JFET gain control "resistor" in Weinbridge oscillator

George Herold · 2008-09-05T17:52:10+00:00

Hi All, This is my first time posting though I=92ve read some of thediscussions. I=92ve been having problems with some of the Weinbridgeoscillators that are in our lockin amplifiers. I won=92t bore you withdetails, but a circuit that was working for years now seems to behaving problems. I used a light bulb as the gain control element.But I=92ve now decided to try moving into the 20th century and use aJFET as the control element. The circuit I=92m modifying comes from J.Williams at Linear (AN-47). My problem is finding a replacement JFETfor the 2N4338. Ohh I can still find these on the web, but for $4.30each!!! I found both the PN4119 and J201 as possible replacements andwas looking for any other suggestions. The important parameter is theon resistance of the FET which should be in the 1k to 3k ohm range.Thanks in advance,George Herold

G

George Herold 18 years ago

e

hat

ers

Hi Phil, I got my morning work done and slapped together your circuit. Since I don't have to wait for any FET's to arrive and can play right away. I used OPA227 opamps.. (mostly because they were lying around.) The circuit oscillated right away. There was some high frequency "fuzz" (2 MHz.) between -1.2 and +1.2 volts... something to do with the diodes no doubt, But I put 1nF in parallel with the feedback of IC2a and that cured it. Now I've looked at the distortion at the output of the three amps and it's pretty ugly!

Yes to say the least there is a bit of magic here. I didn't put this final op-amp into the circuit. The fourth op-amp looks a bit like a differential amp, but with differerent gains in each arm.... and then you are mixing in part of the output from IC2a in also... Is there any theory to guide in the selection of these "mixing ratios" or was this done by trail and error?

Thanks again for the circuit idea. George

Vote

G

Gordon S. Hlavenka 18 years ago

Joerg replied:

Yup. They didn't even ask my permission. Take a look at this photo:

formatting link

The dual rectangular box aft is the Seasparrow launcher, and the binocular-looking thing just above the flight deck is the Seasparrow radar director. Every other piece of RF gear and weaponry is gone.

Gordon S. Hlavenka Join the Revolution at http://www.ronpaul.com

Vote

P

Phil Allison 18 years ago

"JosephKK"

( snip the cretin's mindless personal abuse )

** Go drop dead - you autistic pile of dung.

Then your rotting corpse will match your brain.

..... Phil

Vote

P

Phil Allison 18 years ago

"George Herold"

** Learn how to trim - George

And do NOT answer TWO people the same post.

Very rude.

Oh and Phil, by changing the postion of the capacitor and resistor at the noninverting input of the second opamp, you can make the second "all pass filter" of the noniverting type. Then you don't need the tirhd inverting op-amp stage.

** Then you do not have the 3 needed signals for the fourth op-amp stage which cancels the THD.

You are a fool to mess with that circuit.

You have to then add some gain control to one (or both) of the all-pass stages, but this is easy! So at the end of the day yesterday I had an oscillator running from 2 to 20 kHz with quadrature outputs and -35dB of third harmonic. (The third harmonic is all from the diode limiting and should go away as soon as my FETs arrive.)

** Got news for you - pal.

JFETS are *highly non-linear * devices, so you will have lots of second harmonic if you use one as a gain control element with any more than 50mV of signal across it.

You have FAILED to see how elegant, foolproof and FAST the diode limiting is compared to other methods.

So I'll reserve judgment on Phil,

** Yawnnnnnnnnn ....

and also note that good idea's can come from anywhere.

** But you are no respecter of persons or a recogniser of good ideas.

...... Phil

Vote

G

George Herold 18 years ago

Sorry, I'm still a newbie! I'll try to do better in the futrue.

I'm certainly a fool, but if I don't "mess" with the circuit then how do I learn anything. And I do always get excited by a few initial successes only to later have reality come and bite me.

Well this is my first time using a FET for anything but a low resistance switch. It looks like I'll have something like 0.5 to 1 Volt across it, so I should fail miserablly. (Have you seen Jim Willaim's cure for the channel resistance modulation AN-43 figure 45. Linear Technologies).

Why do you say that? I really like the circuit and "magic" mixing to remove distortion. Other circuits will serve as a comparison.

Phil would it make you happier if I was rude?

George

Vote

J

Joerg 18 years ago

Very sad. Why are they so wasteful? Just think about the scrap value of all those metals. I mean, here we have shady folks who steal an air conditioner out of grampa's backyard to net twenty bucks at the salvage yard and there the Navy just chucks a whole big ship.

Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.

Vote