ATX power up on AC

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

I need my ATX pc to power up as soon as AC is applied, hence bypassing
the need to push the soft power on switch. I understand that shorting
pins 13 and 14 will turn on the power supply for testing, but will
this be safe when the p.s. is connected to the mobo? Won't this be
shorting some signal coming from the mobo?

Thanks.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

"Frank R." wrote:

> I need my ATX pc to power up as soon as AC is applied, hence bypassing
> the need to push the soft power on switch. I understand that shorting
> pins 13 and 14 will turn on the power supply for testing, but will
> this be safe when the p.s. is connected to the mobo? Won't this be
> shorting some signal coming from the mobo?

Your BIOS may have a provision for this. If not, try a small cap
(10-20 uF) across the ON/OFF switch contacts on your motherboard.
This will simulate a short upon AC power up, but quickly charge.
Shorting pins 13 & 14 on the PSU/MB connector works sometime, but
takes away the front panel switch, and has PWR_OK issues.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

That is (should be) a function in the BIOS.

--
Tumppi
Reply to group
=================================================
Most learned on nntp://news.mircosoft.com
Helsinki, Finland (remove _NOSPAM)
(translations from FI/SE not always accurate)
=================================================

"Frank R." <angkin2000@yahoo.com> kirjoitti viestissä
news:8942207.0407190453.100d97b1@posting.google.com...
> I need my ATX pc to power up as soon as AC is applied, hence bypassing
> the need to push the soft power on switch. I understand that shorting
> pins 13 and 14 will turn on the power supply for testing, but will
> this be safe when the p.s. is connected to the mobo? Won't this be
> shorting some signal coming from the mobo?
>
> Thanks.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On 19 Jul 2004 05:53:59 -0700, angkin2000@yahoo.com (Frank R.)
wrote:

>I need my ATX pc to power up as soon as AC is applied, hence bypassing
>the need to push the soft power on switch. I understand that shorting
>pins 13 and 14 will turn on the power supply for testing, but will
>this be safe when the p.s. is connected to the mobo? Won't this be
>shorting some signal coming from the mobo?
>
>Thanks.

One question to ask is "why?".
Often there is another solution. As Thomas Wendell mentioned
there is a setting in most modern bios for "resume on AC failure"
(may be worded slightly different) so that sudden ac failure,
recovery can be controlled. However it is not a good situation
to be in expecting sudden power failures, certainly an
appropriate UPS should be used instead, of a capacity allowing
continual uptime if that is important.

To more directly answer your question, yes you can just short the
PS-On signal to ground. For long-term, permanent modificatio it
might be better to use a resistor between the two instead,
perhaps 1K would suffice though it's just a guess, I"ve done it
before as a test but unfortunately don't recall the maximal value
possible and it may easily vary per particular PSU.

What will it matter if you're shorting signal coming from mobo
since that's exactly what mobo does when it keeps system running?
When system resets it keeps signal pulled low too, so it's more
an issue of expected power failure and recovery (as I initially
mentioned, why you feel the need to do this).

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

You guys are right about the bios. I found the "State after power
failure" and set it to "On". Problem fixed.

To answer the question "Why?": The pc is connected to a Voltage
Regulator and I wanted the pc to turn on using the regulator's power
switch. Actually the pc is for wife, to replace her old AT unit (with
power switch always on), who is used to
one-push-of-regulator-switch-turns-everything-on, e.g., printer,
scanner, external modem, powered speakers. I don't want to explain to
her the difference between AT and ATX styles.

Thank you all for your replies.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

angkin2000@yahoo.com (Frank R.) said:

> I need my ATX pc to power up as soon as AC is applied, hence bypassing
> the need to push the soft power on switch.

I don't know what your specific application is, but I suspect you are
applying power to a series of devices and want them all to start
simultaneously...
http://retroblast.com/reviews/smartstrip2-1.html

--
Mac Cool

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Mon, 19 Jul 2004 20:07:59 GMT, kony <spam@spam.com> put finger to
keyboard and composed:

>To more directly answer your question, yes you can just short the
>PS-On signal to ground. For long-term, permanent modificatio it
>might be better to use a resistor between the two instead,
>perhaps 1K would suffice though it's just a guess,

I don't think this is a reliable approach.

