It worked for three weeks

After seemingly fixing that broken laptop it has failed to boot up this morning!

I noted that it has been running hot again, the heat sink port on the side has been getting seriously warm so if I do attempt another re-flow of the GPU (if that is the fault) I will have to look at improving the cooling. I am thinking I might drill some holes through the bottom casing to allow more air in and might fit an override switch on the fan since it seems to only come on when very very warm.

What is doubly annoying is that over the weekend I managed to receive what I believe are the FITSAT-1 and TechEdSat Cubesats. My earlier attempts last week were only partially successful and the one time I did manage to get really good signals I forgot to set it to record the IQ file.

Over the weekend I was away from home, visiting the in-laws. I had taken a scanner, my FUNCube Dongle and laptop. I appropriated an indoor TV aerial and surprisingly managing to get some clearly audible signals, despite major issues with pager breakthrough and interference caused by the aerials wideband amplifier.
 

As soon I manage to get the laptop working, or recover the IQ files from its hard drive I will post the results.

If at first you don’t succeed… try, try, try again

Hopefully yesterday saw the conclusion to the saga of my broken laptop.

Two weeks ago I had an attempt at reviving a sickly HP/Compaq 6735S a model which is notorious for GPU problems. Specifically the GPU BGA contacts break due to thermal stresses and shoddy manufacturing.

I had stripped the laptop down and using a IR rework station attempted to reflow the contacts with limited success. It survived one power up and then died again.

Like Robbie Burn’s famous spider I was not going to give up that easily so I stripped it down again and this time remembering the 7 Ps adage I didn’t cut any corners.

The picture above shows the motherboard after I removed it the first time, you can clearly see the GPU in the top right. Notice first the pathetic thermal bonding? Also you can see around all four corners of the chip what appears to be red cement.

Initially I was unsure of whether to remove this and indeed it appeared to be tough and nigh on impossible to remove so fearing possible collateral damage I left the cement in place and heated the GPU on the IR rework station attempting to reflow the joints.

This I believe was the mistake as this bonding would have prevent the chip from floating on the reflowed solder pads. So this time I did remove the cement as you can see below. I found heating the cement did make it malleable, the problem being how to heat it without damaging any of the surrounding components?

In the end I used a precision hot-air gun (which they also have at work) on a low-medium heat setting on a very low ‘blow’ setting. Using this and a couple of wooden toothpicks I was able to remove the vast majority of the red gunk!

Once removed I applied a small amount of flux around the chip, working it under the device and then used the IR rework station but on a much higher setting than before to ensure the solder properly melted.

To my pleasant surprise the reassembled laptop booted and it has been successfully been powered down and rebooted several more times since.

While not the prettiest of laptops, or indeed in the best condition it does at least have Windows 7 Professional on it and a nice big hard drive and 4GB of memory so should prove more than capable for my experiments – assuming she carries on working, and she does get a little warm!