Regulated Supply for Direct Coupled Amp

[korneluk]

Folks,
Look at the schematic, does it make sense to use a regulated supply in this design?

http://www.goodsoundclub.com/Site_Im...oupled_SET.JPG

This design came from the 1947 edition of practical Amplifier Diagrams
http://www.pmillett.com/tubebooks/Bo...20Diagrams.pdf


I see two approaches:

1. use a regulated B+ for both stages in place of the existing supply.
2. use a regulated supply just to feed the 6SF5 by removing R7 and adjusting the supply to the correct voltage at C2, and tapping the B+ directly after the choke of the supply and before the regulator and connecting to C3.

Feedback and comments welcome,

[SY]

For a circuit like that, unless you're going to regulate the filaments and heaters as well, it's probably not worth it.

[korneluk]

Originally Posted by SY

For a circuit like that, unless you're going to regulate the filaments and heaters as well, it's probably not worth it.

OK then, say I used a wel filtered DC filment supply on the 6SF5 and a regulated heater on the 6B4G (alternatively replaced by a 2A3), what then?

[SY]

You'd still want the heater supply to be regulated (since the output tube's operating point will depend on the 6SF5's plate voltage). Then the output tube's filament would need to be driven from a regulated AC source- I think Guido Tent is selling a supply like that, or you could design a high stability power oscillator to do it.

[Chris Daly]

I see C3 paramount and C2 and R7 as importantl for audio performance particularly with regard to frequency in this direct coupled design. PSU looks honest and reliable so no need for too much added work. The task will be replicating the PSU transformer and output transformers.

C3 needs careful consideration to avoid high frequency loss.

Cheers / Chris

[korneluk]

Originally Posted by Chris Daly

C3 needs careful consideration to avoid high frequency loss.

What's your recommendation?
Cheers,

[Chris Daly]

20uf total using parallel non polarized film types ie 5x 4uf at the require voltage
I look forward to hearing how it sounds.
Cheers / Chris

[Sheldon]

I'd configure it as a Loftin/White.

No need to regulate at all. All film caps, and no cathode bypass. But with a DHT 6.3V filament, I would use a DC supply.

Sheldon

[kmaier]

For keeping the 6B4G you'll need to go DC on the filament as pretty much everyone I've seen has dual triode sections with the grid and plate paralleled and the DHT filaments in series, i.e., not possible to us AC and have it quiet.

Swapping to a 2A3 could resolve that but only it you use the later (and much more common) version which is dual triodes wired in parallel (including the filaments). With good (2A3) samples you can achieve s/n close to 80dB (ref. 1-watt) with an AC filament supply. I would also switch to a balance pot to null the filament hum better.

For the 6SF5, the current flow thru the tube is ~0.5ma, so a small fluctuation in filament voltage won't result in much change. I wouldn't bother with a DC supply if you have good 6SF5 tubes.... mine have been quiet for 4+ years of operation in a very similar circuit.

As the overall current variation from the circuit is minimal (class A operation) regulating it won't buy much but using good quality film caps and over-rated iron will provide good regulation and low impedance.

For component selection, I like Caddock, TF020 for voltage gain stages, MK132 for higher wattage requirements and MP820 for power stuff like the cathode bias on the output stage. Caps.... currently using the older Axon High-Volt film caps.... no longer made but if you can find them, a good choice.

Regards, KM

[korneluk]

Originally Posted by kmaier

For component selection, I like Caddock, TF020 for voltage gain stages, MK132 for higher wattage requirements and MP820 for power stuff like the cathode bias on the output stage. Caps.... currently using the older Axon High-Volt film caps.... no longer made but if you can find them, a good choice.

TYVM for the recommendations. I have some Axon caps but only for one monoblock, so will likely use an oiler for the first position - C4.