I have a few IBM 36LZX drives (SCSI, 10K RPM 9.1 Gb, M/N:DDYS-T09170) in a 36in tall, steel Micron tower. How do I make them quieter? Most products say "not tested on drives over 7200 rpm". My thoughts are this: Isolate HD cage from frame with gromets (perhaps E-A-R). Isolate drives from HD Cage (suspension type?). Wrap cage with Foam (what type, who knows). Figure out how to keep them cool (how?). Finally -- BB or Dynamat the steel box, which is acting like a steel drum. Anyone with experience on this?

Decoupling the hhd's & cage and lining the cage + case will certainly help. You'll cut out most if not all the vibrational noise and some of the high pitched whine & seek noise. Some high freq. noise will still leak through your vents though.

I looked at the LZX review at http://www.storagereview.com and saw that you have one of the hottest drives around. It might be risky, but you could try encasing your drives in rubber like I did here: