Power Supply requirements

Hi

I'm working with an older P1, 233 mhz, Orlando motherboard,
Packard Bell (corner unit). I'm needing to swap out the power supply
and get it repaired as it has died. Because of the design of the
chasis the size/shape of the P/S in the PB is unique.
What I'm faced with is trying to locate a P/S that will fit for
the time being, as I use the PB as a TV, radio and answer machine in
my kitchen via a remote. I did find a smaller P/S that will fit. It is
from an Hew/Pack P3 computer, however, it is only 90 watts. The
original P/S from the PB was 145 watts. The HP has the same
peripherals as the PB does. I know the HP P/S works. I bought the
Cable Converter ATX 20 pin to P8/P9 at Ebay to use with it. Before I
get into it, I'd like to know if you guys think the 90 watt HP P/S
will handle things for a few months?

Thanks

Bill

ss454_@hotmail.com wrote:
> Hi
>
> I'm working with an older P1, 233 mhz, Orlando motherboard,
> Packard Bell (corner unit). I'm needing to swap out the power supply
> and get it repaired as it has died. Because of the design of the
> chasis the size/shape of the P/S in the PB is unique.
> What I'm faced with is trying to locate a P/S that will fit for
> the time being, as I use the PB as a TV, radio and answer machine in
> my kitchen via a remote. I did find a smaller P/S that will fit. It is
> from an Hew/Pack P3 computer, however, it is only 90 watts. The
> original P/S from the PB was 145 watts. The HP has the same
> peripherals as the PB does. I know the HP P/S works. I bought the
> Cable Converter ATX 20 pin to P8/P9 at Ebay to use with it. Before I
> get into it, I'd like to know if you guys think the 90 watt HP P/S
> will handle things for a few months?
>
> Thanks
>
> Bill

Badly! A few years ago, I thought I was doing the right thing upgrading
the memory on a PB with 90w power. With the memory upgrade, the PB
would run for a bit until the power supply decided it was overloaded,
then shut off. I removed the memory upgrade, and the system ran Ok
again... Ben Myers

Hi!

> I'd like to know if you guys think the 90 watt HP P/S will handle
> things for a few months?

I'm not sure it's a good idea. You're probably going to be asking that old
power supply to run close to its maximum rated output and I don't think it
will stand that. Worse yet, it may kill your system if it decides to blow
up.

The Packard Bell supply *should* be a pretty standard baby-AT style unit,
and I'm sure you could find some good ones around on the secondhand market.
These will have a much higher power rating and a good one will be reliable
for some time to come.

William

On Mar 12, 4:44*am, ss4...@hotmail.com wrote:
>I did find a smaller P/S that will fit. It is from an Hew/Pack P3
> computer, however, it is only 90 watts. The original P/S from
> the PB was 145 watts. *The HP has the same peripherals as
> the PB does. I know the HP P/S *works.

Normal is for an undersized power supply to still boot a computer.
Nobody can really answer your question. But numbers from a $20
multimeter can.

First get the computer to do something or as many things possible to
consume maximum power. Then measure DC voltages where AT power supply
connects to the motherboard. AT wires did not always use a standard
color code. So you must identify the appropriate wires with a
connector chart.

AT power supply outputs 5 and 12 volts (and negative versions). In
your case, only the 5 and 12 volts numbers are relevant. Under
maximum load, any one 5 volt and 12 volt line must measure more than
4.87 and 11.7 when load is maximum. If yes, then the power supply is
more than sufficient.

Better testing might temporarily increase the load with two resistors
from Radio Shack. Using a peripheral cable, connect a 10 ohm 10(?)
watt resistor from the red wire to black wire. Connect a 50 ohm 10
watt resistor from yellow wire to black wire. With these resistors in
place, those voltage minimums remain in spec. Then you know without
any doubt that the 90 watt supply is more than sufficient.

Nobody knows what is in your machine and how much power those items
consume. But the above simple test with a multimeter answers you
question definitively. Only a definitive answer is useful. Good
luck. I'm better the HP supply is sufficient. But from here,
everyone can only speculate - not provide a useful answer – until you
provide numbers from a multimeter.

