Voltage Multiplier & Heater Power Supply - Question

[kagliostro]

Few time ago one of our australian friends was planning the PS for a small tube amp

He had a 50V CT PT that he used to build the power supply and to feed the tube heaters

he decided to connect the two heater of the tubes (in series) between the CT of the transformer and GND, the extremes of the windings were used with a voltage multiplier to obtain B+


https://www.guitargear.net.au/discussion/index.php?topic=51922.15




--

I remembered that Mr. Merlin discouraged the use of a voltage tripler (with a transformer having a CT) because this design will result in an unbalance of the windings load and in magnetisation, this schematic remembered me such a configuration

Our friend however build the PS as he planned and didn't had problems (he says it was due to the available current on the windings of the transformer)

--

However a question remained on my brain

Thinking to how is obtained back-biasing

https://www.aikenamps.com/index.php/what-is-back-biasing



and how acts a Cockcroft-Walton voltage multiplier






I've draw this thing



Will this circuit be functional and spread equally on the whole windings of the PT the load ?

Many Thanks

Franco

[jjasilli]

I'm thinking that there should not be a floating B+ supply.  Also there is no (-)path for your proposed B+ supply.  The (-)path should be ground, at the tail end of the diode bridges.  But then the CT should be NC; no ground connection.

To keep the the CT ungrounded, the Valve Heaven circuit has a floating heater supply (I think??); also not ideal. 

Note: Aiken's back-biasing circuit, sourced from the PT's CT, does not use diodes in bridge formation. 

[PRR]

> Will this circuit be functional

I can't make it work in sim, even guessing where "-" is and what heater and load resistor to use.

Get a 12V CT transformer, add 33r resistors on both legs, then wire the D+C. Use reasonable "heater" and load. Can you get useful DC output?

[kagliostro]

Many Thanks JJasilli & PRR

So ..... not feasible

EDIT:

Referring to this Philips PS



I've think to this different version



Will this be feasible ?

Franco

[jjasilli]

Minor point: looks like the outer series caps are wrong.  There should be 3 in series, both top & bottom, like in your 1st sketch.

Back to grounding/short-circuit.  See:  http://www.valvewizard.co.uk/bias.html  It seems the  Cockcroft-Walton voltage multipliers must be either true FWB's; or, alternatively, disguised pairs of two-phase rectifiers. If the former, you can't ground the CT.  If the latter, the CT is grounded (through a resistance), but then you can't also ground the the tail-end of the diodes. 

Hence I doubt the feasibility of this.

[kagliostro]

Ciao JJasilli Thanks for the reply

Reading more on the matter, initially I've find this article

https://josepheoff.github.io/posts/diode-capacitors-volts-pt4

and noticed that the first bridge didn't had the two capacitors as to be connected to AC (it is a simply FWB no doubler)



reading more the author talks about Parallel Cockcroft-Walton stages)

as to achieve a better stability on the output voltage



this remembered me about Mr. Merlin disquisition about voltage multipliers

(Page 66 & 67 of Designing Power Supplies for Valve Amplifiers)





and from this (the version with CT on the transformer) I go back to the Philips EL6416 PA PS where there is a bridge with the minus connected to ground and the CT is used to withdraw half the HT



 

Franco

Franco

[jjasilli]

Good references!  It seems that if you combine the JRE's Project Blog sketches with Merlin's capacitance coupled rectifier Fig. 3.14.a., then you must have C3 & C4 before the 1st diode formation, or there is a DC short across the PT.  JRE doesn't have that issue, because he is not using a CT.  See:  https://josepheoff.github.io/posts/diode-capacitors-volts-pt3 mid-page.


Caveat:  I don't really know this stuff; just trying to give you a sounding board.

[trobbins]

K, going back to Grant's schematic - the two secondary winding sections each have an AC current level in them - the winding with the heaters across it has a higher level of current than the other section, due to the resistive loading of the heaters.  As long as the secondary winding can cope with the higher current level then that circuit arrangement is fine, with no issues that I can see except that the heaters are 'elevated' to 50% of HT1.

Any configuration that tries to use 'back-biasing' can only work if the current in the CT that is being used for back bias is also the current being consummed from the main B+ supply.  So if the series heater connection needs 150mA to allow the heaters to perform properly, then the B+ needs to effectively class A idle at 150mA loading.   That applies if the B+ supply is a simple full-wave rectifier, or has doublers/triplers supplied from each secondary half-winding.

The Philips arrangement of full-bridge rectifier, and CT line providing a solid 50% B+ rail was typically used for their amp circuits using TV valves where a solid screen voltage was needed and a 50% level was a neat fit.  You need to make a simple schematic of that circuit to appreciate that the 50% CT filter cap is actually being full-wave rectified using just 2 of the full-bridge diodes (the 2 diodes connecting to ground).

[kagliostro]

Many Thanks for the replay and the effort to give help JJasilli & Trobbins

Franco

[isaac_teller]

http://www.bunkerofdoom.com/lit/4x8/index.html

[jjasilli]

"Figure 3 How the doubler works. The bridge rectifier section is familiar and needs no explanation."    http://www.bunkerofdoom.com/lit/4x8/index.html
I could use an explanation.  1. The input & output of the bridge are in the "wrong" places.  2.  Are both the bridge & the voltage doubler circuit grounder through D7 & D8?

[PRR]

> input & output of the bridge are in the "wrong" places.

I don't see this problem.

However overall: my mind is baffled and my sim does not get the results indicated. Except oddly if I mis-copy the circuit it sometimes does better.

As a practical man: if I can't understand it, quickly or tediously, then I can't fix it when it breaks. So to my mind it is more suited to the Brain Puzzles column of Popular Rectifiers magazine than to actually build into something.

[jjasilli]

I don't see this problem.
OK, I now see that the bridge is rotated 90⁰ to the left compared to usual depiction. I did say I needed an explanation. 

[PRR]

> the bridge is rotated 90⁰ to the left compared to usual depiction

You have to read schematics like every part may be flipped/flopped/turned shrunk/enlarged relative to real life, and wires are "rubber band"ed to make connections.

Yes, I do object to some drawing styles, and sometimes have to re-draw to fit in my small mind.

[jjasilli]

> the bridge is rotated 90⁰ to the left compared to usual depiction

You have to read schematics like every part may be flipped/flopped/turned shrunk/enlarged relative to real life, and wires are "rubber band"ed to make connections.

Yes, I do object to some drawing styles, and sometimes have to re-draw to fit in my small mind.

Yes, I should have known better having already read this:  https://en.wikipedia.org/wiki/Topology_(electrical_circuits)#Bridge_topology