Zen v5 non-complementary

[djk]

Go to:

http://164.195.100.11/netacgi/nph-Pa...&RS=PN/4229706

Then click on 'images', you need a TIFF viewer,

http://www.alternatiff.com/

The Sumo Model 9 was 120W/4R class A, at lower impedances it went into AB, and was 70W/8R

The supply voltage for the outputs was about 35V X 2 (floating), the opamp was an LF353 running on an independent ±20V

"You'll need to drive it from a preamp (such as Balanced Zen Linestage) that is capable of high voltage swing."

Depends on how close to the rail your opamp can drive, the LF353 wasn't real good at that and it worked fine. The point here is that 70W/8R is 67V peak-to-peak and the opamp running off ±20V was able to drive it to this level. I think you might be able to get about 100W/8R with ±24V for an OPA2604 opamp. The point here being you could get the amp up and running with a $2 opamp while figuring out what to do for the drive circuit 'for real'.

Also note that it would be real easy to change the feedback loops to be like the Nelson Pass X patent.

[Joe Berry]

Originally posted by djk
The Sumo Model 9 was 120W/4R class A, at lower impedances it went into AB, and was 70W/8R.

The actual claim was that this design stayed in class A all the time. This may be true technically, since there was a bias circuit (the crux of the patent, I believe) that kept the base current into the output stage from ever shutting off. But this is really analogous to class A2 operation in a tube, as it avoids cutoff but does not really enforce linearity.

The supply voltage for the outputs was about 35V X 2 (floating), the opamp was an LF353 running on an independent ±20V

FWIW, the schematic I have shows the floating supplies as 40V x 2. Unfortunately I don't own one, so can't measure the true voltage.

[djk]

"The actual claim was that this design stayed in class A all the time. This may be true technically, since there was a bias circuit (the crux of the patent, I believe) that kept the base current into the output stage from ever shutting off. But this is really analogous to class A2 operation in a tube, as it avoids cutoff but does not really enforce linearity."

By that deffinition all solid state designs would be A2, as they all have grid (base) current flowing.

"FWIW, the schematic I have shows the floating supplies as 40V x 2. Unfortunately I don't own one, so can't measure the true voltage."

The actual voltage doesn't matter, the point was that the front end does not require a huge swing as someone else was saying.

The Sumo Model 9 used the same transformer and chassis as their 250W/8R class AB amplifier. It was rated at 30-30-30-30V at 1.3KVA, no load this would be about 40V, but depending on how the bias was set up this will drop.

I have a 9 with the original box, manual, and schematic. It has the gold front panel.

[Michael Rothacher]

The Zirclotron has no valtage gain. It is essentially a power follower. You'll need to drive it from a preamp (such as Balanced Zen Linestage) that is capable of high voltage swing

By this I meant that a preamp driving the circuit (as-is) to full power would require more output than is typical of many preamps.

The Sumo, it appears, used two opamps (two halves of LF353) as an unbalanced-to-balanced converter each with + and - 20 volt rails for a maximum 40V pp in each half of the phase splitter. Wouldn't that be 80V pp?

I like your suggestion and I'll put a quick-and dirty front-end together and give it a go as my BOSOZ is probaly weeks from completion.

Joe: I've been thinking about the alternative bias you mentioned. Do you use a Vgs multiplier for each half of the bridge?

Thanks

Mike

[Joe Berry]

Originally posted by djk
By that deffinition all solid state designs would be A2, as they all have grid (base) current flowing.

Sorry, I guess the tube analogy wasn't too helpful. Maybe it would be better described as a bias regulation scheme that looks at base current rather than base-emitter voltage. If I understand correctly, its main goal was to prevent thermal runaway, with full-time class A operation as a side benefit.

Originally posted by mrothacher
Joe: I've been thinking about the alternative bias you mentioned. Do you use a Vgs multiplier for each half of the bridge?

I've done it that way. It works fine, and gives you a way to adjust the two sides independently. You can also use a single ground-referenced Vgs multiplier for both sides, and power it from the two floating supplies summed through resistors. You won't see much relative DC offset if the output MOSFETs are well matched.

There are lots of options here. One I haven't tried, but think might be interesting, is to use a single grounded Vgs multiplier powered from a 9V battery. You could put a switch in the battery circuit to activate a standby mode, in which the floating supplies are under power, but the output stage is unbiased (and the bias battery undrained) to conserve energy.

[Michael Rothacher]

Here's a simplified schematic of the op amp front-end described earlier in case anyone is curious.

Thanks

Mike