fanstorm controller

tm · Post by tm » Fri Mar 07, 2003 2:38 am

I recently picked up a fanstorm controller and have spent the past two days messing around with it. A few observations:

1. If you value tidy cabling, this will drive you nuts. On the other hand, it's nice that the cables on probes are long enough to reach right across the case.
2. I can basically say that this does not work very well with the Panaflo L1A fan. The controller does have this "Posi-start" feature, but it's not enough to get the fans spinning, so the lights will be on, but the fans won't be. I could get the fans up and running if I spent some time jiggling the corresponding switch or waiting until the controller supplied the full 12 V to the fan. I'm willing to say that this is probably not entirely fanstorm's fault, as I doubt that they had a fan as slow as the L1A in mind. I tested with another fan I had lying around and that started up ok. Once you got the fans going, then they worked fine, although all of my L1As exhibited the PWM growl.
3. My "deluxe" bay mounting came without the nylon supports to which you can mount the fanstorm board. But that's not a big deal, since the ribbon cable that connects the faceplate with the board is long enough, so I shoved the board into an open PCI slot. If you try to slide the whole assembly (board mounted in the bay mounting, with all the cables) right below a floppy drive, it's a real pain.

The current configuration has the fanstorm controlling 5 fans. The exhaust fans (one on top, one in back) are on one header, and the temperature probe is between the CPU and the PSU. There are two fans in the front each controlled by one probe on each hard drive. And then the video card fan is the last one. In theory, the fanstorm is capable of handling 16 fans total (four of each of the four headers).

Overall, the concept of closed loop fan control is great, but I don't think the fanstorm is well-suited for our needs, given that the L1As won't start with the thing. Now I'm very interested in getting my hands on one of fancontrol's homemade fan controllers

My game box:

Athlon 2100+ (SLK-800a), Epox EP-8RDA, 512 MB RAM, two loud 7200 RPM maxtor disks, one loud lite-on CD-RW, 3com NIC, Radeon 9700 (+Zalman HP-80a +Zalman fan bracket), seasonic 350 W P/S (modded with an L1A), Lian LI mid-tower case, 6x Panaflo L1A

Bxwrapper · Post by **Bxwrapper** » Fri Mar 07, 2003 5:33 am

I can see why you need 8 RPM monitoring things on the other thread, That's a helluva lot of fans in one rig...

tm · Post by tm » Fri Mar 07, 2003 4:19 pm

I probably don't ever need all 8, but I'll retrofit the L1As I have to use them

An update: More observation indicates that even if the L1A starts up off the fanstorm, it can just stop again (again with the fanstorm indicating that it's supplying power to the fan). And the noise generated by getting a fan going is really loud on the L1A, on par with the seek noise from my maxtor hard drives. I hooked up a generic Adda fan to the fanstorm and it was actually quieter, exhibiting none of the "growling" and having no problems automatically starting up.

Has anyone used the L1A with a fan controller that used PWM?

Steerpike · Post by **Steerpike** » Mon Mar 10, 2003 8:25 am

Wow! I'm not the only one with a fanstorm unit!

I bought mine back in November, but didn't hook it up until last week. Very disappointed. First, as you say, it won't start the panaflo low flows. And if you really crank up the heat to force it to run, it's very noisy - pwm growl, as you appropriately call it.

So I hooked it up to a generic fan, and noticed the 'growl' right away - almost like the sound of a loose part rattling around inside the fan, intermittent, annoying as hell!

I then hooked it up to my Zalman 92cm CPU fan (the one that comes with the zalman flower cpu heatsink). Same deal, works but not real quiet.

I then re-connected everything to my good-old standards - zalman fan-mates, and ... whisper quiet.

So while the fanstorm (and probably all pwm controllers) read great on paper, they seem ill-suited to actual fan control!

Pity; I was very optimistic.

The other problem with the fanstorm is the lights. The manual claims the light intensity reflects sensor heat. Well, maybe it does but my eyes had a hard time noticing the difference. I dare say I can tell the difference between full on and barely on, but you can't really tell much else!

So the search continues. I will probably rip out the fanstorm, and put back my digital Doc 5. I like the digital readout of temp, and precise temperature setpoint control. What I don't like about the digital doc is that the fans are either ON or OFF, plus, there is no 'hysteresis' in the controller - the fan will switch on and off repeatedly when the temp is around the setpoint.

I will try hooking the output of the digital Doc 5 to some zalman fanmates, and locate the temp sensors well away from the fan itself, hopefully introducing some hysteresis.

My strategy will be: CPU fan run at lowest setting permanently via zalman fanmate. Extra fans, not normally running, will be triggered by thermal sensors attached to the CPU heatsink and other locations. These fans will be off most of the time but if I stress my system, or it's a hot day, the extra fans will kick in. Having a zalman flower heatsink probably helps a lot with this approach, since even a case fan at the back of the unit will help it cool quite a bit.