ampage Tube Amps / Music Electronics	For current discussions, please visit Music Electronics Forum.

ampage archive

Vintage threads from the first ten years

FET Electrical Performance

:

10/25/1999 6:59 PM

FET Electrical Performance

I have been working on a clean boost circuit and have been experimenting with the 2N5457 FET. With a 2.2K drain resistor, a 470 ohm source resistor, a 1M gate pulldown resistor, a 20 uF source resistor bypass, and a 9V supply, I get 5.4V at the drain. This should be about right. When I feed it with a .25V sine wave, I get a 1V sine wave out at the drain for a gain of 4, which is fine for my intended purpose. The problem is that the distortion is 1.5% THD at the 1V out, as measured using an HP 200CD oscillator and an HP 334 distortion analyzer. The test equipment is known to operating correctly. What is the typical distortion performance of these devices? I mainly work with tubes, so I have no idea what it should be.

10/25/1999 7:27 PM
R.G.	1.5% THD from a single FET gain stage is not unreasonable at all, especially if you have the whole source resistor bypassed. JFET's are square law devices, so the THD is probably mostly 2nd harmonic.

10/25/1999 9:06 PM

RG,

Thanks for the info.

This is pathetic performance compared to say a 12AX7, where you get a gain of 60 with less than 1% distortion, and if you were to use feedback to reduce the gain to 4, the distortion would be less than 0.1%. And at much higher output voltages. A study of the gain divided by distortion for various devices would be interesting - High gain and low distortion would give a large number.

So how do you make a low distortion circuit with FETs? Use differential amplifiers to cancel the even harmonics? The open loop gain of the FET is already so low, applying negative feedback isn't really an option.

I wanted to use the FET because of the high input impedance and the simplicity of the circuit, but it won't exactly provide a "clean" boost.

10/25/1999 9:31 PM

Well - the vacuum triode is the most linear open loop amplifying device ever invented, so having a FET come close is actually a compliment.

The only way to get lower than a triode's distortion is to use linearizing feedback. The study of gain versus distortion by device was done long ago, and the triode won.

Some salient points: You probably can't hear the FET's distortion at 1.5%, being purely square-law distortion as it is. In that way the JFET resembles the triode in having primarily second order. The human ear is very tolerant of that. You might try the JFET and see how it sounds, not how it measures. Many people have built JFET gain stages and like them very much for guitar use.

To get lower distortion with anything, you up the gain and apply feedback. There are some JFET/bipolar transistor pair circuits that can give open loop gains of around 4000 and closed loop distortion well under 0.1%.

For "clean boost", you have a couple of options. You can go with the JFET and see if you like the sound anyway. You can use a JFET/transistor pair and feedback, or use a JFET input opamp.

10/25/1999 9:52 PM

quote:
"For "clean boost", you have a couple of options. You can go with the JFET and see if you like the sound anyway. You can use a JFET/transistor pair and feedback, or use a JFET input opamp. "

I will try just the jfet as you suggest. The jfet/transistor pair sounds interesting also. I assume they are direct coupled somehow? How is the biasing done? The pair with feedback would give a low output impedance, acting as a buffer, correct?

10/26/1999 10:05 PM
Ed Rembold	Scott, One thing you might try is a different Fet- NTE458 are just killer in the gain department, you probably won't even need a bypass cap, which should lower your distortion level. The fets your using are Really low gain. Ed Rembold

10/26/1999 11:12 PM
Jack	The 9v power is your problem. Crank the voltage up to about 10% or 15% below the rating of the jfet and its performance will improve drastically. regards, Jack

Page 1 of 2