regulated power supply for high powered amps.

[marchel]

Hi,

Before this site closes for renovation, I wish to ask if you know a circuit for a fully regulated power supply that could output more or less about +/- 70V and 10ampere.

I also would like to ask if it is worth the effort if I use fully regulated power supply for 200W per channel amps.?

Thanks in advance.

 

[h_a]

You can go for the LT1083 regulator IC; guaranteed 7A, typical 10A output current.

As it is a low-dropout chip, power dissipation stays reasonable. However don't forget that if you really want to exploit it fully you need to handle ripple before the reg and toroid regulation as well.

Have fun, Hannes

PS: some more details:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=124346

 

[AndrewT]

200W and +-70Vdc supplies sounds like you intend to drive 8ohm speakers.

Maximum output to the speakers will be ~56Vpk, you might get 60Vpk or slightly more on short transients.

Peak current in to an 8r0 resistor will be 7Apk.
Peak current into an 8ohms speaker could be around 20Apk and severe reactance speakers could demand peaks even higher than this on fast starting/stopping transients.

If the supply is rated to 10A, where will the extra current come from to drive the speaker?

 

[marchel]

Andrew and Hannes, Thanks to both of you.

@Andrew,

Will adding capacitors after the regulator enough to take care of the peak demands of the amp.?

@ Hannes,

If I build parallels of the circuit you posted , How many of it in parallel do you think is able meet or exceed the demand, of a single 200W@8ohms amps driving a 4ohms load full blast in a home environment?

Thanks.

 

[AndrewT]

Will adding capacitors after the regulator enough to take care of the peak demands of the amp?

yes, but you will then be running the amp off the caps and throwing away most of the advantage of the regulation.

 

[analog_sa]

If I build parallels of the circuit you posted , How many of it in parallel do you think is able meet or exceed the demand, of a single 200W@8ohms amps driving a 4ohms load full blast in a home environment?

I may be missing something but how are these parallel regulators going to meet your voltage requirements?

If you look at a practical implementation which still falls short of your requirements, like the Naim NAP250, you will notice that the regulator is similar in complexity, parts count and power dissipation to the amp.

It may be arguable if the increased cost and complexity are not better justified by building a better amp.

 

[marchel]

I may be missing something but how are these parallel regulators going to meet your voltage requirements?

If you look at a practical implementation which still falls short of your requirements, like the Naim NAP250, you will notice that the regulator is similar in complexity, parts count and power dissipation to the amp.

It may be arguable if the increased cost and complexity are not better justified by building a better amp.

No, you did'nt miss anything , But I did.

I did a further reading about it and it turned out it does not meet the voltage requirement, It seems to me that discrete design is the way to go.

I also did a reading about using regulated power supply for hi power amps, And it turned out that it would be very impractical an very expensive if the amps is to rated to drive lower impedance with reasonable safety margin on the regulator. And aside from that there is also some sonical disadvantages of regulated power supply for ampli application, I dont think I want to go that way anymore.

However, There is something to be gained by using separate regulated power supply for the input and drive stages. Does anyone here know a schematics for low amperage 70-0-70V regulator?

Thanks.

 

[h_a]

Hi marchel,

you're not a close reader, are you

I did a further reading about it and it turned out it does not meet the voltage requirement

The LT1083 does meet your requirement, it's all about voltage differential. And you're right, it's an expensive chip. However caps cost money as well.

Wether regulated power supply hurt sonics or not is your personal preference. With a little care to details both can give excellent results. As you find out by googling easily, there are classic amps with and without regulated output stages with high reputation.

For your frontend you can use the lm3x7 series regulators, they're regularly used in commercial stuff (when the high end manufacturers once do not use 7x15 regulators ). You find an exemplary schematic in the datasheet as well.

If you want to go complex, there are lots of discrete regulators as well, however to get their real performance you need to have a little more knowledge.

Have fun, Hannes

EDIT: oh and please take a lot of care working on a +-70V amp; these rails can really hurt.

 

[marchel]

Hi Hannes,

Your right , I'm not a close reader, I misinterpreted the current limit over voltage diffrence on the datasheet as the absolute limit of the operating voltage regulator IC itself.

I'm more of a speaker and passive preamp DIYer than electronics, SO I'm not so well versed in this area. Nobody's perfect

Regarding the schematics you posted on the link you've given, If I were to use it on a power amp that have + and - rail, I noticed that the lower regulator's input and output is connected to the 0V or ground .

Questions:

Does this mean I can't use a transformer with center tapped secondary, Where the center tap is connected to ground?

Will the circuit necessitate the use of separate windings or 2 separate transformer for each (+and-) rail?

 

[analog_sa]

Hi marchel,

you're not a close reader, are you

Apparently nor am i

Curious though: can it really output hundreds of volts if the differential limit is met?

 

[wg_ski]

Curious though: can it really output hundreds of volts if the differential limit is met?

Yes, but with caution. As long as Vin-Vout is within safe limits any regulator can regulate any voltage. The trouble is keeping it that way under all conditions - including start-up and current limit conditions, where you could try to force a larger volatge difference across it. The app notes usually show what extra circuitry needs to be added to keep this from happening.

 

[richie00boy]

Yes and so can the LM317. However, in practical reality you must limit the input voltage to the max specified because as soon as you put a cap on the output of the regulator, at startup it holds the output at close to zero volts for a short time which will cause differential limits to be exceeded if your input is too high.

It may survive for a while but soon it may well go pop.

 

[analog_sa]

Excellent point. I was also wondering about the output cap.

 

[h_a]

Well, simplest would be to use an appropriate Zener across the regulator to clamp the differential voltage (power-Zener!).

In that case I would definitely recommend to limit current inrush, maybe simplest by using thermistors or relay bypassed resistors. But that also depends on capacitance you put after the regulator. Maybe a suitable thermistor in series to the Zener is sufficient.

However to build such a thing reliably would need a little experience and some trial and error.

Have fun, Hannes

 

[JPV]

Well, simplest would be to use an appropriate Zener across the regulator to clamp the differential voltage (power-Zener!).

In that case I would definitely recommend to limit current inrush, maybe simplest by using thermistors or relay bypassed resistors. But that also depends on capacitance you put after the regulator. Maybe a suitable thermistor in series to the Zener is sufficient.

However to build such a thing reliably would need a little experience and some trial and error.

Have fun, Hannes

There is a device TSR made by a german engineer (http://www.emeko.de/index.php?id=29&L=1) that controls the inrush current in toroidal transformers by controlling the operating point of the transformer in sending current pulses at switch on. One model of this device can also take into account the control of inrush current added by a capacitor bank after the rectifier.
The capacitors of the regulators are in parallel with the bank. In this way, using a TSR and a regulator with zener clamp would perhaps be a reliable and easy approach to the problem.

JPV