JP Vossen on 2 Dec 2015 09:21:20 -0800

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: [PLUG] fans

Yeah, I can't argue with you.  As you say, "it depends."  My case was a
bit different, and I know the WD-40 thing is...sub-optimal, we'll say. :)

I will stress what we all know anyway though, re-seating (and perhaps
cleaning contacts) is certainly worth trying for a "fan sensor that is
not reading reliably"...  Re-seating things often solves a multitude of
sins, especially on devices that power cycle and thus heat and cool and
expand and contract.  This is less of a problem now, with so much more
integrated than it used to be, but it's still a useful reminder.

And there's always percussive maintenance!  ;-)

On 12/02/2015 12:06 PM, Rich Mingin (PLUG) wrote:
> I respectfully disagree. A failing fan due to bearings going, or crud
> near the hub will work harder, get hot, get louder, and move less
> air/spin slower, but will generally have a steady RPM.
> The "sawtooth" RPM cycle (start low, increase to high RPM, reset to low)
> indicates a fan that is working normally (or at least isn't the main
> problem), and a fan sensor that is not reading reliably or at all.
> Fan controllers start the RPM low, at a level known to spin the fan at
> low speed, and increase the voltage/speed until they sense RPM from the
> fan sensor. If they reach full RPM w/o sensing RPM, they will generally
> reset and start over, for at least a few tries, and will then trigger a
> warning or alarm via the BMC.
> So while cleaning the fan is always a good idea, and lubing the bearings
> every year or two is also great (though I don't think I like your
> method), the symptoms here indicate a problem with the sensor/BMC, not a
> problem with the fan itself.
> OP/Jeff: Can you clarify if this is what you meant by "revving noises"?
> Speed ramping up and dropping back? If you meant it was just way louder
> than usual, JP's advice may be better, though a replacement fan would
> work for either problem.
> On Wed, Dec 2, 2015 at 10:39 AM, JP Vossen <
> <>> wrote:
>     Short version:
>     1) Clean it out with compressed air, etc
>     2) Re-seat the fan & connectors
>     3) Possibly hit the fan with WD-40
>     Long version:
>     I had a similar problem recently on a System76 Pangolin laptop running
>     Mint 17.  It was not complaining about the fan sensor, it was just that
>     the fan was revving up too often and was much too loud.  I was very
>     pleasantly surprised to find out that on this laptop the fan is attached
>     to a "trap-door" panel held on by 3 screws and was insanely easy to pop
>     right off and clean.
>     So I first recommend taking a look at that and just blasting out all the
>     built up dust and re-seating the connector(s).
>     Some may disagree with this, but I also hit the fan motor/bearing with a
>     couple of good squirts of WD-40.  For me, that seems to flush out more
>     dust and loosen things up again.  I did it over the garbage can, and
>     blotted up as much as I could before reinstalling it, for hopefully
>     obvious reasons.  :-)
>     That all helped but did not fix the problem, so I started hunting.  It
>     turns out that somehow it was using the Nouveau driver which (I think)
>     was causing the graphics card not to idle.  I'm not sure if this laptop
>     was always using that driver or if it somehow switched, but I am
>     sure that:
>     a) switching from nouveau (1:1.0.10-1ubuntu-2) to nvidia-340
>     (340.96-0ubuntu0.14.04.1) solved the fan too often/too loud problem and
>     that
>     b) that fan problem has only been happening for a few months and that it
>     got much worse in the last week.
>     On 12/02/2015 07:36 AM, Rich Mingin (PLUG) wrote:
>     > So for some elaboration/reassurance, its likely the fan itself is
>     > mechanically fine, its the rpm sensing that is failing. This matches up
>     > with the rpm sawtooth pattern, the warning on startup, and clearing on a
>     > full power cycle. While there is a small possibility that it's the
>     > system hardware controller failing and not the fan/sensor/wire, I always
>     > blame the 5-20$ parts for failing before worrying about the 200-500$ ones.
>     >
>     > Let me know how it goes.
>     >
>     > On Dec 2, 2015 12:21 AM, "jeff" < <> <
>     <>>> wrote:
>     >
>     >     I'm good, thanks!
>     >     Had to replace the kbd already, which was surprisingly easy. Just
>     >     hoping it's only the fan.
>     >
>     >
>     >     On 12/02/2015 12:13 AM, Rich Mingin (PLUG) wrote:
>     >
>     >         It's a simple enough repair, I've done it on a few different models,
>     >         15-20 minutes if you take your time and do it right. Need
>     >         assistance or
>     >         support?
>     >
>     >         On Dec 2, 2015 00:01, "jeff" < <> <
>     <>>
>     >         < <> <
>     <>>>> wrote:
>     >
>     >             My laptop's fan started making revving noises yesterday.
>     >         Today when
>     >             booted, I got a BIOS warning about not being able to detect
>     >         the fan.
>     >             Rebooting fixed it.
>     >
>     >             I'm looking for somebody to tell me it's simply a
>     >         replacement fan fix.
>     >             The fan is attached to a long cooling heatsink, complete with a
>     >             copper tube. The fan is sold by itself or with the entire
>     >         assembly.
>     >
>     >             HP Elitebook 8760w running Xubuntu (for topicality).
>     >
>     >
>     >
>     >
>     >
>     >             Thanks for the benefit of your experience.

--  -------------------------------------------------------------------
JP Vossen, CISSP | |
Philadelphia Linux Users Group         --
Announcements -
General Discussion  --