Compressed air and cleaning fans

Using an air compressor, I was cleaning out the dust from a consignment of used ATX power supplies, when one of my friends pointed out that I was killing the fan by spinning the fan backwards. What he does it use a stick or toothpick to prevent the fan from rotating. This seemed odd to me as I've been using compressed air to clean fans by spinning the rotor in both directions and usually faster than rated RPM for maybe 25+ years without incident. Google found some articles proclaiming that spinning the rotor with compressed air was a bad idea. However, I couldn't find anything describing any failure modes. So, is using compressed air to clean fans really a bad idea? Is it bad to spin a cooling fan backwards? What breaks or fails?

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558
Reply to
Jeff Liebermann
Loading thread data ...

Jeff Liebermann has brought this to us :

Well I was told that so far back that I can't remember (1960s) I think, Long before PCs, mainframes had lots of fans. Although I never saw a failure I always thought it was because you could grossly overspeed the fan if you were stupid enough to put too much air in. B-)

--
John G Sydney.
Reply to
John G

Spinning the rotor causes induction voltage, which can be the wrong polarity, or to high, or both, so internal electronics may be killed.

Reply to
Sjouke Burry

I can see that spining the fan backwards could be bad if the fan had a brush type motor and the brushes hit the rotor segments going backwards.

Reply to
hrhofmann

What Sjouke said is probably true but I believe there is Zener diode protec tion all over the place. If so, would it be a good idea to unplug the motor for this ? Maybe not. Let the Zeners conduct for a half a minute, should d o no harm. On the other hand, disconneecting it might allow a highr voltage to be developed which could cause a breakdown of insulation in the wiring.

the other concern is the bearings. They are not quite Timkens. they are pro bably just brass wjhich means there is a functional limit to RPMs. It's a c hemical metalurgical thing and I would have to ask my Father. Well, to do t hat we need to get some Gypsies and have a seance. (Timken bearing are the type used on non- drive car wheel's shafts. They are tapered and tightenabl e down to zero play and thrust, they are also used on better lathes and oth er machines)

Anyway, the regular brass bearings almost cannot be called bearings, they a re more apty called busings and they wear with every revolution. What's mor e, the even minor friction in that system exacerbates the wear at at least an exponential rate because of the heat produced. you won;t be aware of it because it is localized right there. what can happen is that the life of th e fan is greatly shortened. not that it will fail. but it will become noisy .

I never worried too much about noise until the fan in my laptop got noisy. you know the microphone picks that up ? I do now.

I would say if you can use two hands, use one for the air chuck and the oth er to limit the speed of the fan.

I was going to mention the 327 the olman built a long time ago that always wiped number one bearing at exactly 9,200 RPM. But that is a differenrt pro blem, they couldn't get enough oil to it. they did almost everytihng, gradu ated bearing clearances, high volume, cut the flow to the lifters and cam, and still it wiped out at about the same RPM every time. the problem was th e mass of the oil trying to make it through the crank with the centrifugal forces acton on it. The ONLY way they could have reallt fixed it was to use two oil pumps, or a dual feed system. That was too much work.

So they got the crank machined a few thousandths under and had it hard chro med back to size. this way, the crank ws safe and all they had to do was to replace the bearings, and in some cases the connecting rod. And no, that d id not unbalance the engine in any signifacant way.

But back to the topic, I THINK the major danger is to the bearings. You're an experimenter, experiment. (and of course post up the results)

Find a way to measure the RPMs when you blow that air through there.

Oh f*ck finding a way, you already know. It has a tach pulse. Just get the rated RPMs for a given fan, find what it is putting out, it is most likely one pulse per revolution. dunno, never had to check. But then blow that air and see what you get on the freak counter.

Hell maybe I'll do that just for fun. Just to learn something that day.

You know what you do shooting that air in ther ? YOu make a turbine engine. It is not quite as good as the stor boughten ones, but that is what it is.

Thanks for raising a new, interesting point that nobody ever thought of.

Reply to
jurb6006

** Computer fans have no brushes.

Brush motors ( both universal and permanent magnet types ) are reversible.

... Phil

Reply to
Phil Allison

HEY, how about just using a really powerful vacuum cleaner instead ?

However, if the RPMs are exceeded... But then, it is probably better for the bearings to keep always going the same way. I am SURE tha tis true of the brushes.

Maybe it doens't make that much difference. Another thing, compressed air has no limit on force, vacuum is automatically limited.

Maybe be better ?

Reply to
jurb6006

My thanks to everyone for considering this problem.

I just spun a common 80 mm fan for about 3 minutes backwards using 80 psi air pressure. I could only run it 3 minutes because I was making enough noise to have the neighboring office complain. I have no idea of what RPM it was spinning, but I'm certain it was well above rated RPM. Applying 12V afterwards to the wires demonstrated that it spun normally and that nothing had failed. Incidentally, putting a VOM across the leads, and spinning the fan in either direction produced no detectable voltage.

I would think that any manufacturer of brushless fans that uses internal electronics as a driver, which blows up if the fan is run as a generator either too fast or backwards, is just asking for problems. I'm fairly sure that the driver IC's are all well protected. That may not have always been the case, where such fans were driven by discrete xsistor circuits or the fans had brushes, but todays fans seem impervious to such abuse.

The view of the inside: There should be a schematic in there somewhere.... yep: Well, it's 3 poles instead of the usual 4 poles, but the principles are the same.

Note the totem pole output of each driver transistor pair. Without any applied power, all the transistors are normally open no matter what the motor is doing. The only way that any generated voltage from the motor/gen is going to go anywhere past those transistors is if at least two of them are turn on by some magical source of outside voltage. So, maybe I can kill this fan by applying a low voltage to the leads, making the fan spin slowly, and then using air pressure to spin it in either direction. Back in a few minutes while I try that...

