using regulator to limit AC secondary voltage - A possible or reasonable approach?

[CharlieLaub]

i too like lots of caps in a psu....i still haven't heard from you about your intended load...
if you still like to lose some voltage, a pair of power resistors from the raw b+ to the collectors of the series pass transistors ought to do it...

This is powering a class-D amp into a 4 ohms nominal load.

I will look into your suggestion in the morning. I also tried a zener to the base of the transistor pair and this worked well at limiting the maximum rail voltage under no load conditions

 

[KatieandDad]

Could you not just put a large resistor in series with the primaries if it is only a small voltage that you are trying to drop.

To drop say 2%, you would need something in the order of 0R47 at 12W. This would also act to supress inrush current at turn-on.

 

[Elvee]

So, I happened to be reading the spec sheet for an LM317 the other day, and at the end there was this circuit:

If you choose to act on the AC side, you should do it in a meaningful way.

Phase control using a power MOS rather than a triac has a number of advantages: it reduces the current peaks and doubles the ripple frequency.

This example operates on the primary side, but could as easily work on the secondary:
http://www.diyaudio.com/forums/powe...-elegant-insane-way-converting-220v-110v.html

 

[KatieandDad]

Interesting. Almost like using a lighting dimmer to reduce incandescent lighting level. However, it reduces the power to the load. If used with a transformer the power throughput would be very reduced.

You could use the idea at high phase angles to reduce the 115V input to say 100V but you would have all sorts of issues with noise to contend with.

 

[sreten]

cap multipliers are fine to use with classA where load current is more or less fixed....

Hi,

Rod Elliot talks about using it for class AB in the article.
Reducing R12 to increase voltage drop is recommended.

rgds, sreten.

 

[KSTR]

In my enthusiasm for a "great deal" I have acquired several 500VA transformers that have a center tapped secondary that is rated at 62-0-62VAC with 115VAC applied to the primary. The mains voltage where I live in the USA typically runs at 120-122VAC, and you never know if it will rise another percent or two. This makes the secondary voltage "too hot" for most DIY amplifier applications that I am considering, so I'd like to figure out a way to decrease the secondary voltage without having to do any surgery on the transformer itself.

The transformer has dual 115VAC primaries and, while I could use the transformer with the primaries wired in series to get half the secondary voltage, I would also reduce the rated VA by half. I'd rather not have to go that route.
[snip]

Well, if those were standard non-potted toroids you could simply "unwind" the secondaries by adding a few turns wired in antiphase to drop 10...20Volts or so. I've done that several times in order to use transformers with sligtly too high voltage.
EDIT: Series'd primaries is not a bad thing, while you get a lower voltage and softer output you gain saturation headroom and this will quiet things quite a bit, especially at idle.

 

[Elvee]

Interesting. Almost like using a lighting dimmer to reduce incandescent lighting level. However, it reduces the power to the load. If used with a transformer the power throughput would be very reduced .

Actually no: it is the opposite: with a capacitively loaded rectifier, the transformer only actually works during a short conduction spike near the top of the half-wave. The rest of the time, the transformer is idle (except it spends uselessly V*s's).

With this scheme, there are two conduction spikes/half-cycle, which means a rms current divided by √2 for the same average output current.
In addition, the primary is spared a good part of the useless Volts*seconds, meaning it could have a lower nominal voltage.
For a given transformer, this more than doubles the original VA rating

You could use the idea at high phase angles to reduce the 115V input to say 100V but you would have all sorts of issues with noise to contend with

As I said earlier, the current draw of this kind of rectifier/filter is already far from sinusoidal anyway. Having a bit more or less distortion changes little to the end result.