Amazing Circlotron with Power JFets

[Ejam]

I was about to build Mike Rothacher's Amazing Circlotron amp when the news of SemiSouth's power jfet came. Liking the simplicity of Mike's amp and Nelson's glowing over the power jfets got me thinking about using then in this circuit. My question is about the biasing of the power jfets. From what I've read, normally that they are biased by a resistor in the source but a what Id and what Vgs etc. The curves on the data sheet are very expansive if you get my meaning. So can anyone help with a suggested operating point and bias scheme. Thanks in advance.

Ejam

 

[flg]

Personally, I always thought the Circlotron was a pair of followers. At least the versions I've seen. Not being amazing then But you might do a search on the Grey Rollins Pwr JFET version. I beleive it was called the Nuetron1 but I did a search on that and didn't find it

 

[Michael Rothacher]

Ejam:

Send me some of those bad boys (SemiSouth SICs) and I'll be happy to answer your question empirically.

 

[Ejam]

Just a few clues please!

Mark

Yeh, I love to but at $60 each, have to save up a few more pennies. So what do you think about biasing, Mark? And Nelson, what operating point do you use?

Ejam

 

[Nelson Pass]

The circlotron can be either Common Source or Common Drain.

I bias the JFETs into Class A - if you are spending that much
for a part, you might as well go for the gusto.

 

[flg]

The circlotron can be either Common Source or Common Drain...
,,, you might as well go for the gusto.

Nelson, Does the original 1950's (or one of the other 3,4,5 attempts etc?) patent cover that, or only the later ones. Or just our ideas? I don't care really but I just had this thing about Mark's version being a different one??? Not the Circlotron I liked And actuallly a cool adaption of the F5
I was working on a MOSFET version of such a thing, that was not stable, enough when Grey came out with the LU1014D version of Nuetron or whatever.
It seems to be the same ol' deal though. You can cancel the evens but the odds are left??? get rid of the THD before play canceling games...

 

[Michael Rothacher]

I don't believe I claimed it was different, or novel.

The name is "tongue in cheek" since the article asserts there is no magic. It's also a bit of Roger Corman marketing.

In my mind, it is indeed F5's Circlotronic cousin, which I have said as often as possible.

It sounds great.

 

[Michael Rothacher]

Just a few clues please!

Ejam:

I don't think they're available to hobbyists at any price right now, so we'll have to wait a bit.

As NP said, Class A for sure, but I can't offer any clues on the best biasing arrangement. I have to actually build stuff to understand it.

When they become more widely available, I'm sure we'll see all sorts of application ideas.

 

[flg]

... Not being amazing then But you might do a search on the Grey Rollins Pwr JFET version.

Must be this comment you are refering too mrothacher? I was refering to the grollins version being the "not amazing" version. I thought Ejam might want to take a look at it... Sorry if I ruffled some feathers.

 

[Michael Rothacher]

but I just had this thing about Mark's version being a different one??? Not the Circlotron I liked And actuallly a cool adaption of the F5

I thought that was a question. I was clarifying. There's no rufflage.

There's nothing new about the Amazing FET Circlotron. It, and other circlotron's are push-pull amplifiers, with push-pull cancellation.

On another note, I have found a possible exploit, but that must wait for another time, after much experimentation.

 

[woody]

The audiocircle forum had a thread about using the IXTH20n50d in a circlrotron. This is a depletion mode mosfet with a 400w 20A rating.
What is appealing about this device is that it's about 1.7A at 0V on the
gate and ~100ma at -1V on the gate. On the down side digikey wants
about $6.80 each with a mimum purchice of 60.

 

[Michael Rothacher]

Cool. Do you have a link, Woody?

 

[Michael Rothacher]

Ah, here's the schematic:

http://i700.photobucket.com/albums/ww7/paulhynes/CIRC1.jpg

Paul is driving arrays of 8 Visaton B200's wired in parallel for a total load of .75 Ohms and 105 dB sensitivity.

 

[Michael Rothacher]

To further clarify my clarifications:

Even when operated common-source, the circlotron doesn't contribute voltage gain. It's still a follower. This may be the cause of some confusion. (Would it be a drain follower? Is that even a word?)

This accounts for the distortion performance differences between The FET Circlotron and its F5 cousin. F5 has more open loop gain since both stages contribute voltage gain. The FET Circlotron has less gain available for feedback, and the distortion is correspondingly higher. Of course, there's also less feedback and a balanced input, and you might view that as a good thing too.

 

[Michael Rothacher]

@ Ejam

I pulled up the PDF from 6 Moons, and, as a starting point I guess I'd pop them (SICs) into the circuit as-is, adjusting P1, and dispensing with R7 and TH1.

Did I mention I smoked a few circlotron prototypes before getting it right?

 

[Nelson Pass]

Even when operated common-source, the circlotron doesn't contribute voltage gain. It's still a follower.

What is it following?

 

[Michael Rothacher]

Input (gate) voltage? He said sheepishly.

 

[Ejam]

Positive Temperature Coefficient

Mark

I don't think that removing R7 and TH1 is a good idea given that the SJEP120R100 has a positive temperature coefficient.

Nelson as to bias, I was looking at the curves and given a requirement for 1.2A the Vgs is pretty close to the gate threshold voltage of 1V. Is this a problem?

Regards

Ejam

 

[Michael Rothacher]

You seem to have it in hand.

 

[Michael Rothacher]

...Of course we needn't limit ourselves to 1.2A