R!f@@

Hope you get the project done in time - I don't want to be responsible for you missing the deadline!! Do you really live on the Maldives too!!??

The photos are of one of the dead boards. The Intel Atom dual core chip is soldered unforunately.

My first install had the components mounted on a simple wooden board. This was fine (for 6 months+) until I drove over a small pothole in the road and the PC just stopped - suggested something perhaps shorted??

For the replacement board, I thought it would be better in a proper mini ITX enclosure. I was testing the PC in the car and everything booted up and then again suddenly stopped with the same apparent damage. I seem to remember the ITX case tapped against the plastic dash just as it all went off.

No smell of burning, no smoke.

Yesterday I decided to remove some of the components I had attached to the board which were still in the car. The only thing worth noting was the USB FM tuner. I had soldered a car aerial adapter onto this and noticed that one of the aerial signal wires had become detached. There was noticeable verdigris on the other wire. The car aerial is powered/amplified...

I plugged the FM radio tuner into my home PC to test and it is still working.

Other components plugged into the motherboard included a PCI wireless NIC (working), the USB signal from the touchscreen monitor (monitor working fine), GPS receiver with serial and PS/2 connector (not tested), audio jacks to amplifier (front and rear), VGA to monitor, 2.5" SATA HDD. One 2Gb DDR2 memory.

In addition to the SATA HDD and ATX + 4pin P4 to the motherboard, the PSU was also running 2 x cold cathode fluorescent tubes.

That's all I can think of!

In addition to the regulator/transistor getting hot, the southbridge does get slightly warm even with no other signs of life from the mobos, other than the green "ATX power connected" LED near the RAM slot.

I'm not sure if it is possible, but would expect any problem to have come through a USB slot. Otherwise, it has to be the PSU!!?

Thanks,

David

 

I've obviously tried both boards with a standard ATX PSU and got the same no-working results.

I've also tried the M2-ATX with a standard desktop board using a 12V supply and checked the voltages which were all spot-on.

Please amplify these statements.

What voltages did you measure on your 2 non functional boards both with the car psu and the standard bench psu?

I'm particularly interested in differences so please tabulate them all.

I also take it that remounting has not revived anything?

Have you checked the memory modules and do you have access to a post card?

 

I asked for the second one picture to compare the two boards david.
If you have it. please post it.
ASAP
PS I don't live on the Maldives, I live in Maldives and am a Maldivian
as for my project, I have good reputation in getting the work done.
it's no biggie
I'm a MULTI TASKER

 

Here's the boards side by side.

A couple of the components have different origins, most notably the south bridge.

Note the CPU on these boards has no fan.

http://www.tapsupply.co.uk/CVR4/both.jpg

 

studiot - what I meant by that post was that I had tried both of the non-working mobos with a standard desktop mains PSU and neither one worked.

I also checked the M2-ATX car PSU for output voltage and the results were as should be expected.

I've re-checked and get the following (with an Antec ATX PSU tester and DMM)

Yellow 11.88v
Blue -11.99v
Red 5.00v
Orange 3.28v
Grey 1.92v

I did not measure the voltages on the non-working mobos as I cannot see how this would be any different at the connector and have no idea where to test on-board.

 

All the power supplies are well within tolerance.

But the grey wire is a signal (Power OK) cable and is way out of spec.

The target voltage here is +5.0, however the tolerance is usually 3 - 6 volts. Some systems will operate down to 2.4 volts but none below 2.

 

Studiot & R!f@@ - you are both being very helpful.

Studiot - I tested the grey wire on my spare 240v ATX PSU and got 1.9v with this too so I can only imagine it has something to do with the Antec ATX tester which I expect does little more than the paperclip to green and black trick.

 

The voltage on the grey wire should be dictated by the power supply.

Its purpose is different from the green wire which is momentarily grounded by the on/off switch to initiate startup of the power supply. Once startup is initiated the green wire has a no control dead time up up to 10 seconds.

The power supply itself should hold the grey wire low (near ground) after initiation until all its internal self checks are complete and the ouput rails have stabilised.

So long as the grey wire is low the motherboard (which has some power from the auxiliary 5 volt supply) holds the reset line of the processor down so the processor is in a state of permanent reset.

When the psu is ready it asserts the grey wire (+5) and holds it there. If it looses power it drops the grey wire low and the processor is reset.

Many cheap PSU testers do not show if the voltages are correct, just that they are present or not.

The real question is why is your power_good line low?

Is it
1) The motherboard is faulty and loading it down.

or

2) The psu is faulty.

I have seen pcs where one of those psu tester shows everything hunky dory, but the pc keeps failing. A case of problem (2) cured by a new psu.

 

There should also be +5 volts on the purple wire.

This is the standby supply and also powers those parts of the motherboard which participate in startup.

It should never be off.

 

Hey Studiot - you know your stuff!

Sorry I missed off the purple line - running at 4.99v

The problem is that to test the PSU I used the Antec device to short the green to black and generate a draw on the PSU - I am unable to connect to the motherboard at the same time. So the voltages recorded are purely from the PSU, the M2-ATX DC-DC PSU.

Next I did the same thing with a spare standard, 240v desktop PSU and got near enough the same 1.9v on the grey signal wire.

