How long can I expect a case fan to last?

Denis54 · Post by **Denis54** » Thu Jan 12, 2006 2:56 pm

I have a Scientific Atlanta 8300 HD PVR with a 160 GB HD. As it is located in my TV stand, there is not much room around it and it gets pretty hot when in operation. If I put a themometer on top, it reads 42 C.

I have taken an old 80 mm case fan and installed it in my TV stand next to the PVR. Within a few minutes the thermometer shows that the PVR case is now at room temperature.

If I run the fan 24 hours a day, how long can I expect the fan to last?

Would a high quality fan last much longer?

Would the fan last longer if I power it at 7 V.

What happens when the fan dies? Is there any danger of a short circuit?

Tibors · Post by **Tibors** » Thu Jan 12, 2006 6:24 pm

Fan life is usually defined with an L10 number (in hours). This means if you take a whole load of the same fans and run them all continuously under the same conditions then after that number of hours 10% of the fans have died or gone otherwise out of spec.
* Papst 80x25mm fans have an L10 of 80000 hours at 40°C (~9.1 years)
* Panaflo 80x25mm L1A has (IIRC) an L10 of 50000 hours at ???°C (~5.7 years)
* Cheap fans have a shorter lifespan, but most often the manufacturer won't tell us (the consumers) what it is.

L10 is always for running the fan at its normal rated voltage. Running a fan slower will extend its lifespan. Running a fan in a hotter environment will reduce its lifespan.

Bean · Post by **Bean** » Thu Jan 12, 2006 7:14 pm

awesome. I wonder if having a fan on the horizontal shortens its life?

vine-au · Post by **vine-au** » Thu Jan 12, 2006 7:52 pm

Bean wrote:awesome. I wonder if having a fan on the horizontal shortens its life?

Why would that make a difference? the bearings in a fan have pressure on them regardless of the orientation.

Felger Carbon · Post by **Felger Carbon** » Thu Jan 12, 2006 9:31 pm

vine-au wrote:
Bean wrote:awesome. I wonder if having a fan on the horizontal shortens its life?
Why would that make a difference? the bearings in a fan have pressure on them regardless of the orientation.

When I tried to run my 1500RPM Nexus 92mm fan pointed upward, the RPM dropped by ~400RPM! Serious bearing problem.

I had modded a Freezer 64 to use the Nexus. When the mini-tower case was positioned upright, the Nexus was blowing upward... spectacularly poorly. Dunno if it was just the one fan or a generic problem.

jaganath · Post by **jaganath** » Fri Jan 13, 2006 3:46 am

Why would that make a difference? the bearings in a fan have pressure on them regardless of the orientation.

When my YL D12SL-12 are vertical they make an intermittent "shhh....shhh" noise, as if the hub is dragging on something, when horizontal they do not make this noise. If you think about it, when a fan is mounted vertically it's weight is primarily acting through the plane of the hub, whereas when they are mounted horizontally it acts perpendicular to the hub.

F4lc0N-76 · Post by **F4lc0N-76** » Fri Jan 13, 2006 4:21 am

As far as my knowledge of bearings goes, the horizontal position for an ball-bearing fan would be the worst case scenario, as I assume the ball-bearing in a PC fan always is of radial type. As for the non ball-bearing fans, some kind of gliderbearing made of plastic, the best position for it depends on wether the bearing is radial or axial.

Radials works at their best in vertical position, and axial bearings in the horizontal position.

As I never had the opportunity to dissicate a PC fan, I don't know if the bearings are of radial or axial type

btw, thanks for a great site!!!! I enjoy reading it, in my fight for a quieter gaming PC

and please do have some patience with my english - it's just one of 4-5 languages I use on weekly basis

Aleksi · Post by **Aleksi** » Fri Jan 13, 2006 4:59 am

Running a sleeve bearing horizontally will wear the bearing a bit faster. Papst specifies different L10 specs for their Sintec sleeve (it's called sleeve, but it's really a passive hydrodynamic) bearing when it is run horizontally / vertically.

