JFET Stages (in an Amp Emulator)

[brett]

Hi

i think you wont have any troubles from muamps.. but to be sure, you could use an opamp buffer and be 100% safe

Op-amps make fine buffers, but a JFET buffer would work fine, too, and usually be be simpler.

I seem to recall that many of the traditional (valve amp) tone stacks are designed around high output impedance from the previous stage (up to 100kohms) and high input impedance to the next stage (typically 470k to 1M).  These would suit a mu-amp design.

But different stacks have different requirements.  For example, the BMP tonestack requires Zout and high Zin, but is often seen where loss of highs will be significant (low Zin).  I strongly recommend "playing" with Duncan's Tone Stack Calculator to find out what is what.

have fun!
PS I for one simply don't believe that "tone sucking" is such a great problem.  Is it a bad thing?  We might also call it "tone mending" if we think of it as getting rid of high-order harmnics and focussing on the tone and the first couple of octaves up.  But don't believe me - I'm just some guy on the www.  Look at how many great albums were recorded in the 1960s and 1970s through BJT pedals such as the tonebender, fuzz face (around 20k input impedance), a variety of other pedals with Zin of about 50k, or using high capacitance, low quality cables (e.g. Jimi's "tone sucking like a black hole" curly cables).

[doitle]

Gearmanndude will tell you how important that white curly cable is. Owl strap too...

[anti-idiot]

Hi

I seem to recall that many of the traditional (valve amp) tone stacks are designed around high output impedance from the previous stage (up to 100kohms) and high input impedance to the next stage (typically 470k to 1M).  These would suit a mu-amp design.

But different stacks have different requirements.  For example, the BMP tonestack requires Zout and high Zin, but is often seen where loss of highs will be significant (low Zin).  I strongly recommend "playing" with Duncan's Tone Stack Calculator to find out what is what.

have fun!

PS I for one simply don't believe that "tone sucking" is such a great problem.  Is it a bad thing?  We might also call it "tone mending" if we think of it as getting rid of high-order harmnics and focussing on the tone and the first couple of octaves up.  But don't believe me - I'm just some guy on the www.  Look at how many great albums were recorded in the 1960s and 1970s through BJT pedals such as the tonebender, fuzz face (around 20k input impedance), a variety of other pedals with Zin of about 50k, or using high capacitance, low quality cables (e.g. Jimi's "tone sucking like a black hole" curly cables).

Well, i'm not using different-different tonestacks, just using the Marshall TS, and changing the slope resistor and cap (like the old plexis, the master volume and the Dual Rectifier). Now i don't know how good or bad is it gonna be if i put the 3 tonestacks together and just switch between their outputs. of course there's gonna be lose, but as all of you made me see, i'm gettin' too neurotic. I mean, if Marshall JVM can do it and still sound amazing, why not me? I'll try to finish the schematics ASAP.

Now, for something completely different, what are your experiences with the Presence Control on designs like the Plexizer? Has anybody take a look at the JCM900 4100 Dual Reverb Presence control? it's a Pot-to-cap at the cathode of a tube (post TS, FX-Loop, so i guess it's a recovery stage/buffer for the reverb stage)

Thank you all, you've been great!

[petemoore]

  A 1/2 vbias could be made with 2 =value small resistors, each stage biased through 1 large resistor from a common Vbias, 1 divider means fewer resistors total and may help reduce noise.
  Isn't the upper JFET a weird source follower?
  Yupp, it pretty much follows the voltage of it's input, the resistive divider which creates 1/2v, and biases the lower Jfet.
  So, what's the deal with this schematic?

http://i20.photobucket.com/albums/b228/Anti-Idiot/MarshallFETPreamp.gif
  Not sure which deal...
  The last stage has 1:1 resistance between the S/D, which would increase the amount of current it can put out, and be about unity gain, for driving whatever is connected to the output it seems....as far as I can calc..
  If using the above biasing suggestion [creating a V-bias with say 2 x 10k resistors], a 1 Jfet buffer would be pretty easy put into service [gate biased with a large resistor, see Labs Notebook].

[snap]

dirty little secrets: http://www.geofex.com/Article_Folders/modmuamp/modmuamp.htm

[petemoore]

  That's the one, noiseless biasing.
   Thanks RG, of course I've tried it and it works great !

[earthtonesaudio]

yes, muamps have rather high output impedance..compared to a buffer, of course.. but putting a buffer before the tonestack should be a must independently of what you are using for gain stages (well not for opamps, maybe)

Isn't the upper JFET a weird source follower?  Kind of, but it's also a load for the lower JFET.

So, what's the deal with this schematic?

http://i20.photobucket.com/albums/b228/Anti-Idiot/MarshallFETPreamp.gif   The 3 first stages are standard mu-amps with different levels of gain and filtering.  The last stage is probably supposed to be a buffer, but the top JFET has no gate bias, and so it won't work right as drawn.

What if i use an additional source-follower? (pretty much like this one http://www.runoffgroove.com/tc-tone.png)

http://www.geofex.com/Article_Folders/foolwfets/foolwfets.htm

You could do that, but if you add a single resistor you'll get basically the same result: http://www.geofex.com/Article_Folders/modmuamp/modmuamp.htm ...if it was me, I'd replace the stage just before the tonestack with the last circuit in this link, and replace the output buffer stage with a more traditional common drain follower circuit.

[anti-idiot]

What if i use an additional source-follower? (pretty much like this one http://www.runoffgroove.com/tc-tone.png)

http://www.geofex.com/Article_Folders/foolwfets/foolwfets.htm

You could do that, but if you add a single resistor you'll get basically the same result: http://www.geofex.com/Article_Folders/modmuamp/modmuamp.htm ...if it was me, I'd replace the stage just before the tonestack with the last circuit in this link, and replace the output buffer stage with a more traditional common drain follower circuit.

I'll follow your advice and replace the buffer (i was thinkin' about some stage with a 'presence control')

Ok. I got the 10k/10k, and the 1k between both JFETs. But what's the deal with the 4M7?
So, basically i should use a common 1/2vBias and connect as many 4M7 resistors per stage, right?