_jet.pdf  Do the suppressor grids (or screen grids, cant tell which) of the 6v6's have signal passing through them, or is it carrying DC to the 270K in the tremolo circuit?

[B Tremblay]

Hey Brian or Gary, the only problem with these designs are the number of friggen trim pots I have to smoke... er use.  Can I combine the trimpot of q1 and q3 to get 4.5 for both of them, or is it dependent upon each JFET? Furthermore, if I can get 4.5v off of one trimpot, is there someway I can also get 7v off the same trimpot?

The trimmer is providing a bias voltage relative to the needs of the FET.  What constitutes 4.5v for one FET may be way off for another.  I've often biased a J201 to 4.5v, removed it to put in an MPF102, and the drain voltage will then measure less than 1.0v

Also, if more than one FET were receiving their bias voltage from the same source, the drains would be connected in such a manner that the circuit could not function properly.

Due to the sometimes huge variation between FETs, the trimmers are your best bet.  You could determine the ncessary resistance and used a fixed resistor, but only if you never planned to experiment with different bias points or part numbers.