Need some help understanding JFETs, Voltage and overdrive

[BuddyPrince]

I'm continuing to learn about all this stuff as I'm working on building a clean boost for my bass.  I had it working very nicely when I had passive pickups in the bass, but then put in active pickups and the signal is clipping.  It actually produces a pretty decent overdrive tone, but it's not what I'm going for.  I tried various resistor values, but that didn't help.

Here's the circuit I built from Tonefiend http://www.tonefiend.com//wp-content/uploads/DIY%20Club%20Project%203%20v02.pdf

In place of the potentiometer, I have a 470 Ohm resistor.  I'm using a different JFET, don't remember which one off the top of my head, but it's rated at 6V.



My question is, would a JFET with a higher voltage rating solve the problem?

[slashandburn]

While your waiting on a more complete answer it may be worth trying a 100k pot or resistor on the signal input path, as seen on the Zvex Mastotron).

I'm no authority on the subject but I understand believe this feature  (at least on the mastotron) was to make the pedal play better on Active pickups.   Possibly. 

[GibsonGM]

If I had my choice, I'd add  a source resistor and run this on 12V or 18V for max clean headroom.

Designed like this:  http://www.runoffgroove.com/oldfetzer.html

[BuddyPrince]

Thanks y'all!  I'll try these out.  The Old Fetzer looks nice and simple, and I think I have all the parts laying around.  I'll look into the Mastotron too.

[PRR]

> In place of the potentiometer, I have a 470 Ohm resistor

Why? ? ?

Why not try it as-published, instead of random values?

470r is awful low for most JFETs. It also is a HEAVY drain on a battery. There is a reason the designer drew a pot there. Try it.

In general though, that isn't an all-purpose bass/guitar preamp. The JFET is zero-biased, nearly hard-on. You can get a little more current by forcing the Gate positive, but in most instrument situations that lust leads to clipping. Agree with the other guy: there is usually a Source resistor, and such plans are easy to find.

[BuddyPrince]

> In place of the potentiometer, I have a 470 Ohm resistor

Why? ? ?

Why not try it as-published, instead of random values?

470r is awful low for most JFETs. It also is a HEAVY drain on a battery. There is a reason the designer drew a pot there. Try it.

In general though, that isn't an all-purpose bass/guitar preamp. The JFET is zero-biased, nearly hard-on. You can get a little more current by forcing the Gate positive, but in most instrument situations that lust leads to clipping. Agree with the other guy: there is usually a Source resistor, and such plans are easy to find.

Thanks for the tips.  As a relative newbie I appreciate them...

The guy who published it has the pot in there to help identify the appropriate resistance to get the most boost out of the JFET.  It's not meant to remain in the circuit.  After having the circuit work well with my passive pickups and in an attempt to learn more about how all these things work, I tried a bunch of different resistor values to see what effect they had on the gain to see if I could get a boosted clean signal that I liked.  I couldn't...so here I am... :-)

Seems like there are definitely better circuits out there.  This was a good entry to begin understanding how all these components work together.  I look forward to building the Old Fetzer and hopefully gain even more understanding.

[Rob Strand]

If I had my choice, I'd add  a source resistor and run this on 12V or 18V for max clean headroom.

Designed like this:  http://www.runoffgroove.com/oldfetzer.html

I third adding the source resistor.

[antonis]

Shouldn't be a Gate bias resistor somewhere there..??

[GibsonGM]

Shouldn't be a Gate bias resistor somewhere there..??

Pulldown?

[antonis]

Pulldown?

You may call it as you like, Sir... 

I can't see in OP schematic neither IN coupling capacitor nor Gate bias item so I've to suppose a direct DC coupled zero Volts booster..

[GibsonGM]

Pulldown?

You may call it as you like, Sir... 

I can't see in OP schematic neither IN coupling capacitor nor Gate bias item so I've to suppose a direct DC coupled zero Volts booster..

Yes, the schematic the OP posted seems to be missing some things!  It will only amplify positive parts of the signal, because it  is not 'elevated'.
I meant this one: http://www.runoffgroove.com/oldfetzer.html   which has the 'missing parts'

The OP schematic does not seem like it would give a very 'clean' boost!   

[antonis]

IMHO, the 68k resistor on Fetzer Valve is some kind of "residual" from original Fender tube input stage..
(grid stoppers were used mainly to prevent blocking distortion when input tube grid voltage becomes more positive than the cathode voltage - and also prevent very high freq oscillation by forming a low-pass filter together with Grid-Cathode interelectrode capacitance..)

Some FETs work well with such a series resistor (of much lower value, e.g. 1k) but it shouldn't happen anything if you omit it..

P.S.
Why you tease me to write things you already know, Sir..??

