Scythe Kaze master Mofsets? reaching 91+C with 2 fans attached within specs??

[rge]

I have 3 fan controllers and need to condense to 2 for space, so need to run 2 fans per controller on some.

According to scythe fan controller, specs are:
Maximum Fan Ampere per Channel 1 Ampere = 12 W max.

Fans I tried running 2 of are Scythe GT 1850 (<0.1 amps per fan) and Scythe S flex F (0.2 amps so 2 is 0.4 amps).

With 1 fan, temp on ?mofsets was 50's C when running fan full speed, 70C when running fan low speed.
With 2nd fan hooked up, temp was 70C when running full speed and 90-93C (SEE PIC) when running 2 fans at low speed.

Anyone have a guess if 90C within specs of that component? Tried it with more than one temp probe and all are calibrated accurate, and it will burn your finger quickly to touch it (which is why I measured I was testing rpm readouts with 2 fans hooked up an accidentally touched it). I guess if it dies no loss, going to replace soon anyways, just as long as it doesnt smoke and smell like electric fire like my video card did when gpu seals leaked, resulting in a long conversation with wife about watercooling...

[dejanh]

I know you are asking if 90C is within specs, but I have to say that the bigger issue is that MOSFETs running so hot with such a seemingly light load does not seem normal. It seems defective.

[rge]

That was my first guess, but so far I tested 2 of the 4 on that one controller and both ran ~same temps with 2 fans hooked up, and just testing 1 on my second scythe controller, it is up to 88C just after 2 minutes.

Maybe I need a fan to cool my fan controller, until I replace them

[Main]

No it's not defective - they run as hot as that

I will probably put on some ramsinks and use a 40 mm fan for better cooling since I don't feel comfortable with those temperatures.

[dejanh]

That was my first guess, but so far I tested 2 of the 4 on that one controller and both ran ~same temps with 2 fans hooked up, and just testing 1 on my second scythe controller, it is up to 88C just after 2 minutes.

Maybe I need a fan to cool my fan controller, until I replace them

No it's not defective - they run as hot as that

I will probably put on some ramsinks and use a 40 mm fan for better cooling since I don't feel comfortable with those temperatures.

That's pretty messed up

[eth0s]

90C seems awfully hot. What about a different fan controller: the Sunbeam Rheobus Extreme is good, and has six inputs, for possibly up to 12 fans. I personally use one Sunbeam for only six fans, to be safe though.

Second topic: Watercooling is the best option for efficient cooling, however, you still have fans. Maybe fewer fans than now, but maybe more, depending on your set-up.

[rge]

Actually does look like the 2 scythe F fans are about 10-13C worse than 2 gentle typhoons when tested on same channel, guess since F's have higher amp rating.

ethos: yep I am going to get a different controller. I will get the Lamptron FC-5 (has readouts which I like) if they get in stock soon, if not will get sunbeam...and I like the 6 controls on sunbeam. Or may get one of each of those...only cpu rad fans I am interested in seeing rpms readouts. I am just hoping for no electric smell burnouts until I get new ones.

Main: thanks for info

EDIT: holy crap, dont run 2 led fans on one channel, a few of my 15 fans are .3 amp rated led fans (x 2 still only .6 amps), temps were 103C-109C. I am thinking they should change the max to 0.1 amps on specs instead of 1 amp. Anything above .1 amps nets mofsets temps above 80C when fan controller rpm is turned to lower rpms.

First pic, cant see temp probe, it is measuring temp mofset on second scythe controller still attached to case, in second pic, pulled it out.

[rge]

No it's not defective - they run as hot as that

I will probably put on some ramsinks and use a 40 mm fan for better cooling since I don't feel comfortable with those temperatures.

what a difference heatsinks make...this is the sony regular controller not extreme (both it and extreme have similar large heatsinks on mofsets and this is without any fan), max temps are ~45C no matter where I measure, and this is with same 2 led fans at same ~rpms that was at 109C on mofsets on scythe that does not have heatsinks.

[NaeKuh]

YES the mosfets on the kaze is HOT as hell.

I actually burn't my finger trying to get a fan header back in once, and learned never to try that again.

The kaze master is SUCH CRAP CRAP CRAP!!!!

Thank god i have a lamptron on its way as a replacement.

[gmat]

MOSFETS can take this kind of heat, it's within their usual specs. Nothing wrong here, but still it's not far from the limit (which is around 120C usually if i remember)

[NKrader]

i have kaze server and i run 2 fans on one channel that are below specs and the mos gets really hot like burn my finger hot ... figured its fine lol..

EDIT : and i use the Triebwerk TK-122 fans

[Main]

YES the mosfets on the kaze is HOT as hell.

I actually burn't my finger trying to get a fan header back in once, and learned never to try that again.

The kaze master is SUCH CRAP CRAP CRAP!!!!

