Arcade Collecting > Merit/JVL Touchscreen

Megatouch Force Boot problems

<< < (6/10) > >>

TalkingHeadsFan:
So, I took a fan from a garbage PC in my basement and plugged it into the motherboard; worked. So the fan busted in shipping. I held the fan against the processor to see if it would help cool it and boot further. It did. It went right to the main menu without a hitch. The touchscreen is acting funny but it probably needs calibration.

So now, I need a new fan and thermal paste. So why would the fan separate from the processor in the first place? Would the heat from an overheating processor liquefy the paste and snap it off?

obcd:
Usually, when a fan block it's due to build up dust.
To remove that from the heatsink, often the heatsink is removed from the cpu.
An air compressor is the ideal device to blew everything back to a nearly new condition. A vacuum cleaner doesn't work.
Usually, between the heatsink and the cpu there is a thermal pad as such is faster to assemble. If you remove the heatsink
afterwards, that pad usually splits, leaving parts on the cpu and others on the heatsink. Due to that, thermal coductivity of the
pad decreases. That's why it's better to remove the old stuff and add some new thermal paste to the heatsink and processor.
If you never removed the heatsink, you can try just changing the fan. Let the machine run for a period with the new fan. Restart it, go into bios and check the cpu temperature. If it's reasonable, you don't need to replace the thermal pad with thermal paste.
If the original fan is a three wire (with speed sensor) make sure the new one has a speed sensor as well. Some mobo's check their fan speed and some even feed it a pwm signal to change it's speed. In times of celeron cpu's, speed control usually didn't happen. 

lilshawn:
 ::)

Yet another good one obcd. I've worked on literally 100's of these machines. Not only that we own but what other game vendors have brought to me to fix. I've seen it all. While you info in general is sound, it doesn't apply here... in this case...specifically. and I'll explain why.

The fan failure is one issue, the thermal interface is the other.

These coolers originally used a solid waxy plastic thermal interface material ( TIM) that used the clamping pressure of the cooler and an initial heating of the CPU to melt the TIM and seal the 2 together. This is fine if it's never disturbed. But it's only ment for 1 use. This interface is hard...meaning if it gets knocked or banged it will break. And If the interface between the CPU and the cooler is broken, it no longer conducts heat. There is no fixing this contact or hoping it gets better. Or waiting and seeing. One use only. Its stupid and I haven't a clue why the used it.

The other TIM they used was a graphite coated aluminum strip. This was stupid as all hell too. You have a sandwich of CPU die, graphite, aluminum, graphite, aluminium cooler. See a problem here?

Bottom line:

Your issue is 2 fold here. A CPU with a still good TIM connection but a bad fan can run for hours before heat takes over and seizes up the system. Same with failed TIM and a working fan. But both? The CPU may as well be running bare without a cooler. This is why it runs for a bit then seizes up. This is also why way back in my initial post I mentioned removing the TIM and replacing it.

Final word:

You may get lucky and the CPU may be stable if you remove all that crusty yellow TIM from the CPU and cooler and replace it with some good paste and replace the fan. I recommend contact cleaner and some scrubbing to get it off.  The stuff is pretty tough, iso alcohol won't even touch it.

Once the TIM is good and the fan fixed, you should be in working order. If you find the machine is failing after a few hours the CPU was too overheated and will need to be replaced.

obcd:
Yes, I looked at the videos.
I didn't realise the thermal pad they used was that bad. I did find it strange that the system froze that fast, even in bios.
The cpu fan seemed so obvious to check, I thought it had been done, specially since the rams had been reseated.
Besides that, it's unclear how the fan became defective due to transportation.
My megatouch is the XL version, a bit older hardware still running under dos.

Isn't the parking feature something all harddisks are doing?
As the heads are moved using voice coils, isn't it normal they return to their start position if the coil isn't energised anymore?
Once upon a time (when the animals were still talking), they used steppers to move the drive heads.
You had to run the park program to move them to a safe location.

lilshawn:
Yes and no. Most desktop drives park the heads on a sacrificial section closest to the hub. Some new drives like the WD reds unload the heads like the drive I pictured.

If a standard drive is knocked hard enough, the heads can skid across the platters into the data area. On a drive where the heads are skidded off, they are effectively locked in place in the ramp area. Some magnetically others with a latch. While it's not impossible to knock these drives hard enough to dislodge the heads and have them knocked onto the platter, I'd be concerned of it being physically damaged.

As for how the TIM got broke...I've seen the crap the postal service does to packages...especially if they are marked fragile.

Navigation

[0] Message Index

[#] Next page

[*] Previous page