Downsizing ZEN4?

[Skorpio]

Will downsizing ZEN4 circuit to run with a +25V supply instead of the original +40V give any problems, besides the power output will be reduced to aprox. 10W?

[Nelson Pass]

More like 4 watts

[JCM]

I am in the process of doing the same thing. The power supply voltage that my circuit sees is going to be about 31 volts, down 13 volts from 44 volts unloaded.

I chose not to include the regulator, in favor of air core inductors in series with reservoir caps. That regulator, or if it's configured as a capacitance multiplier, will drop a few volts (4 minimum, I believe) to the circuit, which is the reason NP says "more like 4 watts".

I see no reason it would give any problems using a lower supply voltage. I've built a few of these circuits - not exactly as NP designed them, but simpler than that, and it is just a matter of scaling things up, or down.

[Skorpio]

Thanks for the replies....

What power do you get from your version? And what changes made? Perhaps a JFET follower?

[JCM]

>What power do you get from your version?

If there is 31 or so volts supplied to the circuit, it's going to give a good 28 V pp, there is no doubt of that, a good 12 watts or so.

Unfortunately, I haven't even started on it. I bogged down, and became stuck. I didn't want a regulator - even a capacitance multiplier, feeding the amp, being in a "purist" frame of mind. Now I'm wondering if that's foolish. Maybe the LCLCLC scheme I have in mind won't yield low ripple - although on paper it does - unless I'm in error. Still, there's something - unnatural- in my mind about all that wire - hundreds of feet of it - feeding the amp.

Also, I decided to go with switchable degeneration; but variable would be better. To accomplish that, I need to know the open loop gain of the circuit, to begin with. Then I started wondering if I could make an even simpler current source than the Aleph... looking at that little feedback loop in there ...

I spent a lot on all those parts - and still don't have heatsinks. I bought a great big fat slab of aluminum with the idea of making my own, which I began to have great doubts about.

CHANGES :

* Lower supply voltage

* No regulator or capacitance multiplier, LCLCLC instead

* Probably run the input buffer mosfet at higher bias current

* Use non-polarized capacitors for the 220 uF units

* switchable (or better yet, variable if possible) degeneration

* switchable current source mode.

* Would like to have a one-transistor current source

* 6800 uF output caps - I don't see the need for higher.

* MPSA06 transistor in place of the Zetex, because of higher volt rating.

I even started worrying about the use of a PC board - some swear point-to-point is audibly better - an assertion I find incredible. I wondered about trusting a pcb trace to carry 1.3 amps ...

It might be a good idea to create a kind of experimental amp, with lots of adjustments, in order to find out better how to best build a final version.

On one hand I think "Go with the proven N.P. design", on the other, I think of the spirit of the Zen, which is to show what's possible, not necessarily dictate any one design that must be adhered to exactly.

I may end up not with a Pass Amplifier, but a Rube Goldberg device, a Kludge - if I can just get off the ground

Temporarily stuck.

[Shoog]

I built a Zenv2 as my first amp and used a 24V transformer for a supply rail of 33VDC. Get about 8 Watts.
I added a cap multiplier which helped a lot. I also added a simple buffer circuit to the front end based on the Bride of Zen preamp, this improved the top end a lot and also made it more realistically sensitive. This is different to the Zenv4 which has the front end inside the feedback loop, where mine acts as a seperate buffer.
I would say that the Aleph current source is an integral part of the design and the bass will disapoint greatly without it.
I dragged it out of storage last week and have been listening to it for a full week, its got a softer more rounded sound than my Triode SET amp- very easy to live with.

Shoog.