My SurfacePro 4 won't start!

Wolf K wrote:
On 2018-11-22 17:35, Peter Jason wrote:
On Thu, 22 Nov 2018 16:47:35 -0500, Paul
wrote:

Peter Jason wrote:
On Wed, 21 Nov 2018 15:14:11 -0700, 123456789
wrote:

On 11/21/2018 1:56 PM, Peter Jason wrote:

I've had it on recharge for days. Sometimes it gives a small
glimmer, then stops. Is there any way to fix it?
Have you tried these (first line in a Google search):

https://windowsreport.com/cant-turn-surface-pro-4-on/

Thanks, I just did and I got it started with the
suggestion to hold down the start button for 30
seconds.

What does that do, officially speaking ?

Does it reset some hardware controller on there
or something ?

Paul

I don't know. But it starts OK now. Perhaps the
starter-switch contacts are degraded. This can
happen to roller-door switch contacts too.

I seized the opportunity to download updates going
back to July, and there were heaps of them,
including Intel & firmware ones.

Press and hold for several seconds has also worked for me with desktops
that were turned off instead of shut down. Before I used a UPS a couple
of reboots after a power outage (-- no proper shutdown) brought Windows
back to normal operation.

I suspect that Windows does some kind of self-repair and/or cleanup
before boot.


But that's a hardware conditioning feature for the Power button,
at least on desktops.

The Power button on the front of your machine is momentary contact,
Normally Open (NO). Pressing the button, makes a momentary
logic zero on the signal line.

The chipset sees the signal, and that would normally
kick off startup. The signal is latched. When the circuit
is in the "the power is officially off" state, the necessary
signal width is very short. Just a pulse on the upper
signal is enough.

_____ _______
|___| Power control, logic signal

_________
|_______ PS_ON# To ATX PSU

To protect against your cat pressing the button by
accident, once something intelligent is running on
the computer, the chipset circuit enters the
"the power is officially on" state. To FORCE the
power off requires holding the button down for
four seconds.
_____ _______
|_______________| Power control, logic signal

- 4 sec - ___________
__________________| PS_ON# To ATX PSU

So this is a pretty consistent behavior, that
prevents accidentally brushing against the switch
and losing a session. If the button needed only a
short "pulse" to turn off the power, then even a flaky
wire, one that gave some electrical noise, would
be enough.

The front panel Reset button on the case, is
not conditioned. There's no 4 second timer on that
one, although you'd wonder why.

The longer time constants (like pressing and holding
some button on your router), it's possible those
are measured with the processor. The Surface might
be doing that. At least some Surface products,
use S0iX and don't have the normal ACPI states.
Such Surfaces are "always awake" and it wouldn't
be a problem to modify the above timing
diagrams so that the Surface starts monitoring
the Power button when the falling edge of the
signal happens. Or, a microcontroller on board
could do it, if perhaps battery charging has
a management chip, and the Power button is
routed through such a controller.

But traditional desktops have had that 4 second
behavior enforced in hardware. Either the Southbridge
or the SuperI/O could have the function, and the
function runs of +5VSB. If +5VSB isn't present,
the logic that conditions the power button
cannot work.

The Power button then, is "fly by wire". It isn't
a direct control. Whereas the switch on the back
of your PC on the ATX supply, that's a direct
control.

So if smoke is coming out of your PC, *always*
use the back switch. If the back switch itself
catches fire, pull the cord out of the wall :-)
I had a power switch on a computer disintegrate
in my hands once, and while nothing bad happened
in that case, it could have. If you're one
of these people who hides PCs in broom closets,
you should always have a means to get at the
outlet supplying power, so you can pull the
cable in an emergency (before the smoke stinks
up the house). I've seen some kiosk installs
where you'd have to run for a breaker panel in
the event of a fire event, and that's a hell of
a time to discover you don't know where the
breaker panel is.

Paul

On 11/21/18 12:56, Peter Jason wrote:
I've had it on recharge for days. Sometimes it
gives a small glimmer, then stops. Is there any
way to fix it?


is your battery fried? if it's old enough, probably just that.


--
(aka 'Bombastic Bob' in case you wondered)

'Feeling with my fingers, and thinking with my brain' - me

'your story is so touching, but it sounds just like a lie'
"Straighten up and fly right"