And to mention that the PS cover helps guide the airflow thru the heatsinks and by leaving the cover off the fan isn't drawing the air through the designed path. You can leave the lid off the PC without a problem but the PS (as Toth states) is very dangerous and actually detrimental to the cooling process. A better fan, in terms of more CFM pulled, is the better way to go. I've pulled fans from old NCR microchannel machines that have variable speed, controlled by a thermistor, that speeds the fan up when it gets warmer and I know new ones are still being made. -----Original Message----- From: owner-classiccmp(a)classiccmp.org [mailto:owner-classiccmp@classiccmp.org]On Behalf Of Tothwolf Sent: Monday, April 29, 2002 3:38 PM To: classiccmp(a)classiccmp.org Subject: Re: IBM AT power supply case/cooling question On Mon, 29 Apr 2002, Chad Fernandez wrote: Leaving the cover off of the power supply would greatly increase the chance of contacting one of the high voltage components in the supply. Typically, there are voltages in excess of 300V in those types of switching supplies. It is often present on the chopper's heatsink(s), and at a potential to ground, since most of those supplies simply use a voltage doubler on the mains voltage. All that being said, I don't think leaving the cover off of the power supply will help the cooling situation much, and would expose anyone working on the system to the high voltages present in the supply. There is a possibility of EMF troubles, but there is also the possibility of RFI issues. -Toth

Ok, well after reading everyone's replies, I guess I'll reassemble it as produced by IBM. I didn't realize that some of those heatsinks were live. Looking at the machine, I don't think airflow would be an issue with this particular case and power supply. The side towards the drive bays is mostly open and the top is very close to the back of the case cover. However, I'm still reinstalling it because of the electrical hazards mentioned :-) What I may do is poke a hole in the top of the case, and mount a big fan. How well would a 24volt fan run on 12 volts? Since it's DC it's fine, right? It's just slower? Chad Fernandez Michigan, USA Tony Duell wrote:

Tony Duell wrote: I tried the fan on 12volts and it did seem to work fine. I was worried that maybe the fan would over heat if it were running continuosly. I tried slowing and AC motor down once, and it overheated very quickly :-) Yes, I have thw 24v fan already. Unfortunetly, it won't fit the space..... it's too big :-( Chad Fernandez Michigan, USA

I'd suggest you consider an AC fan with a series capacitor in one side or the other of the supply to the fan. If you're ambitious, you can do quite a little with that sort of hookup, trading off fan noise for effective cooling power. If you put about 0.47uF in series with a typical small AC fan, it will cut the torque considerably, and that will cut the amount of air it moves. If you put a normally-open SSR, controlled by a temperature sensor in some critical area of your enclosure, in parallel with the cap, then have the control enable the SSR whenever the temp is above a certain point, it will speed up the fan and keep running at high speed until the control signal goes away. You then play with the hysteresis until you get satisfactory performance. Dick ----- Original Message ----- From: <philip(a)awale.qc.ca> To: <classiccmp(a)classiccmp.org> Sent: Tuesday, April 30, 2002 4:13 PM Subject: Re: IBM AT power supply case/cooling question

Back in the '70's, I hooked up a series capacitor on a relatively small (about 3-5/8" or so across the diagonal mounting holes) fan that made entirely too much noise. It spun up sluggishly but never failed to do so. I used a 0.47 microfarad non-polar cap with a 600-volt rating, not so much because that was required, but because it was "at least plenty" of headroom. I also had a 180-volt MOV across the same AC line. It worked just fine for a decade and a half. I don't know whether you'll want the MOV, but the cap will slow the fan, though you may want to try different values if that's an option. As far as managing the noise, I've found that you can make nice shock mounts using SILASTIC (The kind used for fixing aquaria comes in tubes the right size that have removable and screw-on tops. Use whatever color fits your fancy.). If you make a shoulder washer to sit between the fan and the case, (A pair would fit kind of like this: =(|=|)= ) where |) is the washer, |=| is the bulkhead, and = is the screw shank) such that the screw won't touch either the fan or the screw, and the fan is isolated from the screw, you'll have considerably less noise. I like to apply silastic and let it "skin" over before putting it in place. Then I tighten it into place "finger" tight and let it sit for half a day. After that, it'll stay in its positon after a half-turn with a screwdriver/wrench. If it's possible, you might want to let gravity do the work for you, i.e. put the fan in place without the screw, then let gravity level it for you. You've got to wait until it's thoroughly skinned over, but not cured, however, if you want to capitalize on this. It puts minimal stress on the fan's frame, hence, minimal stress on the bearing/bushing that isolates the blade from the frame. That way it moves with maximal freedom. Dick ----- Original Message ----- From: "Chad Fernandez" <fernande(a)internet1.net> To: <classiccmp(a)classiccmp.org> Sent: Wednesday, May 01, 2002 9:42 AM Subject: Re: IBM AT power supply case/cooling question the cooling will