"Dale Parfitt" wrote in message
news
Has anyone connected a pair of identical computer supplies in parallel for
additional current capabilities? Did it require a low R resistor to allow
for any mismatch?


Since you've asked a general question, I'll give a general answer.
The problem with paralleling power supplies is that the voltage regulators
will fight each other if they are not set to regulate at the same voltage.
Using a low value resistor to isolate the outputs will work, but it will use
power you may not want to waste and your voltage regulation will not be as
good. A technique that was used on mainframe power supplies was to put low
value resistors (actually, calibrated current shunts) on both outputs into a
common load. Then the regulator on one power supply would be adjusted so
that the voltage drop across its resistor would match the voltage drop
across the other resistor. Once the regulators were balanced in this
fashion, the resistors were removed from the circuit and the power supplies
were connected directly to the load. As long as the power supply regulators
tracked identically, there was no problem.

HTH,

-NM

Norm Mann wrote:
"Dale Parfitt" wrote in message
news
Has anyone connected a pair of identical computer supplies in parallel for
additional current capabilities? Did it require a low R resistor to allow
for any mismatch?


Since you've asked a general question, I'll give a general answer.
The problem with paralleling power supplies is that the voltage regulators
will fight each other if they are not set to regulate at the same voltage.
Using a low value resistor to isolate the outputs will work, but it will use
power you may not want to waste and your voltage regulation will not be as
good. A technique that was used on mainframe power supplies was to put low
value resistors (actually, calibrated current shunts) on both outputs into a
common load. Then the regulator on one power supply would be adjusted so
that the voltage drop across its resistor would match the voltage drop
across the other resistor. Once the regulators were balanced in this
fashion, the resistors were removed from the circuit and the power supplies
were connected directly to the load. As long as the power supply regulators
tracked identically, there was no problem.

HTH,

-NM


Another method that is currently popular among high-end supplies is to
have a current command output and a current slave input. With proper
jumpering, you set up one power supply as a "master" that outputs a
current command, and have a bunch of "helper" supplies that _don't_
attempt to regulate voltage, but just supply current on command.

If you're _buying_ supplies to do this, I'd suggest hitting the surplus
places for general-purpose switching supplies. MPJA has a good
selection, and I recall being able to find the manufacturer's data
sheets with some digging. You can get much nicer prospects that way
than trying to hack a PC power supply, although it's not as cheap as
dumpster diving for PC parts.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" gives you just what it says.
See details at http://www.wescottdesign.com/actfes/actfes.html

Is there a feedback sensing line for voltage regulation
in the computer supplies that is accessible?
One trick I've seen done is to diode OR
the supplies, with the sense lines on the common
supply bus, after the diodes. MACom did this
on the massive switchers we used in our old fiber
nodes at work.

Tio Pedro wrote:
Is there a feedback sensing line for voltage regulation
in the computer supplies that is accessible?
One trick I've seen done is to diode OR
the supplies, with the sense lines on the common
supply bus, after the diodes. MACom did this
on the massive switchers we used in our old fiber
nodes at work.


I'd vote for the diode method, moving the sense line to after the diodes
OR just adjusting the supplies for about .6V more than what you want to
compensate for the diode voltage drop.

"Tio Pedro" wrote in message
...
Is there a feedback sensing line for voltage regulation
in the computer supplies that is accessible?
One trick I've seen done is to diode OR
the supplies, with the sense lines on the common
supply bus, after the diodes. MACom did this
on the massive switchers we used in our old fiber
nodes at work.


There are no sense lines available externally on a PC power supply. You might
take a look at the article on modding PC power supplies at
http://www.pavouk.org/hw/en_atxps.html. It shows a typical schematic (albeit a
bit aged now). Look at it and you can see where you might bring the sense lines
out of a "master" unit and run them into the "slave" units. No guarantee that
it will work, but I'd rather try that approach before the diode or resistor
solutions.
Of course, your PSUs will be different from the schematic in the web page, but
it should be similar enough that you can figure it out.
--
Dave M
MasonDG44 at comcast dot net (Just substitute the appropriate characters in the
address)

Life is like a roll of toilet paper; the closer it gets to the end, the faster
it goes.

On Dec 29, 5:40*pm, "DaveM" wrote:
"Tio Pedro" wrote in message

...

Is there a feedback sensing line for voltage regulation
in the computer supplies that is accessible?
One trick I've seen done is to diode OR
the supplies, with the sense lines on the common
supply bus, after the diodes. *MACom did this
on the massive switchers we used in our old fiber
nodes at work.

There are no sense lines available externally on a PC power supply. *You might
take a look at the article on modding PC power supplies athttp://www.pavouk.org/hw/en_atxps.html. *It shows a typical schematic (albeit a
bit aged now). *Look at it and you can see where you might bring the sense lines
out of a "master" unit and run them into the "slave" units. *No guarantee that
it will work, but I'd rather try that approach before the diode or resistor
solutions.
Of course, your PSUs will be different from the schematic in the web page, but
it should be similar enough that you can figure it out.
--
Dave M
MasonDG44 at comcast dot net *(Just substitute the appropriate characters in the
address)

Life is like a roll of toilet paper; the closer it gets to the end, the faster
it goes.

As long as they use the same IC it shouldnt be too difficult to
figure it out even if the rest of the circuit is different.

Jimmie

DaveM wrote:
"Tio Pedro" wrote in message
...
Is there a feedback sensing line for voltage regulation
in the computer supplies that is accessible?
One trick I've seen done is to diode OR
the supplies, with the sense lines on the common
supply bus, after the diodes. MACom did this
on the massive switchers we used in our old fiber
nodes at work.


There are no sense lines available externally on a PC power supply.

Actually,Alot of ATX supplies have atleast a sense line on one of the
3.3V connections to the motherboard,IIRC it's usually brown,or
orange,and of a smaller gauge than the main wire.
I've seen some older Compaq server supplies that had sense lines on each
rail.

You might
take a look at the article on modding PC power supplies at
http://www.pavouk.org/hw/en_atxps.html. It shows a typical schematic (albeit a
bit aged now). Look at it and you can see where you might bring the sense lines
out of a "master" unit and run them into the "slave" units. No guarantee that
it will work, but I'd rather try that approach before the diode or resistor
solutions.
Of course, your PSUs will be different from the schematic in the web page, but
it should be similar enough that you can figure it out.