DC Filament with Choke

[megawhat]

Hello,

I was reading Merlin's power supply book and I thought Id like to give one of the DC heat methods a try.

my PT has a 6.3v/9A wind.
the AC with run 4x EL34s, then 2x 12AX7's.
Then the full wave bridge rectifies the signal. (KBU404G bridge rectifier)

current my amp has this, followed by a 12000uf/16v cap smoothing the DC before V1 & V2. a faux centre tap is after the cap in parallel with V1&V2.

I'd like to try the method using a choke (not for anything specific, just so I gave it a go).

After reading the book, I understand I should split the 12000uf cap into 2 ~6000uf caps (still low esr high ripple) to go either side of a common mode choke. The choke should be mili-henrys in size but I don't know what to choose or if I need to consider anything when wiring it up.

I then plan on attaching the faux centre tap to a DC reference to elevate the supply.

anybody else tried this?

[Gaz]

That's cool if you wanna try the DC heaters, but is the amp otherwise noisy? With only 2 12AX7s, I can't imagine there's a lot of gain in the pre-amp, and am guessing it's an Orange style amp??

I'm not sure about the choke rating question, but have you researched elevating DC heaters? I'm assuming you cant elevate DC with DC, that it would not only raise the heaters to the elevated voltage, but be redundant as far as noise reduction is concerned. Hopefully can answer you question with some confidence!

[megawhat]

the amp will have 4 12ax7s.
It doesn't exist yet, but part of the reason for building it is to try new things like making a regulated voltage for switching and also this idea above.

I've used normal AC and DC elevated in the past... i just want to try other stuff really.

I didn't think it was possible to elevate a secondary if part way along it was being rectified... but im sure Merlins Blencowes book mentions it...

it would be 6.3vAC wind with no centre tap > 4xOctal Sockets + 2x9pin on AC > Bridge Rectifier > Smoothing Cap > 2x9pin on DC > 100r+100r faux centre tap.

so in that, the faux centre tap is in parallel with the valves run on DC, rather than across the winding with the raw AC. Blencowe puts forward the idea of using a humdinger where i mentioned the centre tap, and also the idea of elevated using the wiper of the humdinger. That isn't the circuit i was asking about though, i just have that working in an amp that i didn't build which i reverse engineered.

the common mode choke would split the smoothing cap and sit between two caps.

[martin manning]

The partial DC heater scheme has been done as you describe here:

[megawhat]

yep, ive seen that before. A version of that is in the amp I own, but the cap is 12000uf.

but the faux centre tap after the BR goes to ground. Not to a DC reference which the Blencowe says is possible.

It also doesn't have a common mode choke it in, which is what I was originally asking about. I've attached the schem from the book. I hope Merlin won't mind... for those that want to read his book http://www.freewebs.com/valvewizard/index.html

It says you split the cap value... so two caps close to 6000uf/16V. It just says the choke should be milihenrys, but how many? and I do i need to consider the current going through it to the 12AX7's

[martin manning]

...but the faux centre tap after the BR goes to ground. Not to a DC reference which the Blencowe says is possible.

Sure, why not?

It also doesn't have a common mode choke it in, which is what I was originally asking about. I've attached the schem from the book. I hope Merlin won't mind... for those that want to read his book http://www.freewebs.com/valvewizard/index.html
It says you split the cap value... so two caps close to 6000uf/16V. It just says the choke should be milihenrys, but how many? and I do i need to consider the current going through it to the 12AX7's

Merlin is suggesting that this common mode choke be home-made using "a few turns of wire around a ferrite ring," so evidently the value is not too critical. As you can see from the Marshall schematic, the smoothing cap doesn't need to be nearly that big, but it will reduce the residual ripple and help keep the rectified DC voltage up. I own one of those amps, and it is very quiet. That circuit is going the extra mile for studio use, where most guitar amps just use AC. With good lead dress and a ground reference (elevated or not) heater hum isn't an issue.

[megawhat]

Cool.
thanks for the help.

Good lead dress etc... I agree with that. I like using elevated AC if I have cathode followers. I like the circuit as it is in that marshall, I fancy trying the ferrite ring bit. I'm told the Morgan Jones book has some details on this.

One thing that worries me about this type of circuit, is power tube shorts. I had a friends Trace Elliot over recently and it was pretty broken. I fixed it, but it looked like an EL34 had shorted the HT around itself and onto the filament, burning out a bunch of connector and small resistor it has between the CT and ground on the filament. I imagine if that happen in the marshall or my Soldano then the bridge rectifier would burn up.

[martin manning]

FWIW, you can buy small commercially-made power inductors and common mode chokes. Without doing all the math, 5-10 mH should work nicely with 4700uF caps. The current rating of the choke would have to be at least equal to the current required by the filaments running on the DC side, but a drawback to this whole idea is that the small series resistance of the inductors will reduce the DC voltage quite a bit. A high current rating will help there, but you may have to be satisfied with running a low voltage on the DC when the AC is correct, unless you have separate transformers. With all that, a DC supply using an LM7812 3-pin regulator would be just as easy, and it could run several 12A_7's on 12VDC using a 12.6 VAC transformer.

[megawhat]

Yeah, I don't think I have the extra space for a filament transformer. Otherwise I'd run most of the preamp on 12V.

I want to run 3 relays too.
Originally, I thought I could make a 12vDC supply to run the relays and 2 12ax7's.
The problem is finding an off the shelf PT with 360-0-360, plus a 15v secondary with enough current to run what i want and a 6.3v tap big enough to run 4 EL34's and however many preamp valves.

I can find a PT's with enough high voltage and current on the main secondary, and then say two 6.3v winds at 9A and 2A. So I planned on using 5v relays off a regulated supply from the 2A secondary, and then heaters on the other 9A one.

The LM812 looks like it has 1A max output, so it would be ok with 4 12ax7 on 12v.
The other issue I've read is when using the same DC supply for filaments and relays, switching noise can get pinged into the valves.

I think ideally I'd need a bigger chassis and a huge PT with various windings to have AC on the power valves, separate supply for preamp, separate supply for switching.