The Antec device does not measure voltage and only features an LED indicator. I used a digital multimeter and probes for the voltage measurements.

I understand what you are saying about the 5v on the signal wire but this does not explain why the mains PSU, which I know works, reads the same voltage across the grey wire.

It also doesn't explain why the M2-ATX DC-DC PSU will fire up the old Asus A7N8X motherboard I have as spares.

 

Ok, to settle the issue with the grey signal wire, I connected the M2-ATX PSU to the old A7N8X motherboard.

Signal wire measures 0v. A quick short of the power switch and the board fires up and the grey wire now measures 4.64v

On one of the fried boards I get 0v on the grey signal wire and this remains so even after trying the power switch.

So the 5v on the grey wire is only there if the ATX connector is on a motherboard and the motherboard is 'switched' on.

 

This seems to be telling you that you have case (1) faulty motherboard.

I know you knew that but now you also know that the mobo is loading down the PS_Good signal (grey) on an other wise good power supply. I don't do much component level repair to boards these days as it is usually uneconomic.

Perhaps old RiP Van Winkle will help here when he wakes up.
I'm envious, as a diver, I'd love to be in the Maldives.

 

Morning guys..this is quick reply, I just got up...will go thru post after lunch.
RiP Van Winkle?
who the heck is that?

 

Dav

Verify me the things below

http://i702.photobucket.com/albums/ww28/Rifaa-ath/test.png

note
The PSU fan and the chipset fan should be running
Voltage measurements should taken after powering the board ON. If the PSU voltages are different verify with a known working PC PSU.
*these are tolerance levels with the given voltage below
*12V should be between 11V to 12.25V
*3.3V should be 3.2V to 3.35V
*5V should 4.75V to 5.25V
@ Voltages are referenced to GND only
If the components are emitting smoke during measurement shutdown and take measurement within different time intervals to prevent thermal damage to the board

U can give me the details of both the boards, since you have two there is better chance in fixing them.
After measurement power down till further notice.
And please be careful
Rifaa

 

I'm envious, as a diver, I'd love to be in the Maldives.

You a diver..wow
I would love to dive.
Hey come here and do let me know.

 

Studiot and R!f@@ - I see a wonderful friendship blossoming!

R!f@@ - thanks for all your work on the image and write-ups. The problem I have is that both boards are dead!! No fans spinning.

No amount of prodding the "power on" pins will do any good! Despite the fact that the motherboards are dead, the regulator I mentioned originally gets hot and the southbridge gets a little bit warm. There's a green ATX power led on board next to the RAM slot which lights but that's it!

 

Hmm ! good. So no power up but a green light means tht Stand by is coming.
Next step.
see the png I labelled.
Put the DMM in diode check with -ve probe on GND.
note down all the readings you get on all the supply connector pins. ie. the 24 pin and 4 pin connector, except GND's.
Post the result from the two boards and keep a note of which is which, these things tend to get mixed up sometimes.
Post back, the sooner the better.
We can even be quicker if you have a hotmail ID.

 

R!f@@

Sorry for the delay - trying to work on a web site too!

Ok, here goes:

OLDER MOBO (more fried possibly as was in the car with power connected for longer after failing)

Pin 1 (OR) : 900 - 940
Pin 2 (OR) : 900 - 940
Pin 3 (BLK)
Pin 4 (RED): 445 - 470
Pin 5 (BLK)
Pin 6 (RED): 445 - 470
Pin 7 (BLK)
Pin 8 (GRY): 20
Pin 9 (PUR): 1
Pin 10 (YEL): 1532
Pin 11 - no connection
Pin 12 - no connection

Pin 13 (OR) : 930
Pin 14 (BLU):204
Pin 15 (BLK)
Pin 16 (GRN): 1
Pin 17 (BLK)
Pin 18 (BLK)
Pin 19 (BLK)
Pin 20 - no connection
Pin 21 (RED): 445 - 470
Pin 22 (RED): 445 - 470
Pin 23 - no connection
Pin 24 - no connection

4-PIN connector reads 1500 on each YELLOW

--------------------------------------------------

NEWER MOBO

Pin 1 (OR) : 405
Pin 2 (OR) : 406
Pin 3 (BLK)
Pin 4 (RED): 450
Pin 5 (BLK)
Pin 6 (RED): 450
Pin 7 (BLK)
Pin 8 (GRY): 17
Pin 9 (PUR): 1
Pin 10 (YEL): 1202
Pin 11 - no connection
Pin 12 - no connection

Pin 13 (OR) : 398
Pin 14 (BLU): 202
Pin 15 (BLK)
Pin 16 (GRN): 1
Pin 17 (BLK)
Pin 18 (BLK)
Pin 19 (BLK)
Pin 20 - no connection
Pin 21 (RED): 450
Pin 22 (RED): 450
Pin 23 - no connection
Pin 24 - no connection

4-PIN connector reads 1200 on each YELLOW

------------------------------------------------

Reading on 3.3v is quite different!

 

Purple and green wire reading are out of order. next step is to get the resistance readings. short the pins of all elec. caps to discharge them (disconnect the short after a few secs.)disconnect the fan cable and remove RAM and give me the resistance of all the pins to ground

 

One more pic please. I like to see the CAR PSU power cable to mobo connector