Intel's ATX PSU spec states the following:

===============================================================
PS_ON# is an active-low, TTL-compatible signal
The power supply shall provide an internal pull-up to TTL high.
VIL, Input Low Voltage 0.0V - 0.8V
VIH, Input High Voltage (Iin = -200 µA) 2.0V
===============================================================

If the PSU has an internal 4K7 pull-up resistor, then a 1K resistor
would result in a PS_ON voltage of 5 x (1 / 5.7) = 0.88V which is in
TTL limbo.

> I"ve done it
>before as a test but unfortunately don't recall the maximal value
>possible and it may easily vary per particular PSU.

I think so, too, as the spec does not specify a value for the internal
pull-up resistor.

>What will it matter if you're shorting signal coming from mobo
>since that's exactly what mobo does when it keeps system running?

That's true if the signal comes from an open collector driver, which
it probably does, but not if it is generated by a regular TTL gate.

>When system resets it keeps signal pulled low too, so it's more
>an issue of expected power failure and recovery (as I initially
>mentioned, why you feel the need to do this).

If a BIOS solution is unavailable, I would lift the PS_ON pin from the
power connector and wire it to a ground point on the mainboard. For
example, you could attach a ring terminal to a mainboard mounting
screw.


- Franc Zabkar
--
Please remove one 's' from my address when replying by email.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Thu, 22 Jul 2004 06:31:52 +1000, Franc Zabkar
<fzabkar@optussnet.com.au> wrote:

>On Mon, 19 Jul 2004 20:07:59 GMT, kony <spam@spam.com> put finger to
>keyboard and composed:
>
>>To more directly answer your question, yes you can just short the
>>PS-On signal to ground. For long-term, permanent modificatio it
>>might be better to use a resistor between the two instead,
>>perhaps 1K would suffice though it's just a guess,
>
>I don't think this is a reliable approach.
>
>Intel's ATX PSU spec states the following:
>
>===============================================================
> PS_ON# is an active-low, TTL-compatible signal
> The power supply shall provide an internal pull-up to TTL high.
> VIL, Input Low Voltage 0.0V - 0.8V
> VIH, Input High Voltage (Iin = -200 µA) 2.0V
>===============================================================
>
>If the PSU has an internal 4K7 pull-up resistor, then a 1K resistor
>would result in a PS_ON voltage of 5 x (1 / 5.7) = 0.88V which is in
>TTL limbo.

Glad someone was paying attention... you are quite right, 1K is
far too high a value, don't know what I was thinking at the time!

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

kony wrote:

> >Intel's ATX PSU spec states the following:
> >
> >===============================================================
> > PS_ON# is an active-low, TTL-compatible signal
> > The power supply shall provide an internal pull-up to TTL high.
> > VIL, Input Low Voltage 0.0V - 0.8V
> > VIH, Input High Voltage (Iin = -200 µA) 2.0V
> >===============================================================
> >
> >If the PSU has an internal 4K7 pull-up resistor, then a 1K resistor
> >would result in a PS_ON voltage of 5 x (1 / 5.7) = 0.88V which is in
> >TTL limbo.
>
> Glad someone was paying attention... you are quite right, 1K is
> far too high a value, don't know what I was thinking at the time!

