Auto power to ATX???

  • Thread starter Thread starter Phrederik
  • Start date Start date
P

Phrederik

Hey all!

Does anyone have a circuit diagram, or adapter to convert the ~12v DC in a
car to the voltages required for a low end ATX based PC?
 
Phrederik said:
Hey all!

Does anyone have a circuit diagram, or adapter to convert the ~12v DC in a
car to the voltages required for a low end ATX based PC?
Click to expand...

There are three basic ways of doing this:

1) Brute force. Use a 12V/115V AC invertor and a standard 115V AC ATX
power supply.

Advantage: Parts readily available at reasonable cost.

Cons: Inverter efficiency <70%, ATX efficiency <70%, overall ~50% High
current drain from automobile supply.

2) Special ATX power supply designed to use 12V DC input.

Advantage: Some can fit directly into standard computer case. Several
manufacturers.

Cons: High cost. Efficiency about 60-65%.

3) Separate circuits to supply +12V, +5V, and +3.3V.

This is the approach several of the references cited by Allen Zord
use. Some did not supply the 3.3V needed for an ATX supply. The ones
that did supply all ATX voltages were relatively low power.

Advantage: Best efficiency of the bunch.

Cons: Most were DIY type enterprises, except for the low-power ones.

General comments: A standard ATX power supply takes the 115V input
and rectifies and doubles it to get about 320V DC. This allows for
reasonable cost switching transistors with a relatively low voltage
drop in the "on" state. This makes for good effieciency.

With the same design, but directly using the 12V DC input, requires
high current switching transistors, with significant voltage drop.

Separate circuits for each output voltage improve the efficiency,
but the approach is more complicated.

I have not addressed the -5V and -12V supply. This is a trivial problem.

How much power do you need? Have you considered a system using an
AT power supply, not requiring the 3.3V? How much work are you able
and willing to do yourself?

Virg Wall
 
Does anyone have a circuit diagram, or adapter to convert the ~12v DC in
a
There are three basic ways of doing this:

1) Brute force. Use a 12V/115V AC invertor and a standard 115V AC ATX
power supply.
Click to expand...

Simple, but not how I want to do it.
2) Special ATX power supply designed to use 12V DC input.
Click to expand...

This is what I was considering and hoping for a "do it yourself" type
schematic, or relatively inexpensive prebuilt unit.
3) Separate circuits to supply +12V, +5V, and +3.3V.
Click to expand...

This is probably the easiest way to go, once the design is figured out. You
can provide the required current for each voltage seperately - having a
beefy +5v and a much smaller -5v supply, for example.
General comments: A standard ATX power supply takes the 115V input
and rectifies and doubles it to get about 320V DC. This allows for
reasonable cost switching transistors with a relatively low voltage
drop in the "on" state. This makes for good effieciency.

With the same design, but directly using the 12V DC input, requires
high current switching transistors, with significant voltage drop.

Separate circuits for each output voltage improve the efficiency,
but the approach is more complicated.

I have not addressed the -5V and -12V supply. This is a trivial problem.

How much power do you need? Have you considered a system using an
AT power supply, not requiring the 3.3V? How much work are you able
and willing to do yourself?
Click to expand...

I can do the work myself - produce a PC board, solder, etc... As for how
much power? I'm still working on that. I was looking a possibly ~1Ghz Duron,
but have been looking at the Via/Cyrix III processors since they draw so
little power... which will lead me to my next post...
 
Hey, Virg Wall, I've been wondering some what the same question, only about
AT power supplies. I've been working on making an audio player to go in my
car. Basically i'm done..... but I don't have the power issue solved yet.
I really didn't want to use an inverter. Would it be fairly easy to build a
12v DCin power supply and output the proper voltages /current that an AT
power supply would output?

How big it is isn't really an issue to me unless it won't fit in the trunk
of my grand marquis......hehe.

Thanks

Nic


V W Wall said:
There are three basic ways of doing this:

1) Brute force. Use a 12V/115V AC invertor and a standard 115V AC ATX
power supply.

Advantage: Parts readily available at reasonable cost.

Cons: Inverter efficiency <70%, ATX efficiency <70%, overall ~50% High
current drain from automobile supply.

2) Special ATX power supply designed to use 12V DC input.

Advantage: Some can fit directly into standard computer case. Several
manufacturers.

Cons: High cost. Efficiency about 60-65%.

3) Separate circuits to supply +12V, +5V, and +3.3V.

This is the approach several of the references cited by Allen Zord
use. Some did not supply the 3.3V needed for an ATX supply. The ones
that did supply all ATX voltages were relatively low power.

Advantage: Best efficiency of the bunch.

Cons: Most were DIY type enterprises, except for the low-power ones.

