P4@3Ghz... yes another time...

**stepir** · 12-01-2005, 07:27 PM

Hi there,

still trying to find a way to power my little thermal energy plant...

Just to hear last time (before buying an inverter) that I'm having another dumb idea:

Would be possibile to powering my system using a M1-ATX and the normal +12V of my battery?

What I mean is using the m1-atx to power the 5v and 3.3v rail and connect the +12v directly to the battery. Also the on/off power switch of the mobo still connected to the m1-atx.

I tried my system (p4,2x512DDR2,300gb hdd sata, 7 usb device) with a standard 220W ATX PSU at home. It was barely able to work... but it did it! and the voltage on the +12V rail was no more than 10,2V (measured with my multimeter).

So I guess that maybe a low voltage ONLY on the 12v rail would not reboot my carpc (I'm thinking to the engine crac for example).

**pRoFiT** · 12-01-2005, 07:47 PM

Engine crank as low as 8 to 9 volts. plus while running your alt puts out upward of 14.4 volts which is also bad for a motheboard. Keep trying.

At first my system would die while running because I bought a PW200 which has no regulator built on the 12volt side. So i had to build one myself.

**DarquePervert** · 12-01-2005, 08:00 PM

Originally Posted by stepir

Would be possibile to powering my system using a M1-ATX and the normal +12V of my battery?

What I mean is using the m1-atx to power the 5v and 3.3v rail and connect the +12v directly to the battery. Also the on/off power switch of the mobo still connected to the m1-atx.

At the best you'll have an unstable system. At the worst you'll fry your CPU. The worst is a very realistic possibility.
As previously stated, the battery's output fluctuates. The CPU requires a constant 12v.

**Haystack** · 12-02-2005, 08:51 AM

maybe if you got two DSATX(spelling?) and wired them together parrellel? anyone tired that yet?

**Chris31** · 12-02-2005, 10:54 AM

Originally Posted by Haystack

maybe if you got two DSATX(spelling?) and wired them together parrellel? anyone tired that yet?

Someone supposed to have tried, dunno know what happened in the end. They seems to shy away when things goes wrong.

2x OPUS 150watt in parallel? safe?

Zoots have put 2 of his DSATX in parallel, regarding the real reliabilty test I dont know.

I have designed a 350W but all on papers at the moment. I have put 2 in parallel to make a 700W PSU (in theory) and simulated it. The simulation software shows more stress on few components, mainly the MOSFETs, inductors and the current sense resistors...those are the parts are most likely to fails.

Im hoping to release the 350W soon if all goes well, so far its looking really good on simulations, its never really is the same in reality. PCB is currently being designed, its the hardest part to do due to many restrictions. It be smaller than zoots PSU, and probably cheaper too.

Dont get too exicted, I still have alot to do, havent built the first prototype yet and anything can happen. Ill probably need to derate it too due to large amount of current needed to operate it during cranking. But Im sticking to 350W if possible, thats my target. 400W is abit over kill and really difficult to design.

**Haystack** · 12-02-2005, 11:05 AM

hmmm sounds like you need an updates page

**Haystack** · 12-02-2005, 11:13 AM

well I read over and it seemed to work fine. I cant think of a single reason why it wouldnt work, minus the power switch. That would be an easy fix. I honestly would have just wired in the rails, if I had a diagram, thatwould have been fine with me.

**Chris31** · 12-02-2005, 12:47 PM

Well strangely enough, I have emailed OPUS and they seem to refuse to comments regarding its reliability.

The impression Im getting it is not safe, its like do it at your own risk. Ill probably risk it myself, but its wrong to promote it.

**Chris31** · 12-02-2005, 01:09 PM

Ill only agree with the PSU paralleling if a small circuit is made specially for it.

The board would take in 2 x ATX connector with ORing diodes or ORing MOSFETs. Those design are normally found in servers where backup PSU are needed or more currents are required.

Heres an example ORing circuit >>>

http://www.onsemi.com/pub/Collateral/AND8174-D.PDF

Those who doesnt still undertand what its all about >>>

Why Use ORing Diodes At All? ORing diodes are costly
and they waste power. It is true that multiple power supplies
will work if the outputs are just wired together in parallel.
Two 5−volt AC/DC power supplies taken off the shelf and
wired together in parallel will give five volts output when
they are both powered up, but they may not share the load
very well. The one with the highest set−point will provide
almost all the current. But if one of them is powered off, the
other will supply the entire load current. It will also bias up
the output filter capacitor of the first (off) power supply. The
ORing diodes are used for one single purpose; to protect the
system power bus. If one of the power supplies has a failure
in the output rectifier or filter capacitors that causes it to
short−circuit, then the ORing diode protects the system bus
from being shorted.
The ORing diode will also prevent the system bus from
dumping a charge current into a powered−down supply that
is installed while the system is still on. This function is more
the business of a hot−swap controller, but it also works with
ORing diodes. The ORing system works best if the design
has forced current sharing to get the greatest PS utilization.
ORing does not protect the individual power supplies from
catastrophic failures.

Theres plenty of info on the net regarding paralelling PSUs, whats good and whats bad. Up to the reader to decide if he should take the risk or not.

Alot of those who said they done it, well I doubt they know much about PSU designs, theres alot more to it.