case power switch replacement

I just purchased a nice Lian Li case and installed into it some spare
parts to build a system. There was a problem booting the system (see my
other message). While working on this problem, the front power switch
ceased to function.

That is, when I toggle the switch on/off by pressing it, and with the
wires disconnected from the MB, there is an open circuit in both
states. However, if I depress the switch half-way, there is a low
ohm reading.

So the switch looks bad. It looks easy enough to replace, and a lot less
expensive and troublesome than trying to find packaging for the case and
returning it to the vender (NewEgg). I wrote Lian Li to ask for a
replacement switch or to tell me who makes it, but I'm not optimistic
I'll hear a reply.

These switches must be made in great quantity. It is a SPST switch that
toggles ON/OFF each time you press it. A simple white plastic body is
inserted through a 32 mm. square hole in the front of the case. They
probably are called subminiature push-button switches.

Does anyone have any knowledge of these switches, such as the extent
they are a standard item and who might make them? Pouring through my
Allied catalog didn't help.

--

Haines Brown, KB1GRM

Previously Haines Brown <> wrote:
> I just purchased a nice Lian Li case and installed into it some spare
> parts to build a system. There was a problem booting the system (see my
> other message). While working on this problem, the front power switch
> ceased to function.

> That is, when I toggle the switch on/off by pressing it, and with the
> wires disconnected from the MB, there is an open circuit in both
> states. However, if I depress the switch half-way, there is a low
> ohm reading.

> So the switch looks bad. It looks easy enough to replace, and a lot less
> expensive and troublesome than trying to find packaging for the case and
> returning it to the vender (NewEgg). I wrote Lian Li to ask for a
> replacement switch or to tell me who makes it, but I'm not optimistic
> I'll hear a reply.

> These switches must be made in great quantity. It is a SPST switch that
> toggles ON/OFF each time you press it.

That would be very surprising and require a modified mainboard.
These switches are typically on when pressed and off when not
pressed.

> A simple white plastic body is
> inserted through a 32 mm. square hole in the front of the case. They
> probably are called subminiature push-button switches.

> Does anyone have any knowledge of these switches, such as the extent
> they are a standard item and who might make them? Pouring through my
> Allied catalog didn't help.

The plastic cap is usually not part of the switch. For the
switch you will likely have to get soemthing similar and adapt.

Arno

Arno Wagner <> writes:

> Previously Haines Brown <> wrote:
>> So the switch looks bad. It looks easy enough to replace, and a lot
>> less expensive and troublesome than trying to find packaging for the
>> case and returning it to the vender

> That would be very surprising and require a modified mainboard. These
> switches are typically on when pressed and off when not pressed.

Ahaa! Again my ignorance has betrayed me. You are quite right: When the
switch is depressed, it closes the circuit; when released, the circuit
is open.

So now I'm back to my original - well, not original, but the subsquent
problem. Originally when I booted, all I got was the MB spash
screen. Then, after turning the machine on its side, when I pushed the
start switch, nothing at all happens. The MB LED lights when the power
line is connected (no matter which position the PS toggle is in), but
nothing else happens: no fans, etc.

I guess the next thing would be for me to test power out. I find that
when I push the on switch, I get no +5/+12 V out from the peripheral
power connections. I suppose this could be due to a) blown fuse in the
PS, b) a short causing the supply to shutdown, c) a broken power
supply. Am I correct to assume that the power to peripherals always
comes right on whenever the power switch is depressed?

a) I guess I can't replace the fuse without voiding my warrantee. I
measured the resistance across the line voltage socket and got only 0.04
ohms, which strikes me as suspicious.

b) A short is possible. The power went after I had turned the machine on
its side. The stand-offs holding the MB seem tall enough and don't
believe I put any pressure on the MB backplate. I suppose I could
disconnect everything to eliminate the possibility of a short outside
the PS and see what happens. I gather to test supply voltages, I have to
make sure the supply has a load of some kind. Is a connection to power
the motor of one hard disk enough of a load?