But 4.7K is too low for a pull-up resistor in this application. A 4.7K
pull-up to 5v would provide almost a mA (1000uA) when the pin was pulled
to 0.8v, which exceeds Intel's spec. A minimum pull-up resistor would be
(5.0 - 0.8)/0.0002 or about 20K. With a 20K pull-up, kony's 1K resistor
would be fine.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Wed, 21 Jul 2004 22:19:00 -0700, ric <nospam@home.com> put finger
to keyboard and composed:

>kony wrote:
>
>> >Intel's ATX PSU spec states the following:
>> >
>> >===============================================================
>> > PS_ON# is an active-low, TTL-compatible signal
>> > The power supply shall provide an internal pull-up to TTL high.
>> > VIL, Input Low Voltage 0.0V - 0.8V
>> > VIH, Input High Voltage (Iin = -200 µA) 2.0V
>> >===============================================================
>> >
>> >If the PSU has an internal 4K7 pull-up resistor, then a 1K resistor
>> >would result in a PS_ON voltage of 5 x (1 / 5.7) = 0.88V which is in
>> >TTL limbo.
>>
>> Glad someone was paying attention... you are quite right, 1K is
>> far too high a value, don't know what I was thinking at the time!
>
>But 4.7K is too low for a pull-up resistor in this application. A 4.7K
>pull-up to 5v would provide almost a mA (1000uA) when the pin was pulled
>to 0.8v, which exceeds Intel's spec. A minimum pull-up resistor would be
>(5.0 - 0.8)/0.0002 or about 20K. With a 20K pull-up, kony's 1K resistor
>would be fine.

The 200uA spec is for a High input, not Low.

In any case, see this PSU design:
http://www.pavouk.comp.cz/hw/en_atxps.html

It has a 4K7 pull-up resistor.


- Franc Zabkar
--
Please remove one 's' from my address when replying by email.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Franc Zabkar wrote:

> >But 4.7K is too low for a pull-up resistor in this application. A 4.7K
> >pull-up to 5v would provide almost a mA (1000uA) when the pin was pulled
> >to 0.8v, which exceeds Intel's spec. A minimum pull-up resistor would be
> >(5.0 - 0.8)/0.0002 or about 20K. With a 20K pull-up, kony's 1K resistor
> >would be fine.
>
> The 200uA spec is for a High input, not Low.
>
> In any case, see this PSU design:
> http://www.pavouk.comp.cz/hw/en_atxps.html
>
> It has a 4K7 pull-up resistor.

Mea culpa. The Low spec is 1.6 mA @ 0.4v, or about 3K. I'd still go
with a 10K to be on the safe side, in which case kony's 1K would be
fine. Maybe 470 ohms to be safe. As far as the schematic at the URL
you gave goes, I think it's use of a 4.7K for a pull-up would be an
exception, rather than a rule.

Anybody care to look in some actual PSUs?

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Thu, 22 Jul 2004 12:18:53 -0700, ric <nospam@home.com> put finger
to keyboard and composed:

>Franc Zabkar wrote:
>
>> >But 4.7K is too low for a pull-up resistor in this application. A 4.7K
>> >pull-up to 5v would provide almost a mA (1000uA) when the pin was pulled
>> >to 0.8v, which exceeds Intel's spec. A minimum pull-up resistor would be
>> >(5.0 - 0.8)/0.0002 or about 20K. With a 20K pull-up, kony's 1K resistor
>> >would be fine.
>>
>> The 200uA spec is for a High input, not Low.
>>
>> In any case, see this PSU design:
>> http://www.pavouk.comp.cz/hw/en_atxps.html
>>
>> It has a 4K7 pull-up resistor.
>
>Mea culpa. The Low spec is 1.6 mA @ 0.4v, or about 3K. I'd still go
>with a 10K to be on the safe side, in which case kony's 1K would be
>fine. Maybe 470 ohms to be safe. As far as the schematic at the URL
>you gave goes, I think it's use of a 4.7K for a pull-up would be an
>exception, rather than a rule.

No, 4K7 is a typical TTL pullup in many different designs, not just
ATX PSUs.

>Anybody care to look in some actual PSUs?

Measure your own. With the PSU disconnected from the mains, measure
the resistance between the PS_ON and 5VSB pins.

Otherwise, here are 25 DTK SMPS circuits:
http://www.dtk.com.tw/tech/dtkfaq.html

4K7 appears to be the norm.


- Franc Zabkar
--
Please remove one 's' from my address when replying by email.