William R. Walsh wrote:
> Hi!
>
>> I'd like to know if you guys think the 90 watt HP P/S will handle
>> things for a few months?
>
> I'm not sure it's a good idea. You're probably going to be asking that old
> power supply to run close to its maximum rated output and I don't think it
> will stand that. Worse yet, it may kill your system if it decides to blow
> up.
>
> The Packard Bell supply *should* be a pretty standard baby-AT style unit,
> and I'm sure you could find some good ones around on the secondhand market.
> These will have a much higher power rating and a good one will be reliable
> for some time to come.
>
> William
>
>
The connectors from the power supply to the motherboard on the older
Pentium/Pentium MMX PB systems are standard baby-AT, i.e. two adjacent
power connectors attached in such a way that the black wire leads from
each connector are next to one another. The problem with many older PB
systems is the shape and size of the power supply itself, often anything
but a standard baby-AT size... Ben Myers

On Mar 13, 10:09*am, Ben Myers <ben_my...@charter.net> wrote:
> William R. Walsh wrote:
> > Hi!
>
> >> I'd like to know if you guys think *the 90 watt HP P/S will handle
> >> things for a few months?
>
> > I'm not sure it's a good idea. You're probably going to be asking that old
> > power supply to run close to its maximum rated output and I don't thinkit
> > will stand that. Worse yet, it may kill your system if it decides to blow
> > up.
>
> > The Packard Bell supply *should* be a pretty standard baby-AT style unit,
> > and I'm sure you could find some good ones around on the secondhand market.
> > These will have a much higher power rating and a good one will be reliable
> > for some time to come.
>
> > William
>
> The connectors from the power supply to the motherboard on the older
> Pentium/Pentium MMX PB systems are standard baby-AT, i.e. two adjacent
> power connectors attached in such a way that the black wire leads from
> each connector are next to one another. *The problem with many older PB
> systems is the shape and size of the power supply itself, often anything
> but a standard baby-AT size... Ben Myers- Hide quoted text -
>
> - Show quoted text -

On Mar 13, 10:09*am, Ben Myers <ben_my...@charter.net> wrote:
> William R. Walsh wrote:
> > Hi!
>
> >> I'd like to know if you guys think *the 90 watt HP P/S will handle
> >> things for a few months?
>
> > I'm not sure it's a good idea. You're probably going to be asking that old
> > power supply to run close to its maximum rated output and I don't thinkit
> > will stand that. Worse yet, it may kill your system if it decides to blow
> > up.
>
> > The Packard Bell supply *should* be a pretty standard baby-AT style unit,
> > and I'm sure you could find some good ones around on the secondhand market.
> > These will have a much higher power rating and a good one will be reliable
> > for some time to come.
>
> > William
>
> The connectors from the power supply to the motherboard on the older
> Pentium/Pentium MMX PB systems are standard baby-AT, i.e. two adjacent
> power connectors attached in such a way that the black wire leads from
> each connector are next to one another. *The problem with many older PB
> systems is the shape and size of the power supply itself, often anything
> but a standard baby-AT size... Ben Myers- Hide quoted text -
>
> - Show quoted text -

I thought it used a standard baby At also.
In any case i once had a Compaq presario with a bad power supply.
A weird shaped supply from a 5184
I opened it up and noticed it had a stadard ATX power supply board.
Purchased ATX Comp USA ATX pwer supply and swapped out the board and
soldered in the Compaq connectors .
Sold the directions over 100 times on ebay.
Sold rebuilds a few times for 60 dollars.
Mine was much more powerful than OEM.

I have done the samer with older dells also.
In solderedtheir connectors with leads to new PS keeping the
respective voltage
leads correct.
All PS for compurers all supply the same voltages depending on their
version atx
At etc.
It is not rocket science to swap out the guts.


I noticed that they all use a full wave bridge rectifier for the frst
leg of the PS
before pulsing it through a 50kcs or higher .
None I have seen use a full wave bridge that has gone bad.

The pulsing circuit is just a multivibrator that saturates a an
internal transistor
to switch ar the rc time constant.
I have seen these go bad quite often and have noticed many different
supplies use the same component.
I have also noticed the generic brands use a cheap knock off.
In any case most of these were not the cause of failure when
inspecting defective PSU's.

I also noticed that many of the primary transformers were total
garbage,
Some were designed with cheap wire diameters and were rated much
higher than the PSU stated ratings.
Others had nice diameters and the transformer far exceeded the PSU
ratings
I make this statement based on visual inspections.
When checking dead units many showed open circuits in the transformer.
They are also easy as hell to rewind by just counting the turns.

The main electrolytic capacitors were not up to snuff did not exceed
the maximum circuit values by more than 20%.
The ones used in the main FWB still seemed to hold up.

The heat sinking of the most critical components were done failrly
well
with bulkly heat sinks.
The actual cooling on most sucked especially when placed in the tower.


Most could have been as good as the industry best for just a few cents
in component costs.

It is not like the old American TV sets that used a 500watt
transformer in a 200 watt circuit.
Thes Atx etc supplies are designed just a tad over what is needed and
not one
iota over.


As you have noticed i like to play with things as PSU.
My friend works on the dump and i get quite a few old computers
for parts to sell on Ebay.

Everyone I tested did not fail on short circuit load except one from
an Emachine.
None failed when using a 2.2KV pulse surge for a 1/1000 second
duration.
This is only pulse test that i am able to do using an old flash
circuit.


