L1A heatsink fan longevity? any fail safe options?

[Bean]

The latest SPCR review has cautionary comments on using sleeve bearing fans on heatsinks. I have been wondering/worrying about this recently and those comments confirm my worries.

I built a computer for a relative and it stays on 24/7 (the kid doesnt remember/wont turn the computer off when not in use). If the fan dies, I have no way to know as I do not live there. If the fan dies what happens - a fire? seriously what are the consequences? what are my options? what do OEM's use for fail safe? (Ok I will slow down and take a breath now )

My system:
I have an L1A@7 volts sitting directly on top of an slk900 temps are probably 36-46cpu and 28-36case. 2400,35watt AMD mobile @10x166,1.3volt. Biostar M7NCG-400, WD80g hd (had no choice due to faulty samsung spinpoint), Antrec Aria.

[MikeC]

If the fan dies, the PC will crash sooner or later as the CPU gets too hot. Very unlikely that the CPU can be heat damaged as long as the HS is on the CPU. Fire? Totally unlikely.

Regarding the heat wear on the sleeve bearing of the fan, realistically, even if the PC is on 24/7, if the CPU temp is kept reasonaly low -- say under 50C -- the fan itself will never get quite as hot. Remember, its own airflow keeps moving the CPU heat away. That's not high enough a temp for a good sleeve bearing to wear out any sooner than a ball bearing fan IMO. With most PCs, idle temps are far lower than 50C, so...

This might be more of an issue w/a PC that's at 100% load 24/7 -- like if you are folding @ home. But on the other hand, most of my PCs have been folding 24/7 with Panaflo or other modified sleeve bearing fans for years -- with not a single fan failure ever. And I have no qualms about letting my CPUs hit 60C on a steady basis.

I have to also say that I was not cautioning about sleeve bearing fans -- instead, I was simply stating general industry practise. It's something to be aware of -- like if I built a hot quad CPU web server in a 1U case, I assure you I will be using nothing but ear-deafening high speed ball bearing fans! But as it would be sited in an airconditioned locked data room with a hundred other equally noisy machines...

Getting back to your original question, if you want to feel totally secure, just go and up the fan speed a bit to keep the CPU temp down and/or make sure the case exhaust fan nearby is pushintg a decent amount of air.

[Krispy]

My Volcano 12 was so noisy when I first switched it on I thought my PC was going to start hovering. What a pig of a fan.

I swapped it for slowish sleeve bearing'd Coolermaster I had lying around & it has stayed there ever since with no problems.
That was almost 2 years ago.

Hope this is of some reassurance.

[nici]

A quality fan will last, or should last, for several years of 24/7 running. Crappy fans may fail regardless of what type bearing it uses..

[ferdb]

Undervolted sleeve bearing fans last much longer than when running at full voltage. The primary mode of failure of sleeve bearings is the oil in the bearing gets spun out by centrifugal forces leaving the bearing short on lubrication and thus accelerating wear of the bearing surfaces. Lower RPM not only reduces this effect and the bearing stays better lubricated, but it also reduces the bearing surface speeds so wear doesn't occur as rapidly anyway. End result for this pair of effects is that the fans last a long long time.

[Bean]

Thanks everyone for your comments. I think I can rest easier now. Especially since I don't have to worry about a fire. whew! thats a big relief.

I was thinking CPU load got pegged during a crash. I guess that its just the opposite - the CPU stops working. is that about right?

I think I will leave the fan at its current speed then. As was pointed out I should be OK with a heatsink temp below 50, which it is - even during prime 95. and as also mentioned, the fan will last longer due to slower rotation. By the way I dont have a case fan (i didnt use the cheap,IMHO, aria pci slot fan), the 120mm power supply fan is the other fan.

Just for kicks, since my anxiety has been eliminated, is there a thermal switch of sorts or thermostat. Basically a device attached to the heatsink and inline with power switch. kinda like found on space heaters. Is there such a thing?

[Ralf Hutter]

FWIW, there are many software solutions to shut down a box based on sensor outputs.

Just one option are the many plugins that are available for MBM. They will shut down the box (or run an alarm or send an email or whatever you want) if the selected sensor output goes beyond a preset threshold. Use the CPU temp sensor or the heatsink fan rpm (if available) and you can safely shut down your system way before it starts running up against thermal issues.

[HellDiver]

Put it like this, I've several Panaflo 80mm fans running more or less 24/7 since 1990. 15 years of constant use, and they're still going. Good choice Commodore, because those yellow label Panaflos are still going strong in a serius number of "big box" Amigas.

I personally also use a more modern grey L1A Panaflo on a PAL8942 on my machine. No issues to report, and I fully expect this fan to outlive both this machine and several replacements.

[patord]

Which review was Bean referring to that mentions caution about using sleeve bearing fans on heatsinks?

I was trying to see if any mention was made about the fact that Panaflo and Papst fans, although sleeve bearing, use "custom" sleeve bearings which should have much greater life (in the type of lubrication and operating temps) then generic sleeve bearing fans. In my experience sometimes as much as 50% longer (exceeding even some double ball bearing fans). I've been replacing bobo sleeve fans in PC's and electronic equipment for many many years with Panaflo or Papst fans, so I know.

But given the fact it is getting harder to get Papst and decent Panafo fans now (atleast the kinds I use to get), I'm having to resort back to normal sleeve bearing fans in some sound sensitive applications (silent PC's and audio gear). But I stock extra's knowing they'll failure sooner then the P fans.

Though the discussions about operating temps in most PC's and the slower RPM's most may run normal sleeve bearing fans are somewhat valid, so maybe everything I just said is moot... it's just the manufacturing tolerances I've seen most generic sleeve bearing fans is sometimes depressing.

[cpemma]

"Sleeve bearing" is a term that's fallen into disrepute because it can mean "cheap bearing". But there are a multitude of alloys that can be used for the sleeve, and the better ones (eg Papst's Sintec) are powder-metallurgy forms that are porous so absorb some lubricant.

I suppose I'm saying you get what you pay for, or rather, expensive may be 50% hype but cheap is usually poor quality.

But as Ralf Hutter said, most monitoring programs (Speedfan, MBM) and most modern motherboard BIOSes have an alarm feature.