Just built a Dr. Boogey (and my perf layout is verified now). Whoa.

[PaulC]

Hey tcobretti,

   Here's a link to a site with some good info about this.

http://www.aikenamps.com/MillerCapacitance.html

This guy really knows his stuff, and he makes some killer amps.  There's some great tech articles he's written on his site that you should check out.  He's got a great way of getting deep while keeping it simple.

  If you like this design then that's all that matters.  There's no reason to try and nail the soldano/boogie.  But if that's really the goal I believe you'd end up with a circuit that looked pretty different.  If I was really wanting to make nailing it a goal I'd start off by building the actually tube preamp first, and then using that as a ref that I could hear/scope out.  If only we had 36 hour days...

PaulC

 

[RDV]

Well fellows, I was so unhappy with my DB I just jumped on that PaulC advice. I used 100pF caps cause I didn't have 150s. I replaced the 22pF I had in the 2nd stage with a 100pF and liked the tone a bit better.

So...

I thought why not do the thing on the 4th Fet as well. I just had to add it on the back of the PCB.

Now I really like it. I can control the tone with the tone controls now. It really seems more amp-like in it's response now.

TRY IT!!

RDV

[RDV]

I just read that Aiken article and it's funny because exactly what you're trying to avoid and or compensate for in tube amp design is what I'm looking to add to the DB so it will react more like the amp in question.

RDV

[PaulC]

just read that Aiken article and it's funny because exactly what you're trying to avoid and or compensate for in tube amp design is what I'm looking to add to the DB so it will react more like the amp in question.

Well in the case of the boogie, and it's father the soldano, those roll-offs are planned.  It's just something that you wouldn't see if you didn't know about the miller effect.  There is another reason though for having those large series resistances in the tube amps.  It has to do with bias shifts and things like blocking distortions.  A large series resistance is always asking for noise problems though, and since you don't have those issues needing the large values with jfets you can scale the values down a bunch to help cut the noise back a little.  Scale it all by 10 and you might notice a slight reduction in noise.

Jfets are effected by Mr. Miller to.  The thing to do would be to figure out the miller value for the jfet you're using, and then add what's needed to get the tube values.  Or just go by that average of 150pf and then tweak by ear. 

Look into the asym diode thing to replace the 3rd stage when you have time.  In the amp that stage isn't ment to add gain - it's ment to chop the top off of the wave.  It's a perfect use for diodes in this circuit since they could do the job as well as a gain stage could, and the diodes wouldn't be adding the noise of a gain stage. 

Later, PaulC

[MartyMart]

Thanks Paul , interesting read .... wierd thing is that I added a diode to ground arrangement onto
my little "G3 Distortion" from last week , as a "final stage" ( 4th ) and it does make quite a difference
I guess "lopping off" the waveform a bit, lost a tiny bit of level but not much ( 3mmLED's )

I think it's time for the "perfect" amp sim, take into account IDSS of fets , Miller capacitance etc etc.
The new Fetzer valve setup is also a good read, seems like there could be a VERY exact and repeatable
amp sim waiting to emerge .....

MM.

[RDV]

The diode thing would probably make the circuit smaller and quieter but since I've got it on PCB I probably won't be doing that one.

Adding the caps however seems to have really brought mine under control.
The way it was before I had very little control with the tonestack. With everything at dead zero I still had a ton of fizz.
Now, with everything at dead zero it's very dull(as it should be) and then there is plenty of treble and mids and bass to be had by rolling up the tonestack. It's still nice even after sleeping so I know it's a good thing.
I may experiment with values a bit but the caps are staying.

RDV

[Xavier]

The diode thing would probably make the circuit smaller and quieter but since I've got it on PCB I probably won't be doing that one.

Adding the caps however seems to have really brought mine under control.
The way it was before I had very little control with the tonestack. With everything at dead zero I still had a ton of fizz.
Now, with everything at dead zero it's very dull(as it should be) and then there is plenty of treble and mids and bass to be had by rolling up the tonestack. It's still nice even after sleeping so I know it's a good thing.
I may experiment with values a bit but the caps are staying.

RDV

What caps have you added and where? It's not very clear to me after reading the post , sorry for my ignorance.....

[RDV]

The diode thing would probably make the circuit smaller and quieter but since I've got it on PCB I probably won't be doing that one.

Adding the caps however seems to have really brought mine under control.
The way it was before I had very little control with the tonestack. With everything at dead zero I still had a ton of fizz.
Now, with everything at dead zero it's very dull(as it should be) and then there is plenty of treble and mids and bass to be had by rolling up the tonestack. It's still nice even after sleeping so I know it's a good thing.
I may experiment with values a bit but the caps are staying.

RDV

What caps have you added and where? It's not very clear to me after reading the post , sorry for my ignorance.....

I replaced the 22pF from gate to ground I had in the 2nd stage with a 100pF and liked the tone a bit better.
I placed a 100pF to ground where the 220k resistor attaches to the 4th Fet's gate as well. I just had to add it on the back of the PCB.
They just filter some of the scritch that was causing me some headaches with the super-high feedback and stuff, though I've still plenty of highs on hand. It just seems a little more controlled to me. 100pF was what I had on hand, so other values could be tryed as well.

RDV

[PaulC]

What caps have you added and where? It's not very clear to me after reading the post , sorry for my ignorance.....

There's an "invisible" cap not shown in the circuits that are a part of the way tubes run.  An average value for normal types of 12ax7 biasing is about 150pf.  It depends on the how the gain is set up for the tube, but this is a good average.  Just add it from gate to ground on your Jfets.  Include it on the input fet.  It sets up a filter with the 68kinput resistor rolling back highs at about 15khz.

Without this cap being added those large series resistors are not doing the voicing things with the jfets that they do with the tubes.  If you're trying to capture the eq curves that go on between each tube stage then you need to add these caps.

