JFETs and triodes

[R.G.]

An old friend of mine did some very nice work on comparing JFETs and triodes. It includes some buildable circuits.

http://desertcomb.com/gas/projects/preamp/overview.htm

[QSQCaito]

Tell him to change that background!! I can't read !

Thanks a lot RG
ill take a look!

Lets see if i can learn something

DAC

[brett]

Hi
That's a substantial piece of work.  The amount of simulation alone is mind-boggling.

I don't wish to criticise what was done - it is a great study of pre-amps.  But many articles like this suggest that the tone from valve amps is mainly due to overdriven valve pre-amps.  I feel it is easy to underestimate the roles of valve phase splitters, valve rectifier power supplies, power valves and output transformers.  In particular, there is lots of evidence that valve phase splitters play a big role in tone, but they get remarkably little attention.

Thanks for the article RG.  You've got some very smart buddies.
cheers

[rodriki1]

I was about to give up thinking about a serious work at this subject.

All my questions before was answered.

i would just like to comment that the work has similarities with the ideas
of the article of john murphy from carvin.

THANKS A LOT.

Rodrigo

[R.G.]

Yeah, the overall picture is remarkably complex. I was impressed at the degree of effort that went into studying the real technology and finding real answers other than that JFETs and tubes are both depletion mode devices. There is always more to do, but I was impressed with Ben's efforts so far.

i would just like to comment that the work has similarities with the ideas of the article of john murphy from carvin.

Yes. I believe that John's article on that was referenced.

More importantly, it appears that tube emulation is one of those things that disappear when you look too closely. The waveforms of one tube can be simulated, but it's the assemblages of tubes, circuits, and other paraphernalia that make for the whole sound. Both Ben and John before him are digging into the whole mess, looking for the forest to tell them how to model a tree.

[MartyMart]

Great RG, he's put an awful lot of work into this and has produced a very nice piece of work.

It's interesting, I've "delved" into the Jfet amp sim thing quite a lot recently and found the results can
be pleasing, but difficult to repeat with success.
The new "Fetzer Valve" article and a couple of things that Gus put me onto are helping, I've returned to
one or two circuits and I'm getting better results, taking into account IDSS etc etc.
I've also tried using diodes within the circuit too, when I get something that I really like, i'll post it up.
I dont think they will ever be replacements, but to have something "close" to the sound of an amp in a
pedal, has to be a good thing   Just look at the BSIAB II  / Dr Boogey etc !!

The LHX2 that I built is also wonderful ( it's referenced in the end list too )

MM.

[billings]

The use of zeners for clipping was something I hadn't seen before.  Has anyone in the DIY community used zeners as clipping devices?

[petemoore]

  Seems like a long, windy, road...tracing the path of tube sound with Jfet emu-ing triodes.
  With no possibility of actually making a scale model of the power supply and other related to tube sound items found in tube amps.
  Why not?
  "Jfets use a very small amount of current compared to amps, so it's impossible to replicate the 'power supply' influence on the sound in smaller voltages" [?]
  Then hook on some kind of speaker and tap the output somehow...
  By the time and trouble scale, calculating that a tube circuit could probably be built with the same effort...

[R.G.]

The use of zeners for clipping was something I hadn't seen before.  Has anyone in the DIY community used zeners as clipping devices?

Then you have not read the Guitar Effects FAQ at GEO (http://www.geofex.com). It was in the first edition of that, which predates the world wide web.

With no possibility of actually making a scale model of the power supply and other related to tube sound items found in tube amps.

Actually, it's pretty easy to scale the power supply down. The difference from tube circuits is that the DC impedance is much higher. So put a resistor in series with the power supply. If you really wanted to get the compression effect, generate an envelope from the signal and cause that to compress the output, or use it directly to lower the power supply voltage. That last is patented, by the way - another worthless patent, but patented still.

Even with all the circuitry needed to simulate the action of a tube, it can be argued that once the design work is over, all low power PCBs of electronics are almost equal in cost and time - it's always (a) buy the board (b) stuff the parts (c) solder the board.

[billings]

The use of zeners for clipping was something I hadn't seen before.  Has anyone in the DIY community used zeners as clipping devices?

Then you have not read the Guitar Effects FAQ at GEO (http://www.geofex.com). It was in the first edition of that, which predates the world wide web.

True, but this is all you say about them in the "Distortion 101" section:

I've never seen a schematic with zener clipping for music, but they are used in the signal processing world.

And without regard to the content of the Guitar Effects FAQ, I was referring mainly to the fact that I hadn't seen zener clipping in any of the circuits I have personally seen, with the exception of the circuit in Mr. Craven's thesis.  On top of that, Mr. Craven specifically mentions that the zener clipping sounded better for his purposes.

So why aren't they used?  Should they be used?

[R.G.]

They aren't used because the voltages are too high.

You don't usually get zenering much below 3V. With a 9V battery power supply, you only have 4.5V best case to work with, most opamps won't go closer than a volt or two to the power supply, so you quickly run into situations where the opamp clips but not the zener.

Zeners tend to have a more abrupt "knee" - they go from fully off to fully on faster than diodes, so the clipping is sharper.

They were used in Stephan Moeller's AC30 simulator for the same purpose as Ben used them - to make a sharp clipping point. I believe Stephan may have softened them a bit with series resistance.

My take? They should not be used unless you have enough power supply to drive them and a design that needs the sharper clipping.  But that's just me.

[brett]

Hi
In previous threads I've mentioned that the diode "knee" is similar in absolute terms (I vs V) for a range of diodes, including Ge and Si.  Therefore, the *relative* sharpness of the knee is proportional to the forward voltage.  Put another way, germanium diodes have a standard knee on top of a small waveform (0.3 volts), while LEDs have the standard knee on top of a large waveform (2 volts), which gives a sharper shape.  As an example: if the knee is 0.1 V a signal clipped by a Ge diode is 1/3 knee, while an LED-clipped signal is 1/20 th knee.  This is why LEDs give "crunch".

The only variation that I found was that 1N4148 diodes have just a little bit more knee than other diodes.  Something to do with being laced with gold, I suppose.

Therefore, LEDs should be the best option for the zener/tube sound in 9V stompboxes.
Cheers