--

Haines Brown, KB1GRM

Previously Haines Brown <> wrote:
> Arno Wagner <> writes:

>> Previously Haines Brown <> wrote:
>>> So the switch looks bad. It looks easy enough to replace, and a lot
>>> less expensive and troublesome than trying to find packaging for the
>>> case and returning it to the vender

>> That would be very surprising and require a modified mainboard. These
>> switches are typically on when pressed and off when not pressed.

> Ahaa! Again my ignorance has betrayed me. You are quite right: When the
> switch is depressed, it closes the circuit; when released, the circuit
> is open.

> So now I'm back to my original - well, not original, but the subsquent
> problem. Originally when I booted, all I got was the MB spash
> screen. Then, after turning the machine on its side, when I pushed the
> start switch, nothing at all happens. The MB LED lights when the power
> line is connected (no matter which position the PS toggle is in), but
> nothing else happens: no fans, etc.

> I guess the next thing would be for me to test power out. I find that
> when I push the on switch, I get no +5/+12 V out from the peripheral
> power connections. I suppose this could be due to a) blown fuse in the
> PS, b) a short causing the supply to shutdown, c) a broken power
> supply. Am I correct to assume that the power to peripherals always
> comes right on whenever the power switch is depressed?

Yes, but the swiotch actually runns to the mainboard and the PSU
is controlled by the MB as well. This allows swicthing off
under software control.

> a) I guess I can't replace the fuse without voiding my warrantee. I
> measured the resistance across the line voltage socket and got only 0.04
> ohms, which strikes me as suspicious.

Indeed. But unless you have a special low-Ohm meter, you cannot
measure a resitsance this low. I suspect a wrong meter setting.

> b) A short is possible. The power went after I had turned the machine on
> its side. The stand-offs holding the MB seem tall enough and don't
> believe I put any pressure on the MB backplate. I suppose I could
> disconnect everything to eliminate the possibility of a short outside
> the PS and see what happens. I gather to test supply voltages, I have to
> make sure the supply has a load of some kind. Is a connection to power
> the motor of one hard disk enough of a load?

You can test most PSUs without load, if the are on only for a short
time. A disk motor is quite enough, if you want load.

Arno

Arno Wagner <> writes:

>> a) I guess I can't replace the fuse without voiding my warrantee. I
>> measured the resistance across the line voltage socket and got only 0.04
>> ohms, which strikes me as suspicious.
>
> Indeed. But unless you have a special low-Ohm meter, you cannot
> measure a resitsance this low. I suspect a wrong meter setting.

Dunno. The DMM I have switches scales automatically.

>> b) A short is possible. The power went after I had turned the machine on
>> its side. The stand-offs holding the MB seem tall enough and don't
>> believe I put any pressure on the MB backplate. I suppose I could
>> disconnect everything to eliminate the possibility of a short outside
>> the PS and see what happens. I gather to test supply voltages, I have to
>> make sure the supply has a load of some kind. Is a connection to power
>> the motor of one hard disk enough of a load?
>
> You can test most PSUs without load, if the are on only for a short
> time. A disk motor is quite enough, if you want load.

Thanks. If the PS fan doesn't even come on, I'm not very optistic that
anything much else is working. I'm about to buy a reasonably inexpensive
duplicate PS so that I can test by substitution.
--

Haines Brown, KB1GRM

W_Tom very kindly passed along a test procedure for a power supply that
was far in advance of the primitive notions I had been carrying with
me. His procedure is found at:
http://groups.google.com/group/alt.w...8c24f533d2ab85

It involved the measurement of the voltage at various pins of the molex
main board power connector while the system was under full load.

At all test points, I got zero voltage. There's 117VAC at the power
cord; the power supply toggle toggle switch is in the 1 vs. 0 position;
the mainboard LED is lit; no voltage at the front panel power switch.

I thought the problem was this front panel switch, and contacted the
case manufacturer about it. I'm happy to report that the manufacturer,
Lian Li, was very responsive. I'm happy to say that a replacement switch
is in the mail, even though it turns out that the probem is in the
supply itself.

At one point, although I can't repeat it today, if I press the front
panel switch, hold five seconds, release for five seconds, and repeat
this, on about the third try the cpu fans makes a few rotations.
--

Haines Brown, KB1GRM