Question about Rectifiers

[dunner84]

Hello,

I am building a multi-channel tube preamp, and the transformer secondary is 350-0-350 which is way too much for 7 12ax7 tubes. Is there any reason why I cant leave the secondaries floating (ground the CT), and build a half wave rectifier like a tube bias supply, and string my supply nodes off of this?
I'd like to use this transformer due to a few peculiarities, and because it fits perfectly in the chassis where space is a big issue. If it won't work out, I can easily source some Hammond 229 series transformers, but I'd like to explore my options with the iron I have on hand first.

[sluckey]

I am building a multi-channel tube preamp, and the transformer secondary is 350-0-350 which is way too much for 7 12ax7 tubes. Is there any reason why I cant leave the secondaries floating (ground the CT), and build a half wave rectifier like a tube bias supply, and string my supply nodes off of this?

You can do that. However, the dc voltage will still be about the same as using a conventional full wave rectifier. I suggest using full wave just because it's easier to filter.

[Latole]

You can reduce B+ voltage with zener diodes stack. I used it some times.

https://robrobinette.com/Generic_Tube_Amp_Mods.htm#B+1_Voltage

[pdf64]

Other than the necessity for a high voltage reservoir cap, I don’t see the issue?
Just use resistive droppers in the HT feed to the tube circuits, same as the preamp supply in a regular amp.
For a given HT decoupling cap value, the bigger the dropper resistance, the better filtered the HT will be.

[dunner84]

Thanks for the feedback. The high voltage e-cap is my main reason for looking for an alternative. I wanted to explore some creative options before stringing e-caps together. Not only does it take up space, they are the most expensive components in my builds.

[sluckey]

350 x 1.414 = 494.9

If you use that PT, whether 1/2 wave or full wave, the first cap will need to be rated for minimum 500V, be better to stack two 100uf/350v caps (still pretty small).

The attached pdf may be useful.

[Latole]

  I wanted to explore some creative options before stringing e-caps together.

 Why not zener stack ?

Oh because you never hear about that you are afraid

[sluckey]

Why not zener stack ?

Oh because you never hear about that you are afraid

Why use a stack of zeners when a single resistor will do?

[Latole]

Why not zener stack ?

Oh because you never hear about that you are afraid

Why use a stack of zeners when a single resistor will do?

Because a single resistor will not do !

[sluckey]

Why not zener stack ?

Oh because you never hear about that you are afraid

Why use a stack of zeners when a single resistor will do?

Because a single resistor will not do !

Oh yes it will.

[Latole]

In my opinion the zener by regulating the voltage will keep the B + more stable than if we were to use a resistor.

The resistance may vary the voltage depending on the current demand ?

Or I'm wrong ?

[sluckey]

In my opinion the zener by regulating the voltage will keep the B + more stable than if we were to use a resistor.

The resistance may vary the voltage depending on the current demand ?

Or I'm wrong ?

What you're saying is right but who needs it when a single resistor will do. The load will be 7 12ax7 tubes. That's pretty light and ain't gonna be varying. It's a simple job for one resistor. Do you know Ohm's law? Here's a simple explanation...

     https://www.fluke.com/en-us/learn/blog/electrical/what-is-ohms-law

And here is a simple calculator to calculate voltage drop resistor. You don't even need to know Ohm's law to use it.

     https://elecurls.tripod.com/drop-res.htm

[thetragichero]

i only ever use servers if i absolutely have to. have repaired too many scorched low voltage boards after the zeners failed short (to be fair, I'm certain modern zeners are much more reliable but i still avoid)
resistor in the filter string will also improve power supply filtering, no?

[glass54]

Hi Dunner
I believe simplest is best.
First if you use a Zener Diode you must decide on Topology ie if its going to be used as a Shunt regulator (as proposed by Latole ??).
This requires a Series Power Resistor (Rs) to drop substantial Voltage and a High powered Zener combination (eg 4x75V to achieve say 300V at suitable current, maybe up to 20mA for Load and min 5mA for correct Zener operation) We know that there are 7off 12Ax7 you are running, but we don't know exact load current. This design leads to substantial heat due to required zener power rating (15Watts), you can possibly check the maths.Then we have total Load and Zener current going through the yet unknown Rs. You must decide if you really want a Zener regulated PS.
Secondly you could use smaller Rs (drop some of excessive voltage from raw B+) and a series Zener to further reduce the B+ to desired voltage. I suggest a (sacrificial) resistor Rs in case there is a smoke incident so that your Transformer stays safe. Then you simply add RC stages for B2, (Second Node,your new preamp main supply) and B3 etc.

