High fan controller failure rate or are people stupid?

Like everything else, controllers are designed to run a range of fans. Some fans can draw a couple of amps or more yet not all controllers are designed for this heavy a load. Overload the controller and yeah stuff can burn.

well... but one would think at least ONE controller would be able to handle whatever fans you throw at it. yet even the ones that get the best ratings (scythe kaze master/server) have plenty of people mentioning that not only did they start to fail but that the manufacturing quality itself is horrible.

at least those with displays... i think i've seen some without displays that cost just as much but have no bad reviews (and which coincidentally - or not - have heat sinks whereas at least the scythe ones don't...). no RPM display for me then -_-

See, the thing is I=V/R, so basicly, you can burn anything, as long as you have a powerfull enough electrical source. And as most computers are attached to the power grid, which may be considered a powerfull source, when the line is shortened /or R is low enough and a more powerfull fan has lower R, when V is fixed/ it turns to a race - what will burn first - your PSU, your mobo /if the line goes through it/ or your controller.
And the fact that a fan is a simple thing just makes it a better tool for the job  

well but apparently, manufacturers significantly skew the specifications as it seems.

e.g. this fan control is supposed to be able to handle 30W per lane. yet it already heated up with at rather low load (if he measured correctly because it seems weird that it would heat up less when the load is higher...): http://martinsliquidlab.org/2011/04/03/sunbeam-rheosmar...

and there was one guy who described he put four scythe ultra kaze onto a fan control that was supposed to be able to handle 30 watts too... well... started burning... (unfortunately, i can't find that review any more)

sounds as if being able to handle "30 watts per lane" really means "30 watts total".

It just doesn't actually work that way. What makes a thing to burn is not the power of the device, attached to it, but the ammount of the electrical current going through it /so sometime the results may seem kind of strange, from an every day logic point of view/.
With several devices, connected in parallel, the formula is I=V(1/R1+...+1/Rn), so the resulting resistence of the cirquit drops when adding consumators in parallel. It's not how many watts they are in summary, but what will be the resulting current going through a specific point of the system. In practice the things are even more complicated, as even if the attached in parallel devices have same specifications, they are never perfectly the same, and if the source is powerfull enough the current may /over/run through the weakest device. It may result in a cascade failure.

so people are frying these things likely because they attach too few consumators?

as for the resulting current being more important - unfortunately, manufacturers don't provide specs on that.
and they actually calculate it the way i did themselves: for instance from scythe kaze master: "Maximum Fan Ampere per Channel 1 Ampere (= 12 W max.)"
this of course doesn't make sense because if it scales inversely proportional, that would mean one would have to use fans with at least 1 ampere to not get above 12W. but of course the maximum current is 1 ampere. so according to these specs, the only safe way to operate this would be with with a 1 ampere fan on each channel?

A couple AMPS or more?...one of the strongest fans I've seen is a server grade 1.3amp fan that sounded louder than a jet engine and blew air faster then that hurricane Irene.

I can't imagine a fan a couple amps...must be some powerful stuff.

Well remember, while a fan controller is a very simple thing, people expect them to be very cheap and the manufacturers still want to pull a profit. So I imagine corners were cut and quality is poor, leading to wildly variable performance and a lot of failures.

sounds reasonable?  



uhm yeah... seems like profit with a capital P. because if something is very simple, it is usually very cheap to produce. just compare fan controls to other pc hardware that costs roughly as much take e.g. a $30 fan control and a $50 hard drive. you can bet that there is a much lower profit margin on the hard drive...

yeah... since the ones with temperature sensors are a bit expensive just for temperature measurement (which i would've loved) and the zalman fan mate (which seems the only really reliable device) doesn't allow for fans to be switched off completely, i'll just stick to the traditional way - hooking up all the fans to the motherboard and controlling them with speedfan.
i just hope i'll be able to switch off a PWM fan that i have, since it says 900-1300RPM.

Best answer selected by sh4dow83.