need build psu +-16, +-48,+-12 15amp help please!!

Hi guys, i am stuck helping a friend, he baught a lemon for the price of a rolls. I have recapped the desk, new opamps etc. I want to build a new psu from scratch as the old one is very dangerous and damn noisy.

I just thought that i should post here and see if anyone has any ideas. Low ripple, multiple output. Simple and strong is what we need. In the past i have only built little suckers.

thanks in advance.

ben

Buy a number of off the shelf supplies. Power one, Accopian, condor... there are others. Its really the best and easiest.

This post is kind of strange to me.

You recapped and changed opamps but you don't know how to build a power supply?

A power supply is an important thing to understand. How do you know the opamps are even working corrrectly? If you don't know how to design a power supply I would guess you would not know how to use a scope.

[quote author="Gus"]You recapped and changed opamps but you don't know how to build a power supply?[/quote]
yeah but
nah but

15 AMP !!!

am I reading that right ... or is it 1.5 amp
then I agree with Gus

but it it is 15 amp I get the problem
you don't need 15 amp of 48volts so the normal methods should give you enough phantom power.
Big , 0ff the shelf switchmodes, with added smoothing could be a modern approach for a large console.
An analog supply is going to need a big transformer with the required tappings
... or multiple transforemrs (probably cheaper) but even a 24v 15 amp transformer is in Truck Battery Charger teritory ....
then an analog regulator is going to need some serious current sharing.

what's wrong with the original supply ??

physically noisy ... like rattling power trafo windings ?
try the above Multiple trafo substitutes.

:shock:
How big is this console ... 15 amp !!

hi gus and kev, thanks for the reply.

Gus- I can build a power supply and use a scope, i just have never built anything that big before.

Kev, the console is a 28 channel sound workshop series 30.

i can get the parts cheap enough,

The old one is most likely salvagable, if i change the transformers. but i don't think that it is or will ever be safe. It has to be the worst design i have ever seen. totally unserviceable, live mains everywhere.

Do you have a schematic for the original ?
where about in Australia are you ?

this is a big supply
but the basic idea for supplies is the same as for the small ones
but yeah
... this amount of current can make things go bang

The basic idea of the original circuit may be still valid today
check if the output transistors are OLD stock or current
so perhaps a NEW build to make it safer and easier to monitor and service could be what you want

this is your BIG desks power supply is ultimately very important ...
perhaps UN-glamorous
but it deserves just as much thought as the rest of the restoration

Thats good to know. I seem to remember NS had a circuit(s) using a 3 term reg with current booster transistors.

Now most of your higher current power supplies used a 732(IIRC) driving a to3 like a 2n3055 the driving a bank of 2n3055s as a power darlington. To use NPNs there is a x2 voltage circuit to up the voltage for the base drive section. A pnp would not need this but they cost more etc...

Your basic lin power supply is often a voltage ref into an opamp then a drive circuit for the power darlington with a sense connection at the output going back to the opamp. The basic Idea is the opamp drive the pass devices untill the scaled sensed output voltage is equal to the ref voltage

The cooler supplies will use a switcher to prereg the voltage and a lin to "fine tune" the voltage with less power being dissapated buy the lin supply.

Yes I need to step back sometime I work on an AC motor drive that has 690VDC from 3 phase 400 something

I will look in some books and see if I can find the NS app note.

[quote author="Gus"]... I work on an AC motor drive that has 690VDC from 3 phase 400 something.[/quote]
:shock:
b ... aaarrgggh !

ouch
that's in serious spanners territory
electrical connections with BIG terminals and spanners ... not solder or small screw terminals

I have a friend that used to work with the power company at the coal site and his tool kit looked more a plumbers kit instead of the usual sparky's tool kit
he even had a shovel .... ?? don't ask !

hi kev, i am on the nth coast nsw. where are you?

[quote author="giantbenny"]hi kev, i am on the nth coast nsw. where are you?[/quote]
:green:
...
errr
...
keep reading
...
the left hand column
:wink:
never mind ... it's early

i did spot that 1 second after i pushe the damn post button

aarrrggghhhh

ah well doesn't matter.

[quote author="Gus"]You recapped and changed opamps but you don't know how to build a power supply?[/quote]

Doing high-current stuff is somewhat a different pair of shoes...

Thats perhaps what you look for: http://www.thel-audioworld.de/module/spr/spr10.htm
Unfortunately the site is only in german :sad:. Roughly summing it up it's a positive and a negative 10 A adjustable voltage regulator board - for power amplifiers. The current limiting is soft, so it can do peaks up to 20A (in order to deliver impuse power for the amplifier). It's 88? each. I believe it might be possible to parallel 2 of them to achieve 20A constant current (combining both outs via high-current diodes, regulators should be thermallly coupled). It's basically a big 3-leg regulator.

Michael

Interesting topic, I'm going to need to build a new psu for a customers console.