Later, PaulC

[Somicide]

Using that bit of advice, I'll draw up (just a section) of how I think it should be with that.  Then you can tell me if I'm wrong 

[RDV]

Boy I'm tired from the gig, but I just had to give a report on this thing. It was amazing!

I used my newly modded DB at a gig that was supposed to be a duo thing as far as I knew, but the singer had set up a set of electronic drums and a bass amp. I'm thinking "oh GOD! no amp(I play through my RP12 into the P.A. on the duo gigs). I was just gonna try out the DB and play clean most of the night, or so I thought.

So me & the little singer do our little singer/songwriter thing for a set, then these guys come in to play(including another guitarist who brought in a half-stack.

So I turn on the DB and crank it a bit to equalise the level to my clean sound and cannot believe my earholes! This thing is sounding like an amp through the PA, people were flipping out wondering where I'd hidden the Dual Rectifier Head!

This has been one of the few times where people really wanted to know how I was doing that!

ADD THE CAPS PEOPLE.

RDV

[Xavier]

Also in Q5? Q5 doesn't have anything to ground.

So my guess is:

Q1 is clear to me, there's not any cap so I just have to add it in parallel to the resistor
Q2, make the stock 20pf  150pF
Q3 is also clear, as there's nothing neither
Q4, there's already a 1uF electro to ground. Do I have to add anything here?

So if I'm not wrong we need 3 caps for Q1, Q2 and Q3?

I might redraw the perf layout to include the cap change.....

[RDV]

I only did Q2 & Q4 because those filter frequencies are within human hearing range. I'm not too worried about 15khz, though I may add the cap to Q1 just as an anti-fizz.

Q4, there's already a 1uF electro to ground. Do I have to add anything here?

Paul is refering to caps from gate to ground.

RDV

[Somicide]

so that's all there is to this cap adding?  Add the cap going to ground, not inbetween/instead of something?

[swt]

Hey RDV And PaulC, this looks really good. The bsiab has an input cap to ground also to simulate the roll off. The only difference is that RDV put his cap from drain to ground...and Paul advice was putting it from gate to ground. I'll try this on the recentely built Dr Boogey, and let you know my results...i only have acces to BF245A, so maybe it's a different beast, but anyways, a change is going o be notice in my case also. Thanks a lot!

[RDV]

No no, I put mine from gate to ground on both Q2 & Q4. I just edited my post from earlier to try to make that clear.

I've drawn up my changes on the original schematic but I'm not posting it out of respect to Electrotabs(unless he's interested in seeing it).

RDV

[swt]

Hey PaulC...is it possible to do the assymetric stuff with fets?. How high do you think i can go on the source resistor? I was thinking on getting less gain on that stage, by means of a higher resistor, and maybe adding the diodes as suggested.

[PaulC]

Hey swt,

    The 3rd tube stage used a 39k cathode resistor which gave a massive chop to the positive side of the wave form.  You've never seen a more asym clipped signal!  It also had very low gain.  Only a couple of dB vs around 35db or so for a 12ax7 biased with standard values.  Because of the headroom issues with the jfets you've got more than enough signal to replace that 3rd stage with a diode clipper to do the same job.  Again - I've not built this.  Just looking at it I'd think it would give you better results at nailing down what the slo/boogie circuits are trying to do.  What I'd do is set it up so Q4 clips first with good symmetry, and then have the diode in place of the 3rd stage kick in after you've got a good clip on Q4.  Then after you've got the diode clipping I'd have Q2 clip a little behind it.  Play around with voltage dividers and the threshold of the diode to time the clipping between the stages. 

   When adding the miller caps to the circuits you'd have them on Q1,2,4.  Since the gain of the 3rd stage is so low it suffers very little from the miller effect, and Q5 is a source follower (or cathode folloer in the tube circuit) which does not suffer from miller effects.

   The Slo had a 68k grid stopper resistor on it's input that set up a filter with the miller cap of the input tube (Q1) to roll off the highs at about 15k.  This helped keep the radios out, and can also help with osc problems.   The only problem with this cap is it will interact with your guitar volume pot making things darker as you roll back the volume.

    But like I said - these are just thoughts from looking at the circuit.  I've not built it.

Later, PaulC

[swt]

well PaulC...i did built it, and so RDV, so thanks a lot for all the help and advices. i'm going to fool a bit with the concept and let you know. one thing i'm planning to try is...i'll use a low drain resistor, maybe 6k8, and 2k2 source to get it biased, just as a small 2x boost, transparent, and use a 100n cap, with a diode and a small series resistor, between drain and gate, to create assymetric clipping. or maybe a back to back pair of diodes, with a resistor in one for the assymetry. i can also try a bound back to back pair to ground...but i think it will over compress, and maybe make it sound a little rough...what do you guys think??

[mojotron]

Thanks PaulC and RDV - I built a Dr Boogie first without then with your observations/tweaks and it does sound great - like a RAT with real tone, very heavy.

I added an extra set of pads for each of the FETs in the PCB layout for the Miller capacitance to go from G to S (not necessarily ground - although in a lot of cases it does not matter). Other things I added to this build:

- I used an MPS102 for Q1 and a 150pF Miller (G-S) cap
- Took RDV's advice on the 100pF in for the 20pF (Q2) cap and added a 100pF Miller cap to Q4
- Added back in the 68k resistor on the input (in place of the jumper) for a 15kHz roll off
- On Q1's source, I used a .47uF + 47k resistor for the AC bypass (in place of the 1uF cap)
   +This input stage is now more like on the Original Mesa Dual Rec schematic
   +This added more control to the gain

Now, this sounds great - very amp-like, palm-mutes are unreal great!! - but I'm still tweaking.