Alternatively you could do a multistage RC, reduce the excessive voltage after the first Capacitor via Rs and follow on with appropriate RC stages as required. Assuming the 7 off 12AX7 draws 20mA, and the supply is a nominal 500V, and if you required 300V preamp supply, your Rs power rating would be approx 4Watts (P=200x.020A). OK you don't get a regulated PS BUT the 7 tubes will draw a fairly low constant current and with a few RC stages before the main tube supply, there will be little PS hum. See attached. Simple and as previously suggested, will work and already tested on a past project of mine 
Kind regards
Mirek

(By the way, Zeners do generate noise and are best used with a paralleled cap  )

[glass54]

Correct thetragichero
You could always add another resistor before the reservoir caps on my drawing (say 10k 3W MO) to improve filtering and adjust Rs as required. just depends on whats in your kitty box  (or even split Rs)
Kind regards
Mirek

[PRR]

The problem is the 490V surge before the tubes warm up.

It is easy to dummy-up with series resistors so the *running* voltage is under 450V. But in the first 11 seconds it really wants to go over 450V. I tried with a large resistor *before* a large first cap, not enuff. You need near 1000uFd and several K of resistance to squeak by. (And if the 12AX7s fall out, it's still gonna smoke.)

[DummyLoad]

In my opinion the zener by regulating the voltage will keep the B + more stable than if we were to use a resistor.

The resistance may vary the voltage depending on the current demand ?

Or I'm wrong ?

adding zener regulation will generate a LOT more noise than a simple RC network. also, the larger the drop across the zener the greater the noise. for preamp only, personally, i'd walk down another path. 

IF it's going to be a high gain pre-amp, you will want to use a choke (pi) filter for low PS buzz/noise - BTW what's your target B+? you don't mention that.

another option that could make use of your 700VCT transformer would be to make use of a choke loaded pi filter - with the choke load filter full wave, your B+ is .9x Vrms of the 1/2 secondary - so 350V x .9 = 315V - perfectly acceptable for pre-amp supply, just pointing out other options.

caveats if you do run choke loaded filter: you'll want to load the inductor with a large-ish 10-20mA-ish (depending on L) load so the choke load is not seen as totally capacitive load during warm-up (initial B+ could reach in excess of 500VDC otherwise).

if you want really, really clean B+ the LCRC filter could be supplemented with a capacitance multiplier - search youtube for some tutorials on capacitance multipliers. take care to not get the time constant of the C multiplier too high, or you could end up with a start-up delay that is much longer than you'd like.

--pete

[thetragichero]

i was thinking about a choke input supply but wouldn't that require a rather large choke? at that point I'd start looking at a different power transformer
if you can deal with ~250vdc b+ something like a Hammond 269ex would be perfect for this (there are probably others that suit your needs better, I've just had one sitting around for six months that i finally found a project for

[Leevi]

One way to drop the voltage is to use a power Mosfet. It can be set to a wanted voltage level.
Of course it will generate heat and the good cooling must be arranged.
/Leevi

[glass54]

Back to PRRs comments regarding peak voltages:
In the project I mentioned above, I had an unloaded DC of +389V and settled for +311V DC Preamp Supply. It was a nominal secondary 260VAC, 6VA transformer (recycled  ) and I only had 1x12AX7 and 1x12AU7. (Separate transformer for DC regulated heaters)
I needed to dump only approx 80V, so a little easier than say 200V or more. The project is still going after 14 years.
So PRR, I take your point and its probably better to match (choose) your transformer for the requirements at hand.
Sometimes the cure (for excessive voltage) is worse and more ugly than getting a very suitable transformer.
Kind regards
Mirek

Just noted Leevi's comments and I agree. (Adds a little complication BUT great flexibility)

[Leevi]

I have met several times issues with either too low or too high B+.


The specifications of the transformers do not always tell what voltage level you will reach
since there are many variables that affect the result like:
- the capability of the transformer, there are big differences depending on the manufacturer
- rectifier tube, SS
- power tubes
- ...


/Leevi

[Bieworm]

The zeners will also cost a fortune when higher wattage is required.  Not to mention all the extra drilling, heatsink...