I'll have to check my notes, but I do remember something like 6 amps for each +/-15v, some ~12 amps for the 6 volts (yep, those thousands of leds suck up a lot of current) and the 48 volts needs only to be around 1 amp and we're on the safe side.

I didn't think at all to use switching supplies and after checking out a few catalogs this looks interesting.

This seems to be just a matter of packing those things up and adding some filtering (polycaps) for the mains, also some added capacitance for the rails since I still feel a little bad about these.

Only one question:

How much capacitance for the rails? Do switching supplies have limitations for rail capacitance? For 6 amp rails, to smooth out small ripple & noise, I'd probably go with a combination of 1000uF + 10uF electros and a 1uF and 100nF polys, just to be on a safe side, no calculations at all..

I've been in a similar boat while researching "what's needed" for a Virtuoso Quad-Eight desk in Tulsa (the last desk Quad-Eight ever sold). It's a 36 x 24 desk.

The original power supply was a huge affair, with +/- 18 V at 10 Amps, +24 at 10 Amps....twice! IE, double the amounts listed since the internal desk PS distro was essentially split into two halves.

Unfortunately, Quad-Eight used a (now) Unobtanium "hybrid brick" from Lambda. One had failed prior to the current owner's purchase, so a "sidecar" supply had been added. In recent times, other of the Lambda bricks have failed, and I've cobbled it together with supplies available from the unused automation system.

My "current" (no pun intended) plan is to build up a pair of 3RU chassis units. Each will house a pair of 15 Amp 15 VDC International Power open frame modules (with the Amek mods to allow 18V operation, and a "slow down" of the overcurrent trigger due to the DC inrush...more on that later) and one International Power 24V @ 12 A module in each chassis.

In one of the cabinets, I'll also stash a small 48V @ 0.5A open frame module for the phantom rail. (That will supply at LEAST 35 mics under the worst case conditions, which is pin 1/2/3 on the mic input all shorted on all inputs).

On paper, so far, so good. BUT, each of the three modules weigh 20 Pounds each, so that's 60 pounds in each rack box. The modules are too long to mount sideways, so the rack box has to be approx. 20 inches
deep. Now I have to worry about the integrity of the rack cabinet due to the weight and depth.

On top of everything are the air movement specs for the open frame modules. I've been in contact with Tony Kordyban, who wrote an excellent book about electronic equipment cooling:

http://www.amazon.com/gp/product/0791800741/sr=8-4/qid=1144831111/ref=sr_1_4/002-1161618-6534464?%5Fencoding=UTF8

So, I've also had to add into the equation some Papst fans to move air through the rack cabinets.

All in all, it's NOT going to be an inexpensive project.

I briefly pondered switcher supplies, but I find them to be unreliable in my part of the world (thunderstorms), and I've had serious problems with multiple switchers each "whining" at their own frequency, thus creating beat frequencies as each supply generates square waves at slightly different frequencies.

DC Inrush....

One commonly overlooked issue is the existence of a bazillion filter/iso caps hanging from the DC rails. That is a MOFO load that the power supply has to see when a large desk is powered up...essentially a huge short-circuit. Amek and International Power teamed up to create a solution, with an extra cap (and resistor?) on the power supply module to slow down the over current "triggering" as a large desk was powered from "cold".

Bri

There is the old HiFI trick from the 70's of using a series power resistor in the primary hot line and shorting it out after a set time delay often done with an off the self timer relay IIRC.

The old school ramping component was a thermistor. All old Telefunken radios had one, so that even the heater voltage would ramp up. I recently replaced one in a Macintosh amp which was 10 ohm. In series with the mains. 20 Amp I think?
I also suggest the modular design. The Harrison MR-4 PSU has four International frame supplies inside. So did the replacement AMEK Mozart supply. And yeah, they weigh ALOT. Especially the AMEK. If you go modern, you should add the extra voltage sensing wires using seperate twin screens for each supply. And inspect the power cabling. AMEK supplied a new cable run to the console and the wires were 8 GA instead of the original 12 GA.

NTC thermistors are used extensively in modern offline switching power supplies to limit inrush current.

Agree---modularize, and remote sensing is indeed a good idea---make sure there are some safety R's at the supply between sense ond power terminals so that if the sense lines get open-circuited the voltage doesn't soar.

The International Power/Amek mod acted to desensitize the overcurrent "trip" circuitry, thus allowing a hefty inrush current drawn from the desk. I have the notes in my files, but it was just an additional 'lytic added to the control circuit boards of the open frame supplies.

Good point re. the sense lines. The Quad-Eight did include sense lines back to the Lambda "bricks", and the International Power units provide sense lines as well. However, there is a link (10k resistor?) between the output and sense nodes, thus preventing a runaway.

On a similar note, last Friday I was assessing a Neve 5104 desk purchased by a new client. It had a power supply provided by a third party (King-something) with a Neve badge on it. If the sense lines were unconnected, it was bonkers! Some rails were dead, others soared. Bad design there.....

Bri