JR, Please don't get me wrong, I'm all for experimenting, and I think it would VERY educational to explore the differences in power supply construction and performance. This is what this hobby is about!

Wayne, is it possible to set up a group build forum, not only for this project but future ones as well?

We need to finalize what we're going to build and how we're going to relate information. One thought: it might be fun for each to build slightly different variations and send them to one of the club meetings (Tulsa?) for direct comparison. This would allow each variation to be evaluated head-to-head under "identical" conditions.

I know I keep saying this, but I cannot tell you guys how excited I am about this. Hands on anything is quickly becoming a lost art. I think it's important to keep this hobby (and skill set) alive......Regards, Colin

There are a few interesting things to try.

First the minimum reactance PS. It will require a CCS to execute. Let's try two CCS as active plate loads and do a simple 6CA4>>L>>100uF Oiler/motor run.

One or two red LED in series will do a fine job of cathode bias. we are running absolutely constant-current through the valve.

Simpler PS, and Cascoded DN2540N5 current regulators have a low output mu-follower output. These two benefits ought to be enough to justify building a second CCS. 400V D-S. Full wave from a 150-0-150 plate winding will keep things safe 'til the cathodes heat up, and the PS turns properly regulated L-C.

Choke from Hammond, 30 Hy/40 mA. PS cap, eBay's midorimeadows has ASC 370 VAC/100 uF for $10.

On to coupling caps. 5-10 uF motor run oilers are good here. No sense making output z too high at low frequency.

Here is the extreme example of where this design can be taken if you need some time to relax, and building is the chosen method. This one is L-C to a pair of CCS, feeding a pair of VR150's for each channel, which is actively loaded 2C50.

It sounds better than it looks, it only took a weekend to build.
regards,
Douglas

....damn Thrintun!

I was thinking probably a very simple circuit with a rectifier and LC filter to start with. About the regulated supplies, have you ever doen one that was all-tube? I'm not opposed to solid state at all, but it is kind of cool to be all tube. Makes for a lot of real estate though, I imagine.

Wayne

The Amplifier active loads are not so easy with a pentode. The g2 circuit in parallel with the plate is the issue( as I see it now ). Sooooo, how about a cascoded pair of triodes? Series curent path and high AC impedance, and still able to retain the mu-follower's signal output. It would probably be best to use a cascoded MOSFET CCS to generate the Ec2, but that would be th extent of it, and that is two DN3545N3 and three resistors...

I have heard another all valve, regulated linestage and it was really good. this should be able to match or better its performance with a simple signal path like my hybrid.

I have another in the planning stage to use the 10Y as the amplifier valve, and this PS and load idea is how it will get done. It will have a few more valves sticking up through the top, but that's OK with me.

You heard it here first!

Of course I'd keep the 816-rectified, L-C unregulated supply...
regards,
Douglas
!

The vlave shunt regulator has cool out the bazooo from a string of 3 0D3's. This will want a HV/unregulated supply of ~600V and that IMO, is pushing it, let alone for entry level.

The 1kV CT can be done quite easily with damper diodes like the 12AX4 or 6AU4 depending on your available filament supply. A delay is needed for an LC like this and dampers are just the ticket. Doesn't glow like an 816 though...
regards,
Douglas

....damn Thrintun!

It is for this app: 12B4A amplifier valve, 15 mA current, biased with my favourite red LED's. 6BK7/6DJ8 half at the bottom of the cascode, sharing its filament supply with the 12B4. Half a 5687 on the top ov the cascode, with its own filament supply. The 5687 cathode is ~250V and that's the reason.

Four valves, not including the rectifiers, and requiring a B+ of ~450.

I feel sad, it is about Christmas, and if I take time out to build this now, I'll get shot. Probably after getting my doo-dads squished in a big vise.

As far as parts go, the Hammond 193C 20 Hy choke ought to do for the power inductor. a 500-0-500 is 'up to you', I have at least two of them. L-C-R-C DC filament supplies are not too much trouble.

If this works as well as I think it will, I'll have to lay out the Green for a nice pair of 10's...
regards,
Douglas

....damn Thrintun!