mosfet opamp

[sonnya]

Hi all.

Just some evening fun.

I thought why not try making a all mosfet amp? only the ccs if JFET based.

It took 1 hour.

First impression is good. The goal was only a few types of components.

1. ZVN3310/ZVP3310
2. only 100R in most parts of the opamp except output wich is 47R
3. Should be able to be build by most in the forum.
4. Reasonable performance.
5. Cheap

Comments are welcome.

Todo:

• biasstage
• dc blocking in NFB

[lineup]

Good.
Nice project for having a different Preamp.

If you can find JFET which is more commonly available
for the current sources, then this would be fine.
U309 is almost impossible to find and buy.
(Mouser, Farnell, DigiKey)

[geraldfryjr]

Very cool I know of a similar project that used +-200v supplies! jer

[sonnya]

Originally Posted by lineup

Good.
Nice project for having a different Preamp.

If you can find JFET which is more commonly available
for the current sources, then this would be fine.
U309 is almost impossible to find and buy.
(Mouser, Farnell, DigiKey)

J309 is identical and available. I have to find a solution for the bias generator - maybe something like LM431!?

[homemodder]

Mosfets are so noisy, you should really use Jfet inputs.

[sonnya]

with 11x gainthe noisefloor is 63nV.. That's okay for me.

[homemodder]

I didnt know zetex supplied noise figures for their mosfets, I must look it up.

Is this the result from LT Spice sim ???

[Samuel Groner]

Noisefloor is 63 nV.

In what bandwidth? Simulation result?

I'd tie the gates of J2/J4 to the source of J1/J3. Gives much better CCS output Z.

Samuel

[aparatusonitus]

Originally Posted by sonnya

I have to find a solution for the bias generator - maybe something like LM431!?

I would use ZVN3310 based Vgs multiplier.

[peranders]

Originally Posted by homemodder

Mosfets are so noisy, you should really use Jfet inputs.

My first thought also some years ago but notice that Mr. Pass is using them in high-end products.