New circuit: Hiawatha (And I can't get it to work...)

[syndromet]

In my quest for learning more about JFETs and the way they work I decided to make an atempt on an amp sim of my dream amp. Its a Hiwatt SA112, but I'm not sure I've gotten it right. If anyone of you wizzards would care to check if this seems to be a woring ciruit, I would be extremely happy.

Orignal schem


JFET version

[Sir H C]

I have to say that the 100k pots will probably be cranked almost to one end with a 9V supply.  JFETs require the gate voltage to be negative relative to the source to turn them off.  Maybe a volt or two (really variable for these things).  So if you have 2V across the 2.2k k, then to have the output in a cool spot you need maybe 4.7k.  That is nearly off for the 100k pot.  Also you have negligible gain relative to the original circuit.  Either the two resistors have to be scaled down (220 and 10k) and deal with the high current, or something else should be done.

[MartyMart]

Looks to me that you only have to go as far as the four 12AX7 stages, just after stage 4 and
presence control is your final output from the 22n cap.
Emu of the two power tubes with fets wont  work correctly, but the rest will get you the "flavour" and
"personality" of the amp.
Take a peek at some of the ROG/Nelson/MartyMart sims etc, perhaps their dual cap/ dual resistor roll off at the end
would benefit some too.
The Bassman and Classic 30 sims that I did got "very" close to the actual amp sound, which whilst not a replacement
for a valve amp, make great little pedals non the less

Marty.

[George Giblet]

Appart from the issue of lower gains etc.

To mimic the original circuit a second 470k should be placed between second JFET's gate and ground.  This will give a gain drop which is in the original ckt but probably undesirable for the JFET version. The closest mimic is to replace the 470k with 220k.  This is al very minor and little academic.

Your replacement tone control is a significant difference to the original.  The main problem with it is the 10k impedance is quite low compare to the impedance of the driving stage.  You might want to try 10k + 100k tone pot + 1.8n cap + 500k pot.

In the original tube circuit tube V3b seems to do very little.  It's just a DC bias for the first power amp stage.  I can't see this doing much and it will only add to the biasing issues with the JFET version.  In the original tube circuit the supply ripple may have added to the tone but even that is highly unlikely and speculative (because that power amp stage is differential and insenitive to the bias point).   Given the following stage has a variable drain resistor I suggest simply connecting the 100k that goes to the gate the last JFET to ground, and of course complete remove the V3b emulation.  You could experiment by connecting to the centre of two 10k resistors connected to +9V and ground.

[syndromet]

Thanks a lot, Marty and George! I really appreciate it.
I have updated the schematic, and I think I understood all you guys said. Seems to improve the circuit and make it an easier build, wich is great!
Let me know if you see any mistakes.

[George Giblet]

With the 470k, I mean replace the 470k on your original circuit with 220k.

The presence control won't work like the original, currently it will only act to cut treble (and there's already enough of that).

In the original circuit the presence control is part of the feedback loop of the power amplifier.  Since you aren't emulating the power amplifier that make is awkward to emulate the presence control.  It could be done using another JFET state following the 22n + 100k at the output.  A simple way would be to connect the presence pot to the source of the added FET stage but it will require some design work getting it to be close to the original.  On the hiwatt circuit the components affecting presence are the 10k, 470R, 10nF and the 2k2 resistor in the tail of the V4a and V4b, and of course the presence control components.

Another option is to add another stage which has a feedback network where you can emulate the presence control and to some extent the power amplifier.   It would be possible to adapt the output stage of this circuit,

http://www.runoffgroove.com/professor.html

[petemoore]

  taking Prof. Tweed as Example...
  what Would the circuit as shown sound like if tubes were used [ie tap a princeton reverbs 12ax7 preamp section, and listen somehow].
  How much of the sound of the amp is in the output section?
  Speakers
  I suggest that designing a Jfet circuit should be likened to designing a tube circuit, design, test, tweek, re-design, test, tweek, field test, tweek...revel in glory?...move on...design.
  The basic premises for working up a great sounding Jfet Cct. are, as I can tell...
  Voicing...lots can be done here
  Distorting and compressing...lots of shades and variations can be produced, how much each stage distorts [how hard it's pressed to clip], the pre-post distort voicing, and how many stages are used to add distortion.
  As far as making it sound just like the whole tube amp by building part of a tube amp using Jfets...I think some morphing may be needed, especially if the amp in question gets alot of it's distortion character from the output section.
  ///Now that I've said that...I'll provide a more 'purist' form, do what Leo did, work, with using the available parts to produce an 'optimized' circuit, let the circuit help you decide as opposed to attempting to force the circuit to do something it won't.
     