General comments: A standard ATX power supply takes the 115V input
and rectifies and doubles it to get about 320V DC. This allows for
reasonable cost switching transistors with a relatively low voltage
drop in the "on" state. This makes for good effieciency.

With the same design, but directly using the 12V DC input, requires
high current switching transistors, with significant voltage drop.

Separate circuits for each output voltage improve the efficiency,
but the approach is more complicated.

I have not addressed the -5V and -12V supply. This is a trivial problem.

How much power do you need? Have you considered a system using an
AT power supply, not requiring the 3.3V? How much work are you able
and willing to do yourself?

Virg Wall
Click to expand...
Sagan ~
 
What is the fastest "AT" PSU based mainboard/CPU available...

I'm sure it would be lower power than most ATX systems, but how would it
fare playing MP3's and DVD's???
 
Well I think that playing mp3's does not require much processing power....
my mp3 player (not yet in the car) has a pentium 133Mhz..... and as long as
it's not in graphics mode, meaning it displays a bunch of funky junk to go
with the music (and I never PLAN to have it that way) it runs just fine. As
for DVD decoding, at least in decent quality/speed/framerates and what not,
I think it'd require alot higher than a pentium or equivilent processor.
 
dang it I keep replying to sender and not group! ARG!

Anyway....


Well I think that playing mp3's does not require much processing power....
my mp3 player (not yet in the car) has a pentium 133Mhz..... and as long as
it's not in graphics mode, meaning it displays a bunch of funky junk to go
with the music (and I never PLAN to have it that way) it runs just fine. As
for DVD decoding, at least in decent quality/speed/framerates and what not,
I think it'd require alot higher than a pentium or equivilent processor.

Nic
 
Robotnik said:
Hey, Virg Wall, I've been wondering some what the same question, only about
AT power supplies. I've been working on making an audio player to go in my
car. Basically i'm done..... but I don't have the power issue solved yet.
I really didn't want to use an inverter. Would it be fairly easy to build a
12v DCin power supply and output the proper voltages /current that an AT
power supply would output?
Click to expand...

How much current at +5V and +12V do you require?
How big it is isn't really an issue to me unless it won't fit in the trunk
of my grand marquis......hehe.
Click to expand...

Here's the reference Alein Zord gave: It's for an AT type supply.

http://home.concepts-ict.nl/~cybersake/carmp3/sch_smps.gif

In looking at the schematic, it supplies 5V at 5A, and 12V unspecified,
but looks like at least a couple of amps. It also gives -5V at 0.2A,
and -12V (most recent cards don't even use this).

My best guess is it would easily fit on a CB about 6" X 6". It would
draw about 10 amps or so from the battery. A cigarette lighter plug
should work if you don't want something more permanant.

I haven't checked the MC34167T ICs used. Will do so soon, if only for
my own curiosity. The rest of the circuit looks pretty standard, with
the possible exception of the two transformers (?), (L3-220U). "U" usually
means micro-henry, but this shows two windings on a core (?). They
are only used to supply the negative voltages, so no big deal.

If you have any questions, I can try to answer.

Virg Wall
 
V said:
.... snip ...

With the same design, but directly using the 12V DC input, requires
high current switching transistors, with significant voltage drop.

Separate circuits for each output voltage improve the efficiency,
but the approach is more complicated.
Click to expand...

Automobile systems have notoriously high noise levels and large
transients. This requires heavy filtering to use safely, since
transients do unmentionable things to CMOS chips. The general run
of MBs expect clean power, and do little or no filtering. Just a
word of caution.
 
CBFalconer said:
Automobile systems have notoriously high noise levels and large
transients. This requires heavy filtering to use safely, since
transients do unmentionable things to CMOS chips. The general run
of MBs expect clean power, and do little or no filtering. Just a
word of caution.
Click to expand...

Good point. The sample circuits for the MC34167T switching convertor
show 20mv p/p with only a 4700U capacitor on the output +5V.

The circuit (.gif) that Alien Zord refered to, has two 4700U capacitors,
separated by a RF filter. The output ripple and transmitted transients
should be pretty good. The input is also well filtered, with a zener to
clamp any spikes.

It looks like there is a tendancy for many systems to do their own
voltage regulation and filtering. We may find power supplies that
need only a good 12V output used for motors, with the rest of the
circuits being supplied by their own switching regulators from the
same 12V source. This would be great for mobile (automotive) use.

The Motorola switching regulator data sheets show a supply using a
simple, self starting, input switching supply to generate 24V input
from the rectified line voltage which supplies several MC34167
regulators set up for the needed output voltages. This would move
the complexity out of the PS, and allow the power to be tailored
to individual circuits.

Thanks for the heads up!

Virg Wall
 
Back
Top