Easiest way to reduce rails.

[Eva]

I suggest simple series pre-regulation based on a couple of 60V 60A mosfets, they are cheap and very easy to drive if you pick the right topology [source as input, drain as output, with floating transformer windings for symetric regulation]

If supply drops to +-75V at full power and you adjust output to 80V then dissipation in the regulators will be very low as they will be fully switched-on most of the time

I've used this stupid technique to power 12V car ICs [max. 18V] with 17V regulated [up to 28V before regulation], using cheap transformers and small filter capacitors, to get cheap and simple 25W/channel PA amplifiers for 100 Volt lines [transformer coupled outputs]

[K-amps]

Originally posted by Tube_Dude


But here nobody pay attention!!

The regulators are for the input stage!!

The guy want increase the rails from 80 to 100 volts and maintain 80 volts for the input stages..

Thanks Jorge, you got it. I should have been more clear too. I want to increase the power of the amp BUT keep the input circuit voltages the same i.e. around 80v. Is there no real cost effective way of doing it without resorting to building a separate PSU or installing regulators? All I need is a 15 volt drop from the main rails.

Eva, any sample schematics I can review?

The amp in question is an Adcom 555. the trafo is 700VA or so and the AC is pretty well regulated, i.e. secondary voltages drop only from 60vac to 56 vac or so. It is the DC side that drops hard.

Each channel has 30,000uF caps which is not huge but its not puny either.

What I am trying to do is use it to drive a couple of Woofers and need about 60 volts RMS produced by the output stage. I reckon with +/- 100 vdc rails I can get that.

Any more ideas are welcome!!!

[K-amps]

Originally posted by millwood
your subject line seems to contradict the body of your message. I assume that you are trying to raise voltages on a power amp (that would then rule out those 317 type linear regulators).

if the rail drops from 100vdc to 62vdc under load, you are using too small of a transformer. 10% of regulation is probably max. 40% is unheard of. Mine does about 4% at rated output.

Sounds like you have overloaded your transformer. Using large caps will help but not much under this type of load / capacity imbalance.

Millwood, I realize the subject is a tad misleading but hopefully not contradicting, as Jorge points out the current situation is that under full load, the DC sags from 80 to 62-3 vdc. But the trafo drop (before rectification) is only from 60vac to 56vac. So it seems like the Trafo has decent regulation...

What I am trying to do is get more power by jacking up the output stage voltages to 100vdc and drop the same to the input stages. Thereby keeping it around 80vdc at the input stages and let the output stage rails go up and down with load.

Again my aim is not to have a rock solid 80vdc at the input stages but merely to drop 15-20 volts from the main 100vdc rails...

I want to keep it simple for my prototype.

thanks in advance.

[sreten]

there are lots of options to lose the 20v for the low power parts,
the difficult bit is maintaining the 80V under high load.

Simplest is using RC filtering, dimension R according to the
class A standing current in the input stages to give a 20V drop,
you then need to add enough C to maintain the voltage high
enough under high output load transient conditions.

Simulation is probably best, but I'm sure 470 uF would work well.

edit : add a series diode to prevent the C reverse
discharging into the main supply capacitors.

edit : from the sound of it a seperate small rectifier with the RC
filtering would also help matters, here you need CRC filtering.

sreten.

[Eva]

You should take into account that this voltage drop in the power supply is sometimes used as a trick to make amplifiers work with 4 ohm or 2 ohm loads without exceeding the SOA of the output devices

I have seen lots of PA amplifiers whose no load supply voltages are way too high to be reliable, but low VA high V transformers and smaller capacitors are cheaper and protect the amplifier when working full power at low impedances

It would be a great idea to check SOA of output devices

[K-amps]

Agreed EVA, the originals are the old trusty Toshiba D424/B554 pairs but for 100vdc rails, I plan to use MJ21193/94 pairs. About 4 pairs per channel. It should work.

Sreten:

A schematic would help me a lot.

thanks yet again all you guys.


K.

[K-amps]

Sreten:

I re-read your post. It makes a lot of sense for my application.

Question: would a lower R value shorted by an 80v Zener help regulation? Also the C, I am thinking to install much larger values like 2200uF or so.

One would think the more the merrier but then they would take longer to charge up and who knows what kind of secondary issues I have with oscillations and/ or power on thumps because the Adcom does not have a delay circuit.

Your thoughts?

thanks!

K.

[lucpes]

Read these for power supply design. It will cover up most of the issues you may deal with:

http://sound.westhost.com/power-supplies.htm

http://www.tnt-audio.com/clinica/ssps1_e.html