[syndromet]

In the original circuit the presence control is part of the feedback loop of the power amplifier.  Since you aren't emulating the power amplifier that make is awkward to emulate the presence control.  It could be done using another JFET state following the 22n + 100k at the output.  A simple way would be to connect the presence pot to the source of the added FET stage but it will require some design work getting it to be close to the original.  On the hiwatt circuit the components affecting presence are the 10k, 470R, 10nF and the 2k2 resistor in the tail of the V4a and V4b, and of course the presence control components.

Another option is to add another stage which has a feedback network where you can emulate the presence control and to some extent the power amplifier.   It would be possible to adapt the output stage of this circuit, http://www.runoffgroove.com/professor.html

I don't know if I understood this correctly. If I keep the circuitry around V4a, and remove v3b from the original schem it should work?

[syndromet]

Ok, an update. I decided to try to use the first power amp stage as a gain stage. Would this change the sound to much? What do you think?

[MartyMart]

Looks good, I wouldn't panic too much about changing a couple of areas, these are not a 1 to 1 "replacements" !!
You're just after the "flavour" or "Tone" of the original, major gain increases and tone area changes, will make it
into something else ..... often thats better anyway 

Nice work,

Marty.

[syndromet]

I'we made a layout for this too. I'll try it when I get the time, and post a few soundclipps.
I'we also given the circuit a name.

[RaceDriver205]

Is that the bancika program again?
That thing is going to make a lasting impression in this forum, I know it. Good on him.
Id use it too if it weren't for the fact that I HATE PERF.

[MartyMart]

I'we made a layout for this too. I'll try it when I get the time, and post a few soundclipps.
I'we also given the circuit a name.

Take off the trace "cuts" on the 9v line ... or you're trim pots wont have anything to "trim"

MM.

[syndromet]

I'we made a layout for this too. I'll try it when I get the time, and post a few soundclipps.
I'we also given the circuit a name.

Take off the trace "cuts" on the 9v line ... or you're trim pots wont have anything to "trim"

MM.

yeah.... I see that now. Guess I used the scetch for the layout. Should be corrected now.

[syndromet]

Is that the bancika program again?
That thing is going to make a lasting impression in this forum, I know it. Good on him.
Id use it too if it weren't for the fact that I HATE PERF.

Yes. I love that program. I personaly love vero. I hate perf, and find Pcb's to be a little too timeconsuming.

To bancika: THANK YOU!!!

[Eb7+9]

have you tried a "boost" cap across 2k2 source resistor in the first stage ?! (something like 5uf ~ 47uF)

also, just an observation ... the 220k that's in series with gate of J2 (mimicking 470k grid blocker) isn't necessary per se as far as reducing Gate current goes - I'd try it both ways, the circuit may be slightly noisier with it in there ...

promising circuit ...

[TELEFUNKON]

imho, it`s no grid-blocker (too big anyway);
it was there for mixing reasons (see orig. schemo)
of 2 channels.

[RaceDriver205]

and find Pcb's to be a little too timeconsuming

Well yes, they do that for sure.
But can you imagine doing all this on vero?:
http://www.geocities.com/race_driver205/effects2.jpg
**Shudder**

[syndromet]

have you tried a "boost" cap across 2k2 source resistor in the first stage ?! (something like 5uf ~ 47uF)

also, just an observation ... the 220k that's in series with gate of J2 (mimicking 470k grid blocker) isn't necessary per se as far as reducing Gate current goes - I'd try it both ways, the circuit may be slightly noisier with it in there ...

promising circuit ...

imho, it`s no grid-blocker (too big anyway);
it was there for mixing reasons (see orig. schemo)
of 2 channels.

In the original schem there is a 47nF cap in parallel with the 2k2 ground resistor on the brilliance channel. Maybe I should add that to the stompbox to?

So you guys think the 220k is for mixing the two channels? Seems logical. I'll try it with and without the resistor when I get time to build this thing.

Thank you for your replays.

and find Pcb's to be a little too timeconsuming

Well yes, they do that for sure.
But can you imagine doing all this on vero?:
http://www.geocities.com/race_driver205/effects2.jpg
**Shudder**

I don't do that kind of monster-stuff... Looks good, though. What is it?

I actually bought an etching set a couple of months ago, and I've only etched an orange squeezer this far.

[RaceDriver205]

What is it?

Lol, many many things.
If you want to know more - http://www.diystompboxes.com/smfforum/index.php?topic=44752.0