PSU as Car Stereo Power

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View
I'm looking to use a ATX12V PSU to power a car stereo head unit and
amplifier.  On the car stereo side, I'm comfortable with the hookups,
but what do I need to do at the PSU?  So far this is what I've gleaned
from (admittedly amateurish) videos and guides on the web:

1) Short PS_ON (Green) to any ground.

2) Twist the two 12V (Yellow) leads headed to the 24-pin connector
together for the + power lead to your stereo setup.

3) Twist any two GND (Black) leads together to be the - power lead to
your stereo setup.

4) All other leads from the PSU should be taped or otherwise insulated.
  (Most sites suggest cutting this off inside the case, but I think I'm
going to try to leave this PSU viable for its intended purpose later.)

Is it really this simple?

Re: PSU as Car Stereo Power

On 09/26/2012 06:25 PM, Grinder wrote:
Quoted text here. Click to load it


I've used a computer PSU as a power supply for other devices.

I just used the wires going to the hard drive connectors


Re: PSU as Car Stereo Power

Grinder wrote:
Quoted text here. Click to load it

You're on the right track, but I can think of some
issues to consider.

1) Yep, using a switch to connect PS_ON# to COM, will turn it on.

2) The Molex pins are rated for at least 6 amps each. Using two pins
    on the main harness (two yellow wires), would give 12 amps. Chopping
    off the connector, and screwing the wires to the head unit, then only
    the wire capacity for current would be a limit. The ATX12V has two or
    four yellow wires, which would be a second way to make a 12V connection.
    But (4) below, suggests it may be better to use the main harness. That
    is, unless you're using a supply which is known to have only one 12V output,
    in which case either connector and yellow wires would be good.

3) An optional feature, would be to draw a bit of current from
    3.3V and 5V. Maybe a 3.3 ohm power resistor on 3.3V, and a 4.7 ohm
    power resistor on 5V. That's about a 1 amp load for each. Check the
    label on the supply - a few older supplies had a "minimum" loading
    spec, right on the label, and if such a spec was printed there,
    I might take the suggestion more seriously. I don't think any of the
    supplies I've bought in the last five years, would need resistors.
    (Some supplies, actually had a dummy load inside the power supply,
    to meet any minimum load requirement.) The power resistors would be
    10W or higher power rating (i.e. big ones). I get them for a couple
    bucks a package at my local "good" electronics store (not a RadioShack).
    This shouldn't be a big deal, but if I don't mention it, someone
    will blame me for not warning you :-)

4) The control loop for the ATX supply, has a phase margin. The
    more bypass capacitance you load the supply up with, the
    greater the chance of de-stabilizing the control loop. The head unit
    could have rather large capacitors inside it for energy storage.

    Normal design for power supplies, would be to have a 45 degree phase
    margin in the control loop. See section 3.2.8 and section 3.2.9 here.
    Section 3.2.8 notes that the 12V1 rail (on a dual rail supply) is rated
    for 5000 microfarads of bypass capacitance, while 12V2 is rated for
    3000 microfarads. Those loads would still be meeting closed loop stability.
    12V1 is for the 24 pin main connector. That would be important, if using
    a cheap dual rail supply. Many supplies use a single output, in which case
    it's likely you'd have at least 8000uF of capacitance driving

    Two of these in parallel (3300+3300 = 6600 microfarads), would be
    sufficient to exceed the 12V1 stability spec. If the head unit uses
    lots of microfarads, to filter the DC and provide for output transients,
    then the loading might be an issue. When you cause the phase margin to
    approach zero, the supply may begin to oscillate (and likely, at a
    relatively low frequency).

    A power supply can have its own stability spec, as well as meet
    ATX. Some brands, their designs "just barely meet" whatever spec
    is in the ATX document. Others, exceed the requirement. But the
    odds of finding a decent datasheet for your average power supply,
    are pretty poor.

    On a project I was working on, it took a bit of time to get a
    spec from our power supply maker, and the power converter we
    happened to use, was rated for 12,000 microfarads. Our loading
    would only have been a fraction of that (purely digital circuits).
    The head unit, on the other hand, knowing it's connected to
    an automotive power system, wouldn't really feel a need to
    limit what it uses. (The automotive power system, might not be
    constrained like an ATX supply is.)

5) The head unit will be looking for 13.2V or so. As it's a lead acid
    battery based power system, and the car battery isn't exactly 12V.
    It actually varies a fair bit. The car battery should be fully charged
    in the winter, to prevent the electrolyte from freezing. On my car,
    with its defective voltage regulator, I have to check this regularly
    during the winter. While the battery is charging, there might be
    as much as 15V across the battery (and potentially, the head unit).
    While the alternator is running, on a load dump, there can be as
    much as a 70V transient. The head unit (or any other automotive
    accessory), needs to be able to deal with these. But running at
    exactly 12V, that's rather low for the car battery case, and the
    head unit might just decide to shut off at such a low voltage
    (to protect the car battery from completely discharging).

    Again, I don't know if the behaviors of such a gadget are documented,
    but these would be my concerns. That it's a 13.2V device, and it may
    have "anti-discharge" features to prevent damage to the car battery.
    Like, when the battery voltage is too low.

    Some ATX supplies, have a potentiometer inside to tweak
    the supply setting. But don't expect more than about a 5%
    adjustment range at the best of times. 5% would take you from
    12V to maybe 12.6V, still not making it to 13.2V. And in case
    you're really clever, analyze the circuit, and see a way to
    exceed that, if you drive it all the way to 13.2, the overvoltage
    circuit may kick in, and shut off the supply. So not only
    do you need to defeat the feedback loop somehow, you also have
    to defeat at least one protection system as well.

