diytube.com

Hi folks,

I'm planning out a 300B SE build, and I have some confusion about something.

I've seen some folks that recommend using two independent rectifiers for each channel so as to minimize voltage drop. Right now I'm planning a 5U4GB, and I'm estimating around 125mA per channel. Based on pen and paper calculations as well as PSUD, I gain 55V or so by running independent rectifiers, which seems like a pretty big win for the relatively minor cost of a second rectifier.

What I'm unsure about is if I'll need a separate filament winding for each rectifier, or if they can share one filament supply. Seeing as they're directly heated rectifiers, it seems like if they share a filament supply they'll be effectively paralleled, and I don't know if there's potential for undesirable interaction or "fighting" (uneven loading or whatever).

I'm planning on using an Antek toroid for this build, and I already need to buy 5V filament transformers for the rectifier as well as the 300B filament supplies, so it really just comes down to three vs four filament transformers.


Thanks,
Ian

Here's a schematic of how I imagine it would work if it's kosher to tie the two rectifier cathodes together with the same filament supply, but with two filter chains:

Hi,

You need seperate filament supplies or they're just paralleled rectifiers.

Cheers!

Is there any downside to them being paralleled other than potential channel crosstalk?

Also, does running them with separate filament supplies provide better crosstalk, even though they're both drawing from the same 400VCT winding?

Crosstalk with a good C-L-C-L-C power supply will be below the tube noise level.

Seperate filament supplies reduce crosstalk further. It's nothing to toss in an extra 5V, 3A Hammond for the other tube.

If crosstalk is a grave concern for you, I suggest monoblocks.

What's your source? If you're a digital fan, monoblocks may be most desireable. If a turntable, best seperation is 40dB anyway from a moderate-high priced (~$500) cartridge (MC or MM).

Cheers!

I'm mostly a vinyl guy.

To be perfectly clear, my goal here was not channel separation; my goal was to control the voltage drop. I can get as much as 465V if i use two rectifiers, but it's closer to 410V with a single rectifier. I just was unclear if that was the correct way to do it, or if there would be trouble.

Also, the design for this amp is getting somewhat ludicrous with transformers! At final tally, I need: Power transformer w/ 2x 6.3V filaments, either three or four 5V 3A filament transformers, four chokes, two outputs, and a partridge in a pear tree :D

I built a big mama 300B as my first totally scratch built amp over 10 years ago. Plenty of iron !



I used a single 5Z3 or RCA 83 Mercury Vapor recifier and have almost exactly 475 volts with either tube,
using the Hammond power transformer. I don't remember the model number off hand.

I've never used dual rectifiers, except for damper tubes. I have built many amps with filtering for
each channel. Works well, you can use smaller cheaper chokes.

right now I'm trying to decide what I want to do for output transformers, and if I decide the right answer is to wait for edcor to make me some CXSEs, I may see if they can also make me a custom power transformer, 800VCT @ 250ma, 6.3V @ 1.5A, 5V @ 1.5A, 5V @ 1.5A, 5V @ 3A, 5V @ 3A.

If voltage drop is your primary concern, and you also are wary of using too many transformers, have you considered using solid-state rectification? I just built a 45 SET using solid-state rectification (and an Antek transformer) and it works very well — I used a separate filament transformers for each 45, with a separate switch to turn them on. This way, I can warm up the 45s before switching on the B+ (warm-up is one traditional reason for using tube rectifiers). To reduce noise, I bypassed each diode rectifier with a 10nF ceramic disc cap. After the choke, the power supply is very quiet.

(I posted about my 45 amp just a few threads down from this one in diy hifi).

Also, it’s possible to use one or both of the Antek 6.3 volt filament taps to power your rectifier tube(s), if you decide to go that route — just use a dropping resistor to get the 6.3V down to 5V. If you’re using the big 400V Antek, the filament circuits are rated for lots of current. The resistor will get quite hot, though (5.2W at 4A).

I believe that you can run two directly-heated rectifiers off one filament circuit, especially if crosstalk isn’t a concern. I don’t believe anything bad will happen.

125mA for 2 channels I could believe - 125mA per channel with parallel se 300B I could believe. 125mA per channel for a single 300B is NOT right.

Parallel rectifier tubes might be beneficial if indeed current draw was 125mA per channel.

I have a 300B project ready to go too. (Tubes, Chassis, Trannies sitting on the shelf).

I'm seriously considering Rod Coleman's Shunt Cascode driver circuit
http://www.bartola.co.uk/valves/2013/03 ... de-driver/

If I were going "the whole hog" design (money no object) I woukld use Rod's Filament regulators too.

Cheers,
Ian