jallen@uxrd14.UUCP (Jon Allen) (05/01/89)
Maybe the reason for the two fans is a backup. Since the reliability
of the DC fans is reported to be low, maybe they decided to put two in
to save the machine from a meltdown in case one fails. :-)
Jon Allen
ARPA: jallen@acpy01.att.com
UUCP: {backbone!}!att!acpy01!jallenwtm@neoucom.UUCP (Bill Mayhew) (05/02/89)
The fan on the left side (the side where the floppy disk is) doesn't seem like all that swift of an idea. If anything, it assists in sucking dust in through the floppy dirve slot. If the fan on the right poops out, the power supply is probably going to get too hot anyway, as the left fan isn't going to pull a whole lot of air from the right rear of the machine; it'll favor the floppy slot. Threre is thermal protection on the power supply in *my* 3b1, but I suppose that isn't the case for all machines. Now, mind you, repeated short on/off cycles, isn't going to do a whole lot of good for the integrity of the file system as the power supply repeatedly cools and heats. It could also be that the thermal protection doesn't invoke soon enough to prevent deformation of the case due to the heat build-up. I've thought about either using a thermistor probe or an air flow switch and an external controller to cut power when things get out of hand. A termistor would be fairly simple to implent. The termistor could be placed in series with a latch-up relay that would prevent power cycling. It might be possible to select a thermistor and relay that wouldn't need any transistors, etc to drive the relay. If I get anywhere with the project, I'll post the results. I don't think there is enough of a static pressure drop in the Unix PC case to make using an *inexpensive* commercially available pressure sensor practical; one might be able to use a vane air flow switch mounted outside the case with the fan blowing on it. Bill wtm@impulse.UUCP
dca@toylnd.UUCP (David C. Albrecht) (05/10/89)
In article <1611@neoucom.UUCP>, wtm@neoucom.UUCP (Bill Mayhew) writes: > > I've thought about either using a thermistor probe or an air flow > switch and an external controller to cut power when things get out > of hand. A termistor would be fairly simple to implent. The > termistor could be placed in series with a latch-up relay that > would prevent power cycling. It might be possible to select a > thermistor and relay that wouldn't need any transistors, etc to > drive the relay. If I get anywhere with the project, I'll post the > results. I don't think there is enough of a static pressure drop > in the Unix PC case to make using an *inexpensive* commercially > available pressure sensor practical; one might be able to use a > vane air flow switch mounted outside the case with the fan blowing > on it. > I considered replacing the DC fan in my 3b1 with an AC one until I did a conversion job on my girlfriends 7300 to full height and at the same time did a refit of the two DC fans two one AC fan for additional 12v capability. The AC fan was not only louder it also had louder vibration modes in the case which I wasn't really able to figure out how to completely damp. I'm not too keen on loud fans to begin with, buzzing cases are right out. I figured if I really want to protect my 3b1 from melting down what might do the trick is a thermal fuse in line with the AC supply (Radio Shack has a 139 degree C and 226 degree C model). Now, 139 degrees C is pretty hot but if I can mount it somewhere appropriate it would probably do the trick. The only other problem other than choice of mounting location is how to attach to it since it comes with wire leads and it might be difficult to solder to without blowing it. However it gets done, though, its likely it will require case disassembly and a little replacement work should it blow. Since this should be a fail safe device an awkward to replace component doesn't seem to be as bad an idea as it might be otherwise. David Albrecht