Voltage Multipliers

[Eli Duttman]

Please look here. IMO, the last circuit on the linked page has potential. Sourcing power "iron" for tube projects can be problematic. I think a tripler version of that heavy duty circuit mated to an isolation trafo could provide for the needs of EL34s, 7591s, and KT88s. The 250 VA Triad N66A seems up to powering EL34 stereoblocks, 7591 stereoblocks, and KT88 monoblocks. The N66A costs approx. 63 USD.

Opinions please.

[tubelab.com]

I have used voltage doubler circuits to generate the high voltage needed to run 845's, 211,s and 833A's. The voltage regulation is not as good as a conventional bridge of full wave rectifier. This is OK for SE Amps.

I have searched for low cost power capable of running tubes of this type. Here are two of my secret weapons:

1) Allied Electronics www.alliedelec.com sells a few tube type transformers for a fair price. The biggest one, number 6K7VG puts out 750 volts CT at 150 mA with the usual 5 and 6.3 volt filament windings. It has enough juice for small stereo SE and P-P amps. It sells for $43. If you go for mono blocks, 50 to 70 WPC is possible.

2) for the really big amps, search Ebay for an "industrial transformer" or and "industrial control transformer". You want one that has a 480 volt primary, and a 120 (or240) volt secondary. They come in sizes from 50 VA to several KVA. Hook the secondary up to the wall outlet, and put a bridge rectifier on the primary. You can use a SS bridge, or one made with damper tubes. (I used 6AX4 because I have lots of them). With a choke input filter I get about 475 volts DC. With a cap input filter I get 560 volts under load. I am using a 500 VA transformer for an experimental 6LW6 P-P amp. I used a 1KVA transformer with a voltage doubler for my 845 amp. These transformers often end without any bids and can be purchased for $20 to $30 if you are patient. New they are about $100.

[DigitalJunkie]

I once used a power tranny from a SS amp (50-0-50,Used the whole 100V secondary,big 300W monster.) and a tripler to get about +400Vdc for a tube amp..I never finished the amp,but I tested the power supply a bit,and it seemed suitable for the larger tube amps.
The only drawback,no filament windings,and i'm not sure if tapping the bias from the secondary would work out very well.
Otherwise you'll need another tranny or two.

[Jeb-D.]

I've used them in some of my designs and built them as well. They work good in some circumstances, but in others it will become impractical. For driver tubes it's usually feasable because smaller value caps can be used. But for output stages the number of capacitors and sizes can get a little out of hand.

[smoking-amp]

500 VA Hammond xfmr 480 V - 120V @ $19.95: (no affiliation with any)

(these are actually more the size of what one would expect for a 300 W xfmr, but are wound with high temp wire insul. and rated for a big temp. rise at 500W)

http://www.meci.com/product_info.php...10b48f3d56a78b


Here is another interesting way to get HV ($5 each), these are SMPS with +50V and -50V outputs at 1 amp. The voltage regulation is wired to regulate the +5V supply, but can be re-wired (not trivial though, requires some resistor re-wiring on the control board) to regulate one of the 50V outputs. If the + and - 50V are connected in series and no power taken off the mid point, then both will regulate.

Use several in series to get HV (I think Lambda specs them for 400 V max float, best to check their web site info for P/N). Each 100V increment will be regulated for using lower voltage tap-offs.
(These supplies also need a minimum load on them to regulate properly, but one should put bleeder resistors on the HV caps anyway.)

http://www.allelectronics.com/cgi-bi...R_SUPPLY_.html


PS: I had the re-wiring worked out for these once, but lost it in my recent move. Its just involves tracing the foil from the UC3842 voltage regulator chip control pin, out to the little daughter card and re-doing the voltage divider pickoff to sample one of the 50V supplies instead of the +5V one.

Also of interest, these SMPS units have provision for power factor correction circuitry on the PC artwork, but is not installed. I have not been able to get the parts info from Lambda, but one should be able to dupe this stuff off a unit from Fleecebay if one ever shows up with the full deck installed.

Don

[smoking-amp]

Oh, I forgot. The SMPS supply has an overvoltage circuit on the +5 output which has to be either disabled, or modified to monitor the +50 V output. Best to sketch out all the monitor circuitry on the little daughter card.

These units are obviously designed for just about any voltage configuration, and all the control circuitry specific to the voltages is on the daughter card.

Don

[andrew_whitham]

Glad that I'm not the only one wondering about voltage doubling. I also want to use a "underspecified" voltage to run tubes.

BUT can anyone help me figure out the relationship between the drawn current, capacitance and voltage drop? I've found lots of info on the various configurations used to multiply AC voltage but precious little maths. Is this properly simple and I'm missing something?

If it helps I'm trying to acheive exactly what started this thread,
In my case 12VAC to get to +/- 30v. Was going to copy MF's circuit ala X-10D etc but I want to draw about 100Ma overall (well 80) so a multiplier is going to sag, right?

Has anyone got any handy equations?

[Gold_xyz]

Only a note...
on voltage multiplier if the electrolytic (polarized) cap are used
then they can degrade.
this effect manifests him to the power-on when the capacitors
are uncharged for a few microsecond the capacitors are reverse
polarized...

bye

[Eli Duttman]

Andrew,

ALL cap. I/P filter PSUs, including voltage multipliers, droop to some degree under load. LARGE filter caps. "stiffen" the PSU, but they introduce RF "hash" into the rail. So, a choke and additional capacitance is needed after the multiplier. Think of the arrangement as an extended CLC filter.

Am I correct in thinking you want a bipolar supply? If so, I hope you have a trafo with 2X 12 VAC secondaries.

[ilimzn]

Gold_xyz makes a very valid and vital point! When using voltage multipliers, it is ESSENTIAL to put anti-parallel diodes across the multiplier caps (all but the final output cap). These would be diodes that go from - to + of the electrolytic caps. Looking at a simple multiplier with a short on the output (which is what happens when the output cap is completely discharged), shows that the 'charge pump' caps get reverse voltage on them without the diodes, with diodes fitted, the diodes actually provide an initial path to charge the output caps directly with the trafo secondary voltage, and avoid reverse polarization of the charge pump caps.

A couple of other points about multipliers:

1) You can't get something for nothing, so drawing 100mA@300V from a trippler that operates out of a transformer designed for a 100V full wave rectified output, means drawing at least 300mA from the transformer secondary winding. This is obvious. What isn't is that the multiplying action relies on charge transfer, and since the voltages are unchanged on the caps, what must change is the ripple current. Charge pump caps are subjected to ripple current that is increased by the same factor as the multiplying factor of the multiplier, with consequences to the ripple voltages across the caps. Current x voltage means power developed on the internal ESR of the caps, which means heat, the biggest enemy of all electrolytic caps. Conclusion: use good quality caps!
IT should be noted that it is possible to include an inherent current limit mechanism into a multiplier by manipulating the capacitance of the charge pumping caps - just keep in mind that a prolonged fault condition that triggers this mechanism, is likely to severely tax, and if persistent, eventually destroy the caps due to large current and voltage swings across them.

2) Do NOT use taps out of a multiplier to supply lower voltages without carefull consideration that precisely equal current is drawn out of the transformer, during both half periods of the AC voltage, or a resulting DC flux will saturate the transformer core, ultimately ending in a burned transformer. In other words, avoid all forms of half-wave rectification or multiplication unless the power required is negligible compared to the power supplied by all the other windings of the same transformer (for instance, it is perfectly fine to use half wave multiplying or inversion for a grid bias supply).