Stacked power supply HELP!

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Ok, I've been racking my brain and researching this for several days and I could really use some help. I am building the Simplex SE 6C33C based circuit by Ari Polisois but I wish to use tube rectifiers in the power supply. The DCMB design requires a stacked power supply in the end producing +270V for the OP stage and +450V for the driver stage. The +270V PSU is "stacked" on the +450 PSU, so the + side of the 450V supply is connected to the - side of the 270V supply.

Ari, derives these voltages from two separate transformers and they are each diode rectified before being connected together. It is my intent to accomplish the same thing using a single transformer with two separate windings (one for each supply rectified by a 5C8S for the +270V and a 5AR4 for the +450V. The issue then is, what happens with the center tap for the floating supply? Can the B+ from the 450V supply be connected to the transformer center tap for the 270V supply? If not, how would one establish the - side of the +270V supply?

I've attached a sketch of what I'm trying to accomplish to hopefully clarify.

Any help would be greatly appreciated.

An externally hosted image should be here but it was not working when we last tested it.
They way you have it drawn will run, with the added requirement that the transformer have at well over 600V rated insulation if you expect it to have a very long life. Unfortunately you no longer have a 270V supply unless the load is referred to the 425V output. The way you have it drawn you a 425 and a 685 volt supply with respect to ground, and it wont do very well to load the 685 output to ground beyond 50mA because the 425 volt supply underneath it all will be totally overloaded long before 250mA. There's a decent chance you need to flip the series with the 270 volt supply referred to ground to get what you want out of it.
Thanks for responding. The load connected to the top 270V supply is a single 6C33C connected to the high side via the output transformer. The cathode of the 6C33C is then connected to the - side of the 270V supply which is also the + side of the 425V supply.
Here is Ari's original power supply for reference. All I'm changing is valve rectifiers instead of solid state diodes and using a little bit different filtering.

I see. In this circuit the 6C33 power tube is supplied from your high current 270 volt supply and it is indeed "floated" from the high voltage signal/driver supply. You should be good to go.

(sorry if you caught this message before my edit, it looked scarier than it is at first glance from a signal perspective)
Last edited:
I think the thing that was really concerning me the most and was unsure of was connecting the B+ of the bottom supply to the (floating) center tap of the top supplies transformer windings, with both windings being on the same transformer core, but as you said, if the transformer windings are well enough isolated from each other it shouldn't cause a problem. If this was a problem I could use two diodes and re-create a virtual center tap (hybrid rectifier) so as not to stick 425V on the center tap maybe.

Thanks for your help Andrew :)
Well unfortunately any way you re-work the circuit it still demands that the 6C33 has a stable supply with the cathode referred to the plate supply of the driver in order for proper bias and drive, so you can not get around elevating that power winding up to that level. It isn't intended to be "understood" that a power tranformer with two relatively high voltage windings (or any, actually) absolutely can withstand being run in series with any combination of rectifiers or multipliers. That is mainly up to the end user to guarantee! 700V is a lot, hopefully the outut transformer can hold it off with very low leakage too. I would definitely check that with a mA meter from secondary to ground when you get it running. If you see more than microamps you might want to avoid the butt end of the speaker or wires when it's live. (or tie one end to ground for safety) ;)
Last edited:
Yeah, I see your point. In Ari's design he specified an op transformer with the ability to withstand 450mA. I've got a pair of Eastern Audio 600 ohm transformers that I was going to use but something tells me they're not up to this challenge. My original intent when I purchased them was to build the Borbely 6C33C but I became intrigued with the DCMB design. The deeper I get into this design though, the more I find it slightly terrifying, because of the stacked supply.
Fear of 700 volts can not be considered psychosis. You can make this design work with common parts and research to assure they are compatibly designed and built but I wouldn't expect any healthy accidents(.) If you build it I would highly recommend grounding the "bottom" of the output transformer winding so you can be reasonably assured all that tension stays in the box.

Do you have manufacturer's data on the transformer parts?
Last edited:
Indeed! If nothing else I will end up with an effective device for tazing elephants. :p

Grounding the O/P transformer is a good idea.

If you're referring to the power transformer, the plan was to have Edcor custom wind one, I'm not sure of what their insulation ratings are. I would need to e-mail them again and find out.