Many generic sleeve bearing fans have L10 figures of ~20k-25k hours at 25C (for example the Nexus 92mm made by Dynaneon has L10 of 20k hours at 25C IIRC). The lifetime drops naturally when the temperature rises. For some ADDA sleeve bearings the L10 at 40C is a bit over 10k hours IIRC. The exact figure varies between different fans.

In general generic sleeve bearings aren't a long lasting solution at high temps. If they were, then companies like IBM, Dell, HP, Agilent would use generic sleeve bearing fans instead of Panaflos and NMBs because it would save them money.

Proper bearings like Panaflo FDB (Yes specifically Panaflo FDB. For example the ADDA fluid dynamic bearing really isn't a proper one, but technically they can call it FDB), Papst Sintec / BB or NMB ball bearing are good choices for people who want quiet fans for higher temps.

Even at 70C the L10 figures are years for the above fans. Basicly Papst Sintec and Panaflo fluid dynamic bearing have the longest lifetime for non-ball bearing fans (40C Panaflo FDB L10 = 6yrs by spec, 9yrs by field data. Don't remember the Sintec L10, tough I have it at home)

jaganath · Post by **jaganath** » Fri Jan 13, 2006 6:52 am

As I never had the opportunity to dissicate a PC fan,

I think you mean 'dissect'. "Dessicate" means 'to dry out', ie "dessicated coconut" is dried coconut. I hope you'll forgive my language fascism.

BTW, thanks for the info on axial vs. radial bearings. Very interesting....I might "dissect" one of my case fans later on to find out....

it's just one of 4-5 languages I use on weekly basis

Cool! What are the other 3 languages?

Many generic sleeve bearing fans have L10 figures of ~20k-25k hours at 25C (for example the Nexus 92mm made by Dynaneon has L10 of 20k hours at 25C IIRC). The lifetime drops naturally when the temperature rises.

Why is this? I assume the lubricant evaporates more quickly?

Proper bearings like Panaflo FDB (Yes specifically Panaflo FDB. For example the ADDA fluid dynamic bearing really isn't a proper one, but technically they can call it FDB)

Aleksi, I know it's a lot to ask, but could you give a brief (brief!) list of all the different bearings used in PC case fans and a short explanation of how they work? ie ball bearing, sleeve, passive hydrodynamic, active hydrodynamic (?), FDB, ceramic etc. I'd be ever so grateful

Aris · Post by **Aris** » Fri Jan 13, 2006 9:03 am

jaganath wrote:Aleksi, I know it's a lot to ask, but could you give a brief (brief!) list of all the different bearings used in PC case fans and a short explanation of how they work? ie ball bearing, sleeve, passive hydrodynamic, active hydrodynamic (?), FDB, ceramic etc. I'd be ever so grateful

you know, if you really want a detailed answer, Dorothy Bradbury is the one to ask. her SPCR account name is Jafb2000

Aleksi · Post by **Aleksi** » Fri Jan 13, 2006 9:28 am

I agree with Aris, Dorothy will give you more information than you will ever want to know. The answer isn't as simple as many would think, so it really can't be brief

. She knows more different manufacturers and their bearings than I, simply due to her years of experience. She can give better comparisons than I.

Just an example why the thing isn't so simple:

- Ball bearings have differences between manufacturers and within manufacturer (different grades)
- sleeve bearings have differences (compare a generic sleeve bearing and a Papst Sintec sleeve bearing (huge amount of oil compared to a generic). Coincidentally the Papst is also a "true" passive hydrodynamic bearing design.
- compare fluid dynamic bearings (Panaflo FDB and ADDA FDB. The ADDA can technically be called FDB, but it doesn't have the real benefits of a Panaflo FDB, as the ADDA version is a sort of an modified sleeve bearing. And the Panaflo FDB is a "true" active hydrodynamic bearing.

yadda yadda yadda. Get the picture?

You can simplify it to a few categories of bearings, but it's important to explain why A is different from B to make it accurate.

Rory Buszka · Post by **Rory Buszka** » Fri Jan 13, 2006 10:10 am

Don't forget sunon's Vapo bearing with the Magnetic Levitation system. Ideally, the journal never touches the casing. The only point of contact, theoretically, is the very small cap at the end of the rotor shaft.