The Quaditron...?

[Circlomanen]

DEEP EXPLORATIONS IN CIRCULAR TERRITORIES!

I wanted to create a pushpull version of Son of Zen, but dont like "complimentary" P and N channel Fets. Soo this is my idea!
Any and all feedback, thoughts or whatever, is very welcome.
This drawing is without feed
back and bias circuitry to make it easy to understand. Its just the basic idea.

[Circlomanen]

This is a more complet drawing. Probably a little harder to grasp at first sight. The inputs are cross coupled. positive input is one on the top section to the right and the one on the down section to the left, and the opposite for the negative input connections. The feedback is crosscoupled to.

[Nelson Pass]

Very imaginative.

[tschrama]

You're making me dizzy ....

somehow it looks so simple that I wonder why noby else came up with it, on the other hand, it somehow slips my mind to ready undetstand it...

[GRollins]

I did something similar...called it a "Double Cross." My version had a front end, used both N & P-ch devices, and was X once I got the front end straightened out. Lost the schematic during one of my computer adventures a year or two ago, but might be able to locate a hard copy. If you use N & P devices, you'll only need one (bipolar) power supply.
Note that the output is high Z, and as such will have limited application; high efficiency speakers and possibly tweeters. I dropped the concept and went on to other things. Since then I've thought of a thing or two that might make me want to revisit the concept, but it will be quite some time before I get to do so.

Grey

[Circlomanen]

If you use N & P devices, you'll only need one (bipolar) power supply.
Note that the output is high Z, and as such will have limited application; high efficiency speakers and possibly tweeters.

N and P devices is a tradeoff in a circuit that depends on close matching!? I rather have two floating supplies than P Fets.
Why high output-Z? I think the feedback will lower the output-Z a lot. With low gain (4-8X) the output-Z should drop down to 1-2 ohm region!?
Se Mr Pass article about the Son of Zen V.7 :http://www.passdiy.com/pdf/zen-v7.pdf
This circuit should have more gain and with feedback lower output-Z.

[Tom2]

Circlomanen,

IMHO
Your circuit appears to be two differential amplifiers in parallel. The signal currents coming out of the drains to the node with the load resistor and supply connection would seem to cancel (add up to the value of the CCS). Thus the resulting current flows back into the supply (it has to because the supply is in series with the CCS) and no current would pass through the load. In other words calculate Kirkoffs current law for that node and the current through the load will be zero.


Tom

[Circlomanen]

Originally posted by Tom2
Circlomanen,

IMHO
Your circuit appears to be two differential amplifiers in parallel. The signal currents coming out of the drains to the node with the load resistor and supply connection would seem to cancel (add up to the value of the CCS). Thus the resulting current flows back into the supply (it has to because the supply is in series with the CCS) and no current would pass through the load. In other words calculate Kirkoffs current law for that node and the current through the load will be zero.


Tom

Yes!! I came to the same conclusion on my way home from work!
To look at the bright side of things: 100% losses means no distortion and happy neighbours!

[GRollins]

As far as N and P devices, match N devices to each other and P devices to each other. This is no more than you'd do anyway, since you're using them in a differential configuration. The Ns and Ps do not have to be matched to each other.
The final output impedance will depend on the amount of feedback, but will in any event be higher than a follower output stage. It's not necessarily a deal-killer, just something to keep in mind.
I was looking for ways to replace the load resistors in the SOZ and hit on the idea of using an opposing pair of Drains, which led to P-ch devices and so on. There are advantages and disadvantages. As usual, it's a question of tradeoffs. I ended up with (I think) three variants on the idea before I tore it all down and did other things. I'm on the verge of having enough stuff that I'll be able to leave amps assembled once I'm done playing with them, in case I want to go back to them, but not there yet. In particular, heatsinks are always in short supply and get reused endlessly.
I believe I've remembered the original input stage that I used (which Nelson saw), but still can't pull together the versions that followed. Haven't had time to go looking for printed schematics--been playing farmer, getting tomatoes etc. in the ground for all the veggie-loving folks to eat over the summer. Got the sunburn to prove it. And the fire ant stings, too. (Those who say that fire ants 'bite' are incorrect. They are members of Hymenoptera, like my bees, and use stingers to get their point across...)(Ahem)

Grey

[grataku]

Isn't this design a little redundant? With two phases 0-180 ( + and - inputs) you are kind of limited to differential-type design.
Now if you could itroduce two more phases to the input signal the 90 and 270 than you might have something different but this is audio so I don't thik it'll work.