If you're refering to the output transformers, I dont have much data on them at all. I bought them on ebay cause they were dirt cheap, so I don't really know what to expect from them. These were the specs that were given on the auction:


Class : Single ended design and have air gap

Primary impedance :600 Ohms

Secondary impedance : 0-4-8 Ohms

Frequency range : 15--50K Hz

Output : 50W ( maximum )

Maximum DC current :300MA

Part of me wants to just use these transformers and build the Borbely circuit and save the Simplex circuit for later, the curious side of me really wants to hear what DCMB can do.
Wow, I've been pondering a similar 'stacked power supply using a single power
transformer' idea, and here is this post.

So it's OK to stack the supplies if there are two separate windings on the pwr xfmr, and if the windings are sufficiently well isolated from each other to withstand any leakage currents. But what if you try to derive two supplies from a single tapped winding, and then try to stack them?

The pwr xfmr I'm thinking of using is an old UTC LS-76. For HT, it has two taps on a single CT winding, one is 830VCT @ 270mA, the other 790VCT (same current rating). The center tap is the same for both, according to a simple resistance test.

If I do not use the CT on this winding, but instead use two SS diode rectifier bridges (each made of four 600V Cree Schottky diodes), I can define the relative grounds where I want them. But there would be no isolation between the windings, as they are taps off the same secondary winding. So my question...

If I need about +450VDC at 50mA, I could get this from rectifying the 790VCT taps with a diode bridge. I could ground this diode bridge to chassis earth. That would be A+.

The I'd need an additional +450VDC at 200mA supply using a diode bridge from the 830VCT taps, let's call it B+. I'd need to float this supply on top of the A+ supply -- The B+ diode bridge would be grounded to the A+ supply's +450VDC.

A+ would measure +450VDC to ground.
B+ would measure +900VDC to ground.

- The lower current supply will be closer to ground.
- The higher current supply will be farther from ground (on top of the lower current supply).

Does the higher current supply draw current through the lower supply? Or does it use the floating ground as a virtual ground, and its current is drawn from there? Don't the electrons all come from Earth, so the top supply's pool of electrons MUST travel through the bottom supply?

I've used this stacked supply approach with two separate power transformers, and it's worked perfectly well in an amplifier that's been running for at least 5 years. But what if you try to stack supplies with the same power transformer? I'm afraid I haven't been able to figure this one out. (My lack of knowledge is probably the root cause of that.) My hunch is that the currents of one supply will flow through the other supply, and they would interfere with each other. But I'm sure I'm missing something...

So long as you don't keep any secrets from the custom house you probably wont have any problems there.

Of course the auction jobs are totally up in the air. You can get some idea through inspection and testing. Before you ground the output put an ammeter on there to ground and get a baseline leakage measurement so you have some idea where the insulation stands at this point. Can you post a picture of the OPTs. (without covers if any) For 1kV isolation they should have nice insulation and plenty of margin on the windings if no bobbin.
Last edited:
>But what if you try to derive two supplies from a single tapped winding, and then try to stack them?

That's what a common center tap split supply is all about.

I'm sorry, I'm not following you... I could see how you could easily derive two DC supplies with a common ground from the two windings with the common center tap. But if you ground the center tap of the secondary winding, now both windings are grounded at their center tap, correct? No way to float one supply over the other that way. Is that what you were referring to?


Rongon: think of how you would use that CT winding to make a symmetric power supply (+V and -V) using half-wave rectification (a single diode per rail).

Now, instead of choosing the usual center tap as the ground, choose the V- as the ground. Now you have a stacked power supply from a single CT winding :)

Of course, half wave rectification kinda sucks :( Also, if both halves do not draw equal current then the core will see a DC current which some transformers do not like.

My thoughts exactly. I've been planning on having to spring for some of Ari's OP transformers developed specifically for this circuit. I bought the Eastern transformers before I talked myself into this whole Simplex situation. I'm still considering building the more conventional Borbely circuit since I already have all the valves for it and I think the ebay transformers would be suitable for it. I'm about to get married too, which means my budget is tight, so it may make more sense (for now) to go with the simple situation and reserve the Simplex for next winters project. I certainly don't want to abandon the design by any means.
Thanks Andrew! Confirms once again that common sense should prevail.

I hope the OP's OPT's work out. I've noticed that my various old Stancor and Thordarson chokes have insulation ratings of 1kV -- even the open frame ones are rated at 600V and up. The comparable current production Hammond chokes have max ratings as low as 300V. That's a little scary...

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.