No damage. I tried 3.3v, 5v, and 12v applied, while I did the air hose trick. With 12v applied, it made some rather impressive noises thanks to the overspeed, but no damage. I had a difficult time stopping the fan with 12v applied, but managed to do it for a few seconds. Lesser voltages produced similar overspeed or stall, but to lesser degrees. The transistors were certainly turning on when the power was applied and the fan was spinning by itself, so any voltage produced by the spinning magnets should have made it past the totem pole switches, and into the osc/driver circuitry. So much for that theory.

The bearing/bushing burnout theory is also interesting. This particular fan, Mechatronics Model F8025H12B2 12v 0.145A claims to have ball bearings inside. I'm not going to take it apart because I know that it doesn't have ball bearings. It uses bushings like all such cheap PS fans. The only fans I've ever seen with real ball bearing are the larger and more expensive fans by Rotron which burn

20-40 watts instead of 2 watts. At the higher power level, any imbalance is going to put a nice gouge or dent into a bushing, while a ball bearing arrangement will distribute the vibration loading and not be affected.

About the best I can guess(tm) is that fan overspeed will cause the lubricant to get hotter than usual, which might blow the seal, leak out, and produce premature failure. I've never seen that since most such fans don't need my help with drying out the lube or leaking oil. Still, the few seconds it takes to blow the dust off the fan is not going produce any significant overheating and certainly not long enough to pump out the oil.

Summary: Spinning the cooling fan rotor may have been a bad idea in the distant past, but todays fans can easily take such abuse.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558
Reply to
Jeff Liebermann

Methinks you have hit upon the answer. Overspeed and reverse rotation may have been a problem for the fans of the 1960's, but todays brushless computah fans are largely immune to such abuse. Running a commutator type motor backwards will probably do some damage to the commutator because the brushes are inserted at a slight angle (to help clear out debris between the brushes and commutator sections). However, none of the computah fan motors I'm currently dealing with have commutators or brushes.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558
Reply to
Jeff Liebermann

For Pc fans, it's harmless to clean them with compressed air, unless you're trying to overspeed them by angling the air stream just right, turning the fans into turbines.

I did once basically explode a squirrel cage blower with the pressure side of a vacuum cleaner. It ran so fast that the fins expanded and touched the housing and became a bent up mess before flying out of the housing across the room. I was not expecting that.

Reply to
Cydrome Leader

Most ollder larger computers used ac fans. I've blown out a few fans, no problems seen.

Greg

Reply to
gregz

I'll bet this is a carry over from the days when running a brush motor backwards would degrade/ruin the brushes. I know; I ruined one that way.

Reply to
RobertMacy

** FFS - wot an utter ignoramous.

Peal of the paper label and take a look sometime.

Brushless DC fans mostly use ball bearings.

... Phil

Reply to
Phil Allison

Mythbusters, anyone?

To the best of my knowledge, there is nothing in any motor that //mechanically// favors rotation in one direction. "Forwards" and "backwards" are determined electromagnetically, not mechanically.

The maximum speed before the bearings are damaged depends on the bearing quality, I assume. The only damage I can image is knocking the shaft off-kilter.

Reply to
William Sommerwerck

"NO ONE expects a squirrel-cage blower explosion..."

Reply to
William Sommerwerck

I've taken quite a few power supply and CPU fans apart in order to clean off the ring of tar (formerly oil) that accumulates on the rotor shaft and re-lubricate the fan. Usually, the shaft fits into what looks like a sintered bronze sleeve. I don't recall seeing ball bearings on the smaller (80mm or less) but may have been looking at fans with both a sleeve and a ball bearing (1B1S), as described below at bottom of page:

Anatomy of Computer Fans

Ball vs. Sleeve: A Comparison In Bearing Performance

Fan Design (review) including info on bearing types:

Let me see if I can dream up some other ways to destroy this fan with compressed air before I dissect it. Stay tuned.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558
Reply to
Jeff Liebermann

Sounds like fun. Let's run some fans at Mach 2 in a wind tunnel and watch them fly apart. (Anything worth doing is worth overdoing).

Oh-oh. I think you may have hit on a potential problem. As long as the pressure is equal on all the rotor blades, the rotor shaft remains centered in the middle of the bushing. However, with an air hose, one could apply all the pressure on one side of the frame, producing an asymmetrical side load on the rotor shaft. In other words, I'm producing a bearing load condition for which it was not designed. If the rotor shaft were dry and the lube gone, the high speed rotation of the rotor shaft might put a gouge into the edge of the bushing. A ball bearing would be unaffected.

I've seen a variation of the above potential problem with bushing type fans, as in ATX type power supplies. The components inside the power supply are not uniformly distributed around the fan intake area. The result is that the pressure drop at different points around the fan is not uniform. This produces an asymmetrical fan load which results in a side load on the bushing. If lube were present and functioning, no damage would be done. However, if the rotor shaft were mounted vertically, and all the oil drips to one end of the shaft, the potential for damage is quite real.

I guess the proper cleaning technique with bushing type fans is to move the air hose nozzle rapidly around the circumference of the duct, thus preventing bushing wear at any one spot.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558
Reply to
Jeff Liebermann

I've seen (very cheap) brush-type motors that had asymmetric brush springs, that would possibly buckle if the motor was reversed.

Reply to
whit3rd

I took apart the fan and much to my amazement, found two ball bearings. I just hate it when I'm wrong. I have some more junk fans somewhere. I'll see what I can find. In the past, they were mostly bushings, but I guess that's changing, especially since this fan was fairly new.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558
Reply to
Jeff Liebermann

Just one of those things you have to learn the hard way.

Reply to
Cydrome Leader

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.