I also did a makeshift heat test.
Just wrapped aluminum foil around unit to restrict air flow.
I used 20 eight ohm 10 watt resistors in parallel for the 3 volt leg.
This is my way of checking for thermsl overload not short circuit.
Every PSU I tested shut down after 20 minutes.
Out of 6 that i tested 2 never recovered.



Heat seems to be the major destroyer of these units.

**** poor design seems to be the other.


400 watts in a little case place in another case seems to be pushing
the limits of
design.


Not withstanding the fact that low voltage with long leads can lead to
PSU problems ,I still feel the PSU should be external to the computer
or adjacent to the computer but not placed in the case. It should also
have cooling that blows real air
even if it is alittle noisy. Important computers for indistries are
placed in an air conditioned rooms.

Hi!

> I thought it used a standard baby At also.

I've got a Legend 100CD here that definitely *does*. It's a pretty
uninteresting DVE-branded supply with the good old baby-AT case size,
connector position and fan location. It does not have a monitor power
connector, nor is there a cutout in the machine's case for one.

Some other PB systems have used an uncommon supply shape, but those are all
486 or 386SX based to the best of my knowledge. I have a PB 386SX with a 60
watt (!!) rated DVE branded supply. It's a small and unusual shape, and
would be hard to replace. However, it definitely has standard plugs on it.

> I opened it up and noticed it had a stadard ATX power supply board.

I saw a similar kind of thing most recently in a PIII Deskpro EN desktop.
Inside the supply was a Hipro brand board, about the same size and shape of
an ATX board. It could be retrofitted in the way that you describe as long
as the wiring was right. Compaq changed some pin functions in the ATX
connector.

> I have done the samer with older dells also.

There are actually some of these on the market. StarTech.com sells some with
the Dell pinout.


> I noticed that they all use a full wave bridge rectifier for the frst leg
of the PS
> before pulsing it through a 50kcs or higher .

.....

I have filed this information, as there are some computer power supplies I
want to repair. They are all Delta SMP-332AB units pulled from PS/2 Model 95
systems. More often than not, they just died silently after a power failure
and restoration. There are no blown fuses or signs of distress inside. I did
have one looked at by someone much more knowledgeable than I, and I found
out that the TL494 controller is being told to stay shut down. I'm not sure
what got them--internally they are built like brick outhouses.

> The main electrolytic capacitors were not up to snuff did not exceed the
maximum
> circuit values by more than 20%.

They definitely suffer from the heat, and most power supply makers use 85
degree celsius rated units, instead of the superior 150 degree ones.

> The actual cooling on most sucked especially when placed in the tower.

Many of them seem to have a so-called "fan noise killer" circuit. I'm not
sure how this is supposed to work, although some do increase fan speed with
a thermal or current draw monitor. Others just sit there and run quietly
until they blow up from the toll the heat takes on them.

I'd rather have fan noise and a cool PSU than one which runs very hot and
fails in an impressive manner.

> Thes Atx etc supplies are designed just a tad over what is needed and not
one
> iota over.

Some are much better than others. PC Power and Cooling has some truly
superior units, some of which broke the 90% efficiency barrier in a test I
saw conducted against multiple "80 Plus" power supplies on the web. No one
else was close.

Most recently I built a server with a PCP&C Turbo Cool 860. The thing is
built like a battle tank. You'd hurt yourself if you dropped it on your
foot--it's really that solid. And it manages to run very cool, with little
fan noise, even at high load. (High load being five hard disks, two AMD
Opteron dual core processors, 4GB fully registered ECC RAM, a large slow
turning fan and an Adaptec PCI Express hardware RAID card.)

And then there are those PSU manufacturers, who, for want of a better term,
lie about their products:

http://greyghost.mooo.com/psuthoughts/

> As you have noticed i like to play with things as PSU.
> My friend works on the dump and i get quite a few old computers
> for parts to sell on Ebay.

I am always in favor of keeping things out of the dump. While I don't know
how to repair everything, I do what I can to keep stuff out of the dump.

It doesn't have much to do with the discussion, but the other day I pulled a
rusty, trashpicked McGraw-Edison (later Toastmaster) "Eskimo" box fan out of
my storage shed. It hadn't run in a while, and it looked like junk. I'd
guess that it is somewhere around 40 years old, with a real metal case,
grilles, a sturdy motor and fan blades that would surely remove a finger if
one came their way. The little motor was stiff, but it came out of it with
some careful work, cleaning and re-oiling. It now runs perfectly, although
it looks really rough.

What amazes me is the way these fans--which clearly say you're supposed to
oil them every few months--ran for many years if not decades when today's
"permanently lubricated" ones quit within a few months to a year.

But that has nothing to do with this. It's just another commentary on how
things could still be made well...

William