What causes PCB failure on a HDD?

[Skip Da Shu]

Originally Posted by Adrian Wong

I also believe heat is a contributor to HDD death and I have a fan blowing on all my HDDs. However, the Google study says that heat doesn't affect HDD reliability...

The google study says heat doesn't have a measurable correlation to failures on HDDs UNDER 5 years old... after that there's a very direct relationship. So think about that. No factor for 5 years then it is a factor! What that tells me is that it IS a factor just a very small factor that starts to INCREASE with the age of the HDD. It's just that it's not statistically measurable in their study till the drives got to their 5th birthday! YES I have several GNU/Linux OS drives that have seen their 5th BDay. Hmmm, think I have 1 or two that may have hit their 10th and these pooches run 24/7/365 in a non-A/C'd room (currently @ 95F). However they have AIRFLOW. It's that dead hot air that'll cause the internal temps of these old dogs to climb, climb, climb and then die. If you can just move a wee bit of air across them they just keep running. But then my old dogs are good ol' lead solder and epoxy you can't pry apart with a screwdriver HDDs! Check back with me in 10 years to see if my 320/500GB HDDS are still alive (also see if I'm still breathing)!

Think I'll go throw the ol' dogs a BDay party

[1031982]

Well, I have the replacement installed.

I decided to try the PCB swap out to see if I could get anything, and it was no good. Thing is, on the old HDD's board the contacts to the motor and some of the contacts for the power looked burned.

I checked my PSU with a multimeter, and everything is fine so it might be the ROHS degradation mentions by empire23. It could have been a defect as well. I can't tell, and shipped the bad drive back already.

With the fan in my case is has space between the fan and the HDD's mounts. No contact with the drives or the mounts.

As for heat, both my drives were and are now at 34-36C. Under high usage, the Raptor went up to 41C. Well within operating range.

[empire23]

Originally Posted by Adrian Wong

Wow, I didn't know that! So all those ROHS-certified parts won't last as long?

In my opinion and the majority of the electronics community, yes.

And to add to our burden is the fact that lead free solder requires a higher temperature to liquify, thus you have a hotter solder. Add the fact that you're wave soldering, and you'll have to bump that figure up even higher than even normal tin+lead solder needs in order to be wave solderable. This extra heat generally introduces more stress in the board, and sometimes heat shock within components.

Last if not least, IIRC lead free solder is hard to get Eutectic. Eutectic generally means that when a solder crosses the bounds of liquid form to solid form (alpha+beta), it immediately solidifies as a whole and doesn't form a slurry which can compromise the joint in the long term. Eutectic solders are extremely reliable due to their grain structure when cooled and more or less come out just right, so have a non-eutectic joint is inferior to a eutectic one. All of this generally means lower reliability with pb-free.

[zy]

materials.. i hate materials... nice info btw.

[empire23]

Did pretty well for materials eng in undergrad. Hopefully the 2nd level subject next sem will be more forgiving than i reckon it'll be. But it's still nowhere as close to what the mechs learn.

[Adrian Wong]

Originally Posted by Skip Da Shu

The google study says heat doesn't have a measurable correlation to failures on HDDs UNDER 5 years old... after that there's a very direct relationship. So think about that. No factor for 5 years then it is a factor! What that tells me is that it IS a factor just a very small factor that starts to INCREASE with the age of the HDD. It's just that it's not statistically measurable in their study till the drives got to their 5th birthday! YES I have several GNU/Linux OS drives that have seen their 5th BDay. Hmmm, think I have 1 or two that may have hit their 10th and these pooches run 24/7/365 in a non-A/C'd room (currently @ 95F). However they have AIRFLOW. It's that dead hot air that'll cause the internal temps of these old dogs to climb, climb, climb and then die. If you can just move a wee bit of air across them they just keep running. But then my old dogs are good ol' lead solder and epoxy you can't pry apart with a screwdriver HDDs! Check back with me in 10 years to see if my 320/500GB HDDS are still alive (also see if I'm still breathing)!

Think I'll go throw the ol' dogs a BDay party

That makes sense. I always felt that heat would be a major "cause of death" for HDDs. Even back in the days of 5400 RPM hard disks, I would install fans to keep them cool.

Only lost one drive so far, and it was due to a PCB failure.