Adapting "low" voltage power transformers for tube amps

[thorpej]

A legendary Silicon Valley electronics emporium is having a moving sale with by-the-pound discounts (yes, really), so of course I had to pop in today after work and poke around.

In addition to some test leads for my scope, for a song I scored 3 identical power transformers marked 9-80-120 that look to be from a MusicMan 1600B SS bass guitar amp. I haven't mapped them out yet, but the data on them I can find indicates the primary has 100V and 120V taps and the secondary is 67VCT with a secondary current capacity of 2-3A.

My Silvertone 1484 employs 2 voltage doublers in it's power supply. There are 2 secondary windings on the PT at 80V. Each secondary is connected to a bridge doubler comprised of 2 diodes and 2 capacitors. The two doublers are then connected in series to create the ~450V B+.

What I'd like to do with these MusicMan transformers is similar... and they have enough current capacity that driving two doublers off of them should not be an issue. They should be good for 375V @ 500mA (or 425V at something less than 500mA if I used the 100V primary).

I *think* I can get away with simply building two separate doublers in parallel (by tee'ing each end of the transformer winding) and then connecting the outputs of the two doublers in series. See attached schematic. In this configuration, I would leave the center tap of the secondary disconnected.

The only thing I'm unsure of is whether the inputs to the two doublers need to be isolated from one another as they are in the Silvertone power supply.

My electronics-fu is not particularly strong, so a sanity check from the AK experts would be greatly appreciated

(I figure, even if I can't do what I want with these things, they hardly cost me anything and I can still find something to do with them...)

 

[babblefish]

Can you share who this legendary emporium is?
Oh, nvm, found it: HSC Electronics.
Need to check that out. Like I need more stuff...

 

[thorpej]

Can you share who this legendary emporium is?
Oh, nvm, found it: HSC Electronics.
Need to check that out. Like I need more stuff...

Hah, yah, that is the place. When I first moved to California 21 years ago, I recall going to that place and being gob-smacked. No different when I went in today. I had to exercise serious restraint

 

[babblefish]

Lucky you weren't around here 30-40+ years ago. Electronic surplus stores were all over the place, some double the size of HSC. I remember in San Francisco alone, there were at least 4-5 surplus electronics stores. Sigh, they're all gone now. The reason for so much surplus was because the Bay Area had several military bases plus there were a lot of companies manufacturing electronic products where most of the electronic surplus/excess came from. And not just electronic parts, but raw material such as aluminum, titanium, SST, hardware of every imaginable type. Most of the military bases have long since closed and most of the manufacturing has either gone overseas or to a another State where tax incentives, labor costs, and cost of living is better. Unfortunately, some of the laws in California also makes it unprofitable and/or difficult to do business here.

 

[kward]

With a single secondary winding, consider something like this: (from http://www.electronic-circuits-for-hobbyists.com/voltage-quadrupler-circuit.php)


Also see this writeup: http://en.wikipedia.org/wiki/Cockcroft–Walton_generator. I used the Cockcroft-Walton approach (full wave voltage ladder) configured as a doubler in one of my previous amps.

 

[thorpej]

With a single secondary winding, consider something like this: (from http://www.electronic-circuits-for-hobbyists.com/voltage-quadrupler-circuit.php)
View attachment 672557

Thanks, I like this one much better, but I'm not sure I follow it... In particular, how does C4 get charged? And it seems like C3 only gets charged to Vp, not 2Vp. I totally see how C2 gets charged to 2Vp (C1 gets charged to Vp through D2 when the bottom is positive, and when the top is positive it adds to the voltage on C1 to charge C2 to 2Vp).

Also see this writeup: http://en.wikipedia.org/wiki/Cockcroft–Walton_generator. I used the Cockcroft-Walton approach (full wave voltage ladder) configured as a doubler in one of my previous amps.

The half-wave ladder is quite easy to follow, and with large enough caps, ripple shouldn't be a problem in a P-P amp.

 

[kward]

The upper and lower sections are each voltage doublers. By wiring the upper and lower section in series this way, the DC voltages from each section add, in the same way the voltage of two batteries in series add. (And there is a connection from the bottom leg of the transformer to the junction of D2 and D4.)

 

[thorpej]

The upper and lower sections are each voltage doublers. By wiring the upper and lower section in series this way, the DC voltages from each section add, in the same way the voltage of two batteries in series add. (And there is a connection from the bottom leg of the transformer to the junction of D2 and D4.)

Yah, when I look at the standard Greinacher doubler, it's pretty obvious -- On the bottom-positive cycle, the first cap charges to Vp though the first diode, and on the top-positive cycle, the voltages add and charge the second cap to 2Vp though the second diode.

Ah, I see it now. It's obvious what's going on if you orient each half in the same way as the doubler schematic. Super, thanks.

 

[thorpej]

Oh, since this operates as a half-wave rectifier, I suppose that for smoothing purposes I need to divide the AC frequency by 2 to figure out how large the caps should be, yah? For example, if I have a 365V B+ from a full-wave rectifier and a circuit that draws 120mA, normally that would call for a 50uF cap for 60Hz AC for 5%-ish ripple (I'm planning to use this in a P-P guitar amp). But in this topology, I only get Vpk once per cycle instead of twice per cycle, so now we're talking about a 100uF cap because it takes twice as long to replenish the charge. AND, since we're talking about 2 caps in series at the output, it's really 2 250uF caps.

Sound about right?

 

[kward]

I think that's the right idea. What I do in cases like this is estimate by calculating ripple for the full wave rectifier case (not voltage doubled). Then double ripple for every time you double voltage. So the ripple would be about 4x larger than the non quadrupled case.

OR--since this circuit is quite simple, just throw it into the simulator. Meeting your criteria of 150 mA load at ~365V B+, gives 5Vpp ripple using 250 uF caps.

here's a snap of my screen showing output voltage across the load: