• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

SMPS for B+ from PC PS

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all,
Just saw this article from 2004 showing how to convert a AT PC SMPS to regulated high voltage for use in Tube Power Amp
http://www.siliconchip.com.au/cms/A_102096/article.html

Only problem - you need to pay about $5 for access.

I just did and got the full article with schematics & description of mods

Quote from article:
"The circuit is capable of excellent performance. It maintains full regulation at up to 125W, with ripple at 2V peak-to-peak, or 0.3% at full power. This is quite acceptable, as most of the ripple is at twice the switching frequency (60kHz) and so is inaudible.
The 100Hz hum component is only 0.08%, which shows the excellent regulation of the TL494, since the rectified mains source contains 13% of 100Hz ripple at full power. Over-current protection is retained, with a LED added to indicate when it is active."

Final schematic below.

Has anybody tried this? How well does it work?

John
 

Attachments

  • hv power supply - power section.jpg
    hv power supply - power section.jpg
    38.8 KB · Views: 820
Ping Eva to chime in -- or N-Channel as they are the SMPS guru's

one thing I would be very careful of is the insulation value of the secondary wiring -- this is going to be very high -- and a lot of insulation of the windings on the transformer. specify double insulation on the high voltage side (according to Sanjay Maniktala's "Switching Power Supply Design and Optimization"
 
Considered making a new thread but since it's already here, although not in the power supply section, i decided to revive this one.

Can anyone provide more detail on modifying ATX PSU for HV B+ as done by SiliconChip (transformer winding ratio, etc)?
The online article is no longer accessible.

Schematic above shows a quite a few additional parts but this capture from siliconchip looks like it's still stock ATX PSU except for toroids. Does the schematic use the parts already on the PSU PCB?
An externally hosted image should be here but it was not working when we last tested it.


Basically, i would like to know how to modify ATX Computer PSU to provide HV B+ and 6-12 volts for heater.
 
Looks like you have to take apart an existing SMPS T1 transformer, and remove the secondaries (count the number of turns it takes to make the +5V and the 12V so you have an idea of the turns per volt you'll need for the new secondaries).

What I've done is obtain a second T1 from another similar SMPS and connect its 5V secondary to the main T1 of the power supply you want to modify. Usually there's a pair of diodes inside a device that looks like a transistor (kinda like the solid state version of a 5U4 though lower voltage) fed by a centertapped secondary, the centertap going to ground. Connect the ends of this winding (but not centertaps) of the first T1 to the second T1, so this transformer's primary becomes the secondary to give you B+ once you rectify it. And you must use fast diodes (1N4007's won't cut it).

You'll need to figure out which output is more important to regulate. There's really only one feedback loop in these supplies, most look at the +5V output and control the switching accordingly. A higher load on the +5V would then make the B+ output rise, as the controller compensates for voltage sag on the +5V to get it back to +5V.

Or you could find and disconnect the circuit that monitors the +5V output (usually a precision resistor voltage divider) and use a much higher resistance resistor so the new B+ will produce the same voltage the controller saw with the old voltage divider off the +5V. But the +5V output will not be that well regulated, but you might be able to get it to make 6.3VDC (but be aware that these supplies sometimes include a shutdown circuit if the +5V gets too high, you could fool it by making another voltage divider so it thinks the 6.3V out is 5V). Tube heaters are not that fussy, they can take about 10% error or so.
 
Yes that is exactly what they did. They disassemble the transformer, rewound for high B+ and 12V (for heater) and replace the secondary parts accordingly. I found that out after i bought the online mag.

Thank you, you just confirmed what i was wondering. Connecting transformers back-to-back is possible even with SMPS. I thought since it runs at very high frequency, the extra inductance load might pose a problem.

Another idea i had is to modify Ebay Boost Converters to reach higher B+. I made a thread here. Perhaps you can help me with this? I think this is easier than rewinding the ATX transformer, provided that it's feasible.
 
Last edited:
Can you share exactly what did you do to the Boost Converter?
My plan is simply to upgrade the diode and output filter caps to higher rating and changing the ratio of the voltage divider feedback to get higher B+. If you did this already, then obviously my plan isn't going to work.
 
I can't get it to work decently, that is, the regulation is quite poor. I tried to replace the 12VDC regulation feedback loop with a resistor to act as a voltage divider from the new B+ to make it look like 12V (this 12V circuit has an 18K load, so I used a 22oK resistor between this and the new B+. But no dice, no load on the B+ and I get 150V, load it with a 10 watt lightbulb and it sinks to 107V. If I can't handle a 10 watt load, then this B+ isn't going to be much good for an amplifier. I can do much better with an old fashioned 60Hz power transformer... Maybe there's too much inductance in the circuit with the 2nd transformer? I'd say my idea has a serious flaw.

Maybe you can disassemble the main transformer, but they are usually dipped in some sort of varnish or lacquer or some such, and you'll probably ruin the ferrite core prying it apart.

What I did was use a half wave rectifier on the 2nd transformer, a filter cap, and I had to phase the connection between the 1st transformer 5V source ( the one inside the SMPS) and to what was the 5V source on the 2nd transformer. Get it backwards and regulation gets really bad.
 
Last edited:
What I did was use a half wave rectifier on the 2nd transformer, a filter cap....

back-to-back works, but here's your flaw:

with a half wave rectifier the 2nd transformer's core will saturate as soon as you load it, so you have to use either a 4-diode bridge or a full wave doubler; I built a circlotron amp like that even w/o regulation on the B+ with acceptable sagging; circlotrone needs 4 independently floating B+'s so I have actually 4 secondary transformers connected ! and the 12V and 5V tuned to 15V / 6V for heaters.

if you try to regulate to B+ (or better a mix of B+ and heater voltage) you have to put a choke between the rectifier and the cap and take the regulation from behind the choke; if you look at ATX ps schematics, those pwm controlled half bridge / forward converters have a common mode choke behind the rectifiers to improve cross regulation of multiple outputs and (!) another individual choke on each output. You can possibly do w/o the 1st common choke but the 2nd choke is mandatory if you want to regulate from B+. Otherwise pulse width regulation of B+ won't work (obvious if you look at the signal shape before the rectifier).

in case of the 6V/12v use for heaters it may also be necessary to put a NTC in series, otherwise the 7x inrush current of 4 sweep tubes plus a couple of smaller ones is possibly going to trigger the overcurrent protection
 
Last edited:
Something else I found out was that if this high voltage B+ output gets shorted, it will take out the primary side switching transistor!

Before I blew it up, I was getting 250VDC no load, with 10K load that would drop to 225V, and with 2K it dropped to 120V. No feedback loop to the regulator control circuit. I ran simulations in Circuitmaker and got it to reproduce the results I had when I tried voltage doubler circuits (which made for really bad regulation, even no load I'd get 30VDC...). The simulation, and in real world, I got the best results with a half wave rectifier, and a CLC filter circuit. 3.3uF, 1.7mH, 3.3uF.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.