Thank god i have a lamptron on its way as a replacement.

crap my aZZ - Kazemaster Server is the only real fancontroller available, the others are just fanspeed adjusters ( that goes for the Kazemaster ACE for that matter)

Furthermore we have Kazemaster Server installed in very hot enviroments - standalone colocation servers - I have never measured the temperature of the mosfets - but I know they most be glowing hot.

[Main]

i have kaze server and i run 2 fans on one channel that are below specs and the mos gets really hot like burn my finger hot ... figured its fine lol..

EDIT : and i use the Triebwerk TK-122 fans

yeah you will be fine - but it doesn't hurt to take measures against it of course. I run 3x140mm LianLi fans on one channel - no problem.

If the kazemasters started to die we would have seen it by now - they have been in the market quite a while.

[Humminn55]

crap my aZZ - Kazemaster Server is the only real fancontroller available, the others are just fanspeed adjusters ( that goes for the Kazemaster ACE for that matter)

Thank goodness you're here. We really do need someone who absolutely knows everything about anything anyone posts about.

[Diverge]

It's probably a good idea to put the probe on the body of the part, and not the metal. The metal is the collector of the transistor, and it could mess up your readings. Most heatsinks that are mounted to transistors use thermal grease that is either non-conductive, or use a non-conductive insert between the metal tab and heatsink.

Here's the data sheet of the transistors used in my old scythe fan controller http://pdf1.alldatasheet.com/datashe...NICS/D882.html

Tj (die temp) is rated for up to 150C

[faster3200]

This might explain why my perfectly in spec fans managed to kill one of the channels on mine. They really need to revise these things.

Thank goodness you're here. We really do need someone who absolutely knows everything about anything anyone posts about.

[Fitseries3]

i use mine for 3k ultra kazes and deltas.

you just need a heatsink on that bad boy haha!

[damunk]

Yeah they are crap, had one myself and returned it within a day. With only a few fans hooked up it got so hot that it needed it own cooling, which is a bit retarded if you ask me.

I remember going to scythe forums to see if it was just me having the problem and I can remember the first page being filled with posts like: Kaze Master Burning?!

Paid a bit extra for a TMS200 and never regretted it.

[WaterFlex]

Just have checked my Kaze temps out. 40C

[rge]

Just have checked my Kaze temps out. 40C

Is that temp reading with fan speed down turned? If the controller is not reducing the speed significantly, ie fan on full blast or near full blast, the temps wont be bad, mine were ~50'sC for scythe F at or near full rpms. The temp increases rapidly and progressively as rpms are turned down, especially in 800-900 rpm range. But interesting if your temps are much better.

@fitseries3, nice job on heatsink, how did you attach with thermally conductive adhesive?

edit: @diverge, thanks for thermal specs. The black mold is roughly 5C cooler than metal (see pic), here metal was 110C, mold 105C. btw, all measurements are made with thermocouple coated with thermally conductive, electrically nonconductive thermal grease in which this thermocouple was calibrated for surface readings. Back in realtemp thread when milling holes in IHS for cpu IHS readings, and then reading how intel did it, then using calibration plates to check + true surface probe, thermocouples wont accurately read surface temps unless coated in thermal grease, since they will average air temp + surface temp, unless a true flat temp probe like wife had access to at her lab, which is also where got calibration plates. And even mx2 coating the thermocouple tip will read within about 1C of thermal grease, so works in a pinch.

[Diverge]

The reason they get hotter when the fan is speed is turned down is cause the voltage is dropped across the transistor. So if your fans are running at lets say 7V, the other 5V is pretty much being converted to heat by the transistor. When it is at fullspeed, the only drop across the transistor is Vcesat, which is the saturation - when the transistor is fully on.

[Elpy]

Sigh.. So I might as well throw away my kaze server even though I havent even used it yet... Gonna have 7 GT 1450rpm's so... Any suggestions for some good controller? Would prefer speed control based on temperature... So aquaero?

[Fitseries3]

i used 2 small zip ties to affix the HS to the mosfets. use whichever TIM you would like.

[Diverge]

Sigh.. So I might as well throw away my kaze server even though I havent even used it yet... Gonna have 7 GT 1450rpm's so... Any suggestions for some good controller? Would prefer speed control based on temperature... So aquaero?

GT's only draw a fraction of current that a normal fan does, you should be fine.

edit: they only draw 0.049A, typical 120mm fan draws like 0.3A. you could put them all on one channel and still only draw 0.343A... a little more then typical fan.

[Conumdrum]

GT's only draw a fraction of current that a normal fan does, you should be fine.

edit: they only draw 0.049A, typical 120mm fan draws like 0.3A. you could put them all on one channel and still only draw 0.343A... a little more then typical fan.

Thats what the specs GT says. But I don't believe it. Time to make a special cable and check amps myself.............

Grr my multimeter fuse is blown from another past experiment, more tomorrow once the fuse is replaced.........