[GibsonGM]

No teasing, Antonis!  My hope is to move Buddy over to the Fetzer, so he can see how it "should" be designed.  Not to say the original circuit is "bad", but it seems to be kind of "off", doesn't it?  You noticed that it has no bias...Runoffgroove has provided us with a very easy way to fix all of this!

I have to agree, the 68k is very familiar to Fender people    Maybe it is there to build the Fender spirit into the FET?  Ha ha. 

[antonis]

Maybe it is there to build the Fender spirit into the FET? 

Maybe yes..

If "spirit" could only depend on transconductance..

[BuddyPrince]

A few questions...

1. Gate bias resistor...help me understand...this is a resistor before the gate (base) that, along with other resistors biases the voltage of the transistor.  i.e. They set the operating voltage/current of the transistor?  Am I understanding correctly?

2. In the Old Fetzer schematic, is the 1M pull down resistor acting also as a biasing resistor?

3. How does one go about determining/calculating the values of resistors used to bias a circuit like the Old Fetzer?

Actually, the original circuit I posted was a great very clean boost, with way more boost than I needed.  It worked great with my lower output passive pickups.  Not clean with my high output EMGs...

I gotta pick up new soldering iron tips, hopefully I'll get the Fezter built and working in the next day or two.

[antonis]

1. Yeapp.. 

2. Yeapp..   
(without "also"..)

3. Only from desirable Input impedance.. 
(also restricted by leakage Gate current to not be significally off-set..)

If there is Source resistor, like in Fetzer Valve, Gate is set to GND via 1M resistor and Source is set a couple of Volts more positive than Gate via Source resistor voltage drop due to Drain-Source current..
(FETs should work almost predictably in such a way - but there often are surprises..)

[GibsonGM]

The source resistor you see on the Fetzer is actually there to 'elevate' the gate by a few volts...this sets the gate to be in the "middle" of the supply voltage (sort of).  This is the FET equivalent of the cathode resistor in tubes.    Here is where it gets dodgy!   
Read (I should have posted this link instead of the 1st one!)  http://www.runoffgroove.com/fetzervalve.html

There is also this, very informative: http://diy.smallbearelec.com/HowTos/BreadboardBareAss/BreadboardBareAss.htm

The above 2 pages have solved the bias problem, and my talking much about it will only serve to confuse you

In a triode gain stage, like a REAL Fender amp, you use the tube curves from the data sheet to find an 'optimum' (for you) bias point.  This is set such that the input signal can rise above AND below this point, so the negative swing of your signal can also be amplified.     They do it a little differently with FETs, but it is very similar, too, and as Antonis noted - the Drain > Source current is what makes this happen.   The 1M pulldown is really so the gate can confidently know that THIS POINT is ZERO VOLTS (ground).     The FET does the rest...the tutorials explain how.

I really have no inkling how your original circuit sounds clean, but if you say so, I will believe you.   It can't amplify the negative excursions of an input signal...it's clipping them.   Probably sounds good, LOL...since the source is referenced to gnd.   Maybe the gate floating like that has some effect, I dunno.   Anyway, if you liked it, you'll probably LOVE what ROG and Smallbear have to offer!     Especially if you tried TWO stages, in cascade, with a 'gain pot' in between, with a bypass cap only on the 2nd stage (or not at all)....if you don't want nasty clipping, having a small gain for each stage and the ability to control the 2nd stage will make a big difference, IMHO.    Smallbear is showing something a bit like this low on the page, more for adding a tone control, which you could simply omit and use a pot in there.

[PRR]

> It will only amplify positive parts of the signal,

No, with a full study, it turns out to be a fine small "clean" amp. The Gate _can_ go positive, but at +0.6V it sucks huge current, and 0.4V-0.5V may be real heavy load in a guitar chain. It also goes negative just fine. The result seems to be that a 1+V peak-peak input makes a 2.5V p-p output, of such limited swing (around the very high 5mA biaspoint) that THD may be just a few %. So a mild boost.

[antonis]

IMHO, 470R load line slope should be wider (upper part closer to Id axis) leading in FET working into "linear" region (variable channel resistance) for V_GS slightly above -500mV, say..
(just a guess/feeling of course - influenced perhaps by parameters wide spread.. )

[GibsonGM]

Really, Paul?  I threw it in LT Spice.    As I increased Rd (I first began with a higher Rd, hence my clipping negatives statement), I got positive spikes  (well amplified) and clipping the negatives at around .6v or so.  Dropping Rd to 1k or so did result in a well-shaped wave, lower values good shape but still offset....still nothing crossing below zero for the obvious reasons...kind of thinking, how does the next device (or amp) deal with this?   Something I've never looked at.   But I do see that low Rd will get you a good wave form.