They make line-operated supplies for faking a car power system.
Ham radio operators for example, some of their gear was run by
linear regulators (big heatsinks on the back, wasteful of power).
And they run ham radio gear needing >= 13.2V. So there are other
ways, besides the ATX supply, for making a power source. Even
a unit like this, could have a capacitive load stability criterion.

(It looks to be a switcher, or SMPS, rather than a linear. A switcher
can be anywhere from 65% to 87% or so efficiency, whereas the linear is
usually worse than that, and a real power waster.) (4).jpg

A linear DC supply for the ham shack, looks like this.
Big ass transformer, to step down the line. Bridge rectifiers
(the diamond shaped things) for AC to DC conversion. And the
black heatsinks on the right of the picture, discard the waste
heat from the power transistors bolted to them. A very wasteful
bit of gear, good for keeping the room warm in winter. Waste
heat is a function of loading, so less heat when no current
is being drawn. This might have been 13.8V out or so.


The spec page for this Pioneer head unit, says

    "Power source............13.2Vą0.1V DC"

and of course that's ridiculous, as the voltage in a
car is not regulated anywhere near that well. That's
merely the nominal value, and the voltage can go
all over the place, depending on battery state.


Re: PSU as Car Stereo Power


Quoted text here. Click to load it

P&M, reply by post

If you're interested in radio and not the CD player, bear in mind that
car radios do better in cars than they do in the house, because the
car's metal body acts as a ground plane, I'm told, and supplements the

I went through this when I wanted to play 98.1, I think it was, WRC in
the house, and no table radio or fancy receiver would get it, though
both factory car radios did   I was about 45 or 50 miles from the

While I was still thinking about the project, WRC changed frequencies
and to a weaker transmitter that even the car wouldn't get.  So I
never actually tried the car radio indoors and never tried to
duplicate the ground plane (which I have no idea how to do.  Tin

I also don't know if this affects both FM and AM.     I have a $250
table radio that gets loads of FM stations but  only 4 or 5 AM
stations in the daytime, but I haven't compared it to the car
radio.reception   Maybe there just aren't many AM stations in

Re: PSU as Car Stereo Power

On 9/26/2012 6:25 PM, Grinder wrote:
Quoted text here. Click to load it

I've set a supply up, and it is generally working.  Here's the PSU I used:

I say generally because it is powering an amp, subwoofers and head unit
fine when it turns on.  Getting it to turn on is a little finicky, however.

If I do this, it will turn on about 90% of the time:

1) Set the power switch on the PSU to off.

2) Unplug from mains for 60 seconds.

3) Plug back into the mains.

4) Turn the power switch on.

Once it's up an running, shutting it off at the switch and turning it
back on may or may not get the power supply running again.

If I leave the PS_ON disconnected from the GND so that I can manually
make the connection a bit after the power switch is thrown, it seems to
work everytime fro the startup, but not necessarily for a "restart."

Does this pattern look familiar to anyone?  Is there a cheap and easy
way to make this system more reliable?

Thanks for your input.

Re: PSU as Car Stereo Power

Grinder wrote:
Quoted text here. Click to load it

In your captive application, you have two switches. The toggle
switch at the rear, controls +5VSB and the supervisor voltage
the switching regulator chip uses. You should apply the toggle
switch at the rear, with the connection from PS_ON# to COM open
circuit. That gives time for the supervisor voltage to

Once it's been given a second or two to stabilize, then you can
use the switch you connected to the PS_ON# and COM terminals,
to cause the main output rails (+12V) to come on. When you
do that, the switching chip has a timer, and it will ignore
short circuits on the output, for the next 35 milliseconds.
If it takes more than 35 milliseconds to charge up a large
capacitive load, then the switching regulator may latch up
in an overload state. And then the rear switch needs to be
slowly cycled (From ON to OFF, wait 30 seconds, back to ON again),
before the overload state will be cleared. If you need to clear
the overload state, again, it would pay to disconnect PS_ON# from
COM, so that the next time the supervisor comes up, it's not
being requested to "soft" switch on immediately.

Yes, the design should probably be able to tolerate "leaving
PS_ON# in the jammed on state". But modern power supplies will be
a little too clever for that. If you respect the fact there
are two switches, and do the switches one at a time, it's more
likely to work.

If you're still finding sequential application of the switches
leads to a failure to start (no fans spinning), then you may be
overloading the supply for longer than the 35 millisecond
grace period. A large capacitor bank in the head unit could do
that. A good head unit probably has "anti-thump" as a feature,
and the overload is probably not current going to the speakers.
But if the head unit has a capacitor bank for better dynamic
performance, that might be enough to piss off the ATX supply.


Re: PSU as Car Stereo Power

On 10/7/2012 10:52 PM, Paul wrote:
Quoted text here. Click to load it

Thanks Paul.  I'll put a PS_ON switch in and see if I can make a more
predictable startup process.

Re: PSU as Car Stereo Power

Hi Grinder,

The PSU should power on if you have the PS_ON and the GND connected. You
can use the power switch to switch on or switch the PSU. I recommend
that you switch on the SMPS (PSU) first and then switch on the head unit
and the amplifier. This would prevent the PSU from overloading. You may
need to add a switch If between PS_ON and GND to control the PSU if you
experience the same issue again.

Site Timeline