Need linear supply 3.3V/15A or 5V/25A

[Palindrome]

Hooray! Thanks! Looks like it might be a classic uA723 design so a hack
should be a piece of cake.

It has an interesting number and placing of TO3 outlines.

My first thought was that there looked to be 6 identical power resistors
in the middle, hinting at load sharing between 6 series pass elements.
That seems a tad OTT.

I'd agree that the black blob could be a 723 DIP.

I'd be quite interested in a better resolution image of the board, if
anyone has one. It may be a pretty conventional circuit - but it appears
to be a fairly unusual physical configuration. The extension of the
heatsink to behind the main caps suggest that there are more power
devices there (in addition to the main rectifier) - but I can't see
where their connection points on the board can be.

As I said, it is always interesting how other designers lay out things!

 

[oopere]

Oh, I thought you were a Yorkshire man from the the address. They might
state an address of 'oopere' with some pride..

Sorry, I'm afraid I didn't catch the relation with Yorkshire. I'm
located some thousands Km south of that!

Pere

 

[Joerg]

Joerg wrote:



Well, your problem is already solved, but you gave me
an idea with the above. You've got to stop doing that!

Suppose you re-wound the secondary of a microwave oven
xformer for six volts and a bucket of amps?

I haven't done it myself, but you planted the idea, so
I may play with it. Get ready to post some of your
"sound effects". You're _really_ good at that!
You've done a few recently that stimulated some memories,
the most painful (but not loudest) of which was turning a
6800 uf into confetti. And yes, I checked the polarity
before applying power. And yes, I got it wrong. :-(

Mine was a 470uF/330V or thereabouts. I was a kid and blissfully unaware
that even large capacitors have a finite ESR value. Kilowatt-size power
amp and the cap hung on for many minutes. Then the fluorescent lights
mysteriously dimmed. But they had no dimmer. What the heck is that?
*BAM*. The cap had decided to turn into a spacecraft. It flew right past
my head (no safety goggles), smacked into the plastered ceiling, came
down and phsssss ... burned a hole into the carpet. Some white fluff
sailed down, like snow. "Son, what was that?" ... "Uh, err, nothing, pa".

Had to replaster that part of the ceiling and buy a new carpet, then lay
it. In our family, if you screw up you first try pull yourself out by
the bootstraps if reasonably possible. Transporting a 15'*10' roll of
carpet on a bicycle over about five miles is something I learned was
reasonably possible. Meaning no cop saw me

 

[Palindrome]

Those are pretty standard circuits. Ours is still in transit but we've
bought many like this before from Condor and so on. There are one to six
TO-3 transistors on the back and on "newer" models with their bases and
emitters soldered right into the board. For mechanical reasons I don't
like that, the old wires were better. But oh well, slim pickens with
linear supplies these days unless you roll your own which I always do
for the production version.

The circuitry is the classic uA723 veriety with foldback limiting.
Nothing special there. On 5V supplies there is often a crowbar but
mostly the cheap non-TL431 circuit so the trip point is set above 6V. On
other models you get to roll your own external crowbar.

What I was commenting on was the /physical configuration/ being unusual
- not the circuit design. eg the size, shape and positioning of the
heatsink v the position of the series pass devices.

There will indeed be "one to six T03.." You can see some of them
relatively clearly. Looking at the heatsink, I would expect some to have
been positioned to the left of centre - but I can't see any there. It
could be my eyesight, of course

Now there may not be 6 of them - what looks like 6 identical power
resistors may not be the what they appear to be.

I do agree with you about the mounting of the TO3s to the board. I hate
that too. But again, that is /physical configuration/ - not circuit
design. I wouldn't have laid out the components as this designer has,
although I may indeed have started from a very similar circuit diagram.

Hence the request for a high resolution image. Not because I am curious
about the circuit used - but because I am curious to know where the
designer has put things -especially things that I can't identify but
expect to be there..

 

[Joerg]

What I was commenting on was the /physical configuration/ being unusual
- not the circuit design. eg the size, shape and positioning of the
heatsink v the position of the series pass devices.

There will indeed be "one to six T03.." You can see some of them
relatively clearly. Looking at the heatsink, I would expect some to have
been positioned to the left of centre - but I can't see any there. It
could be my eyesight, of course

Now there may not be 6 of them - what looks like 6 identical power
resistors may not be the what they appear to be.

I do agree with you about the mounting of the TO3s to the board. I hate
that too. But again, that is /physical configuration/ - not circuit
design. I wouldn't have laid out the components as this designer has,
although I may indeed have started from a very similar circuit diagram.

