PSU Test

[netburst]

Hi guys,
I read somewhere that the psu can be turned on by joining one of the 2 pins in the 24 pin array block to test weather the PSU is working or not... Any ideas which 2 pins are these???
Cheers

 

[Groveling_Wyrm]

Pins 14 and 15, which are green and black, respectively.

 

[wouse101]

erm on my hiper i repaired (just the fan, not the power circuit), 14=blue, 15=black and 16=green, which from the manual is: -12V, COM (0V /GND i'm thinking), and PS_ON#. Its 24-pin too.

which would mean in this case, you short 15 & 16?

i managed to get it to get in standby mode, but couldn't get it to power on. the fans just "blipped" and nothing else, there is another connection called PWR_ON, but i never got round to testing as i was only on my lunch break!

i can't remember which pins i shorted( most likely PS_ON# and COM, but don't quote me) but i made sure, by reading the pin outs in the manual, and checking what voltage was on the pins with a multimeter , that i wasn't shorting a supply rail.

 

[aziraphale]

You'll have to to draw some power from the PSU in order for it to stay powered on. Just connect a CD ROM drive and then it should work...


Edit deleted

 

[wouse101]

ahh cheers, thought i might need to.

 

[aziraphale]

Ok second input. It works. Just measured the rails and its powering on....

 

[aziraphale]

It even works without any device attached...

 

[wouse101]

how'd you do that? which is it? with or without?

 

[aziraphale]

You connect the green wire to any ground and keep it connected.. nothing attached. Then you could use a multimeter to measure all rails.

 

[wouse101]

got it, CD-ROM on output, fans power up, CD-ROM Ejects when pressed.

good enough for me.

found out that PS_ON# & COM shorted and PWR_ON & PS_ON# work both the same......found that COM to PWR_ON with a voltmeter is 0V....odd, perhaps more too it than that.

 

[aziraphale]

......found that COM to PWR_ON with a voltmeter is 0V....odd, perhaps more too it than that.

PWR_On is pulled up by default. You should read around 5V. If not then you've pinpointed the (or a) problem...

 

[wouse101]

hmm might have to double check that.

if its pulled up, should it not be low anyway in order to be dragged to a higher voltage?

 

[aziraphale]

No it's 5V high and it's going 0V once you connect it to gnd...

(btw; last transmission tonight. I'm in europe and it's getting late... I'll check back in about 10 hrs on the thread, and I'll be ready to experiment a lot more tomorrow...)

 

[wouse101]

no rush mate

 

[aziraphale]

Well I DID some experimenting today. I took my old PSU and converted it to a "lab" power supply. +5 -5 +12 -12 +3.3 and GND are now accessible via plugs. This gives a wide range of possible voltages. To turn it on I connected GND and the Green power_on wire to a switch. As power_OK indicator I use a LED connected to the grey Power_OK line (5V high), 220 resistor to GND. What I found out however is that most PSU's need a load to stay powered on. Usually a load on the 5V rail will be enough. For that purpose I use a 10 Ohm 17W resistor between 5V and GND. You should use (little bit of calculation) at least a 10W resistor. I mounted the resistor (which is pretty big) to the metal casing in front of the fan. IT doesn't get too hot like that. For a little gadget you could use the purple "standby" chord to indicate that you're turned off...

HOWEVER: that is pretty dangerous tinkering. Before you open a PSU you should be aware that the capacities inside will most probably still be charged and they can provide a strong and dangerous shock. Furthermore I would suggest to use fuses at least on the GND. 1A at most. But it would be better to use fuses on all connections. A short without any fuse between random rails would be pretty "hot". Especially with high power PSU's.

You see; our discussion yesterday has inspired me quite a bit

 

[wouse101]

good work.

Bit of over kill with the 10OHM resistor. 17W is a lot when 10ohms on 5v is only putting out 2.5watts.

but if thats all you had kicking around, then fair play.

 

[aziraphale]

Well at least the resistor doesn't get hot...

 

[wouse101]

true, overkill never hurt anyone