Explain to me Mosfets vs Jfets

[trad3mark]

Hello world,

I'm re-visiting an old project of mine at the moment. A while back, along the lines of the fet based amp simulators, i made a j201 based simulator of a Silvertone 1485 (cos jack white is a hero). Anyway, at the time, it was all ready for a box, but the box i had was too small. shortly after, i swapped the trimmers out for use in another project. I'm planning on revisiting it, but now i'm short on jfets. disaster.

SO, onto the topic question of, Mosfets vs Jfets. I've spent the last half hour on the interweb trying to see if I could use a mosfet in place of a jfet, and i'm not really getting anywhere. I've seen pedals that use jfets and ones that use mosfets, so, without going into the theory behind it too much, in real terms, what kind of issues will i be facing?

Will i be able to, for example, swap a J201, with some sort of mosfet? I'm familiar with the whole fetzer thing on runoffgrove, biasing it etc etc, and the whole triode background, but could a mosfet be used as a "triode > fet conversion"? is it done the same way?

MOST IMPORTANTLY, if so, what kind of differences will i notice? gain differences? tonal characteristics difference? more/less noise?

Cheers,
N

[R.G.]

I'm re-visiting an old project of mine at the moment. A while back, along the lines of the fet based amp simulators, i made a j201 based simulator of a Silvertone 1485 (cos jack white is a hero). Anyway, at the time, it was all ready for a box, but the box i had was too small. shortly after, i swapped the trimmers out for use in another project. I'm planning on revisiting it, but now i'm short on jfets. disaster.

Best solution is to order some JFETs.

SO, onto the topic question of, Mosfets vs Jfets. I've spent the last half hour on the interweb trying to see if I could use a mosfet in place of a jfet, and i'm not really getting anywhere. I've seen pedals that use jfets and ones that use mosfets, so, without going into the theory behind it too much, in real terms, what kind of issues will i be facing?

The real issues are that although they both have "FET" in the name, they are different animals. JFETs are depletion mode devices, similar to vacuum tubes in this one respect. That means if you hook an electrical source across them, they will conduct all they can unless you do something to the gate to turn them off. MOSFETs can be either depletion mode or enhancement mode, but all the ones you can get easily are enhancement mode. That means in the absence of any encouragement on the gate, they will block all current flow through the channel entirely. The only circuits where you can sub in a MOSFET for a JFET without doing any other changes are probably in cascode or follower circuits where the voltage the gate will connect to is significantly bigger than the turn off voltage for the JFET and also bigger than the turn ON voltage for the MOSFET. As you might guess, these circuits are rare.

In particular enhancement mode MOSFETs will NOT work in cathode self biased setups like most of the "I'll stick in a JFET and call it a tube" circuits.

Will i be able to, for example, swap a J201, with some sort of mosfet?

Not for the common MOSFETs you can actually get like the 2N7000 or BS170. There are esoteric MOSFETs that do exist that might do that, but they are not common in the electronics biz.

could a mosfet be used as a "triode > fet conversion"?

See "The MOSFET Follies", geofex.com; some circuits, limited applications, and you have to know the limitations. But a JFET circuit is already pretty far from a triode, in spite of some similarities. Different animal.

is it done the same way?

No.

MOST IMPORTANTLY, if so, what kind of differences will i notice? gain differences? tonal characteristics difference? more/less noise?

Other than no sound at all without redoing the biasing? 

JFETs are much transconductance than MOSFETs. Whether the circuit uses this to make voltage gain for you depends on the circuit. Both are fast enough that the frequency response (tonal differences, mostly; some distortion issues will enter into this too) is far above audio, so the circuit and biasing call this tune. JFETs are probably quieter in an "ultimate noise test" setup, but they're both good enough for rock and roll.

[trad3mark]

awesome. At the very least, the simple answer is no. cheers for the help R.G.

off to order some fets. i only need 2, so it sucks in a way.

[ayayay!]

Also if this helps:  Learn your acronyms.    

http://www.acronymfinder.com/JFET.html

http://www.acronymfinder.com/MOSFET.html

[slideman82]

but they're both good enough for rock and roll.

That's the whole point in this forum, I think. I love these devices, specially JFETs. I told a friend "I wouldn't trade my SS amp for your Boogie 5:50!". Also many users have built several JFET based distortions and they do sound great!

The easy way: if you have a gain stage or a source follower (buffer) with a MOSFET with a voltage divider (that's a resistor from supply voltage or drain to gate and other resistor from gate to ground, although there are other ways to polarise gate) you could try to place a JFET instead of it, but you'll need to adjust drain resistor till it sounds great (or till you measure about half the power supply, that would be about 4.5 or 5V) in the gain stage. If it's a JFET stage with a single resistor from gate to ground, you can place a MOSFET instead of it and adjust bias but it won't sound. I've already tried it a while ago.