Hence the request for a high resolution image. Not because I am curious
about the circuit used - but because I am curious to know where the
designer has put things -especially things that I can't identify but
expect to be there..

The only photo I could think of is the large unit at the top here but
it's not hi-res at all:

http://www.power-one.com/resources/products/datasheet/lin.pdf

All the stuff is there, they placed more temp-sensitive things to the
far right where it is cooler. I could take a pic of the power supply
next time I am at that client.

 

[Palindrome]

The only photo I could think of is the large unit at the top here but
it's not hi-res at all:

http://www.power-one.com/resources/products/datasheet/lin.pdf

All the stuff is there, they placed more temp-sensitive things to the
far right where it is cooler. I could take a pic of the power supply
next time I am at that client.

Thanks, that's a very kind offer. Although I must confess that it is
really just fairly idle curiosity on my side - so please don't waste too
much of your time on it.

 

[Jasen Betts]

Sorry, I'm afraid I didn't catch the relation with Yorkshire. I'm
located some thousands Km south of that!

"oop 'ere"

 

[Frithiof Jensen]

Oh, I want the linear feel, just not this much feel

Ok ;-)

How about a load of parallel LM317's as per LM317 application note with a
differential amplifier to measure the voltage drop across the stack used to
set the output level of a big COTS switcher so the LM317 only drops a more
than whatever the LM317 spec guarantees is adequate?

At least that should cut some of the mass out?

 

[oopere]

"oop 'ere"

 

[Joerg]

Ok ;-)

How about a load of parallel LM317's as per LM317 application note with
a differential amplifier to measure the voltage drop across the stack
used to set the output level of a big COTS switcher so the LM317 only
drops a more than whatever the LM317 spec guarantees is adequate?

At least that should cut some of the mass out?

Well, we got a big honking 5V supply and I'll do the usual hack. If I
can do it without having to remove the circuit board it'll be done in
under 30 minutes ;-)

 

[ehsjr]

Mine was a 470uF/330V or thereabouts. I was a kid and blissfully unaware
that even large capacitors have a finite ESR value. Kilowatt-size power
amp and the cap hung on for many minutes. Then the fluorescent lights
mysteriously dimmed. But they had no dimmer. What the heck is that?
*BAM*. The cap had decided to turn into a spacecraft.

I hate it when they do that.

It flew right past
my head (no safety goggles), smacked into the plastered ceiling, came
down and phsssss ... burned a hole into the carpet. Some white fluff
sailed down, like snow. "Son, what was that?" ... "Uh, err, nothing, pa".

I've heard of that filter circuit:

D1
+ --->|---+---o B+
|
/ \
| |
| |
| | C1
| |
| |
/ \
| ^ |
| / | \ |
|/ | \|
|
Gnd ------+

Had to replaster that part of the ceiling and buy a new carpet, then lay
it. In our family, if you screw up you first try pull yourself out by
the bootstraps if reasonably possible. Transporting a 15'*10' roll of
carpet on a bicycle over about five miles is something I learned was
reasonably possible. Meaning no cop saw me

Now THAT's an accomplishment! Gives "reasonably possible" a
whole new meaning.

That "family policy" is great. These days it seems the parents
always want to bail out little Johnny instead of teaching him
responsibility. What a shame - and what a disservice to one's
children. Your family did it right.

Ed

 

[Joerg]

I hate it when they do that.


I've heard of that filter circuit:

D1
+ --->|---+---o B+
|
/ \
| |
| |
| | C1
| |
| |
/ \
| ^ |
| / | \ |
|/ | \|
|
Gnd ------+

ROFL! That made my day.

Now THAT's an accomplishment! Gives "reasonably possible" a
whole new meaning.

That "family policy" is great. These days it seems the parents
always want to bail out little Johnny instead of teaching him
responsibility. What a shame - and what a disservice to one's
children. Your family did it right.

Yes, they even buy them new cars nowadays after they totaled #1. My
parents never bought me a car and I would have never asked them in the
first place. After I got my masters my father wanted to help me with the
roughly $2.5k for a used Chrysler but I politely declined. I had enough
in my savings after working my tail off.

 

[ehsjr]

Joerg wrote:


Well, your problem is already solved, but you gave me
an idea with the above. You've got to stop doing that!

Suppose you re-wound the secondary of a microwave oven
xformer for six volts and a bucket of amps?

I haven't done it myself, but you planted the idea, so
I may play with it. Get ready to post some of your
"sound effects". You're _really_ good at that!

<snip>

I'm goig to post a new thread about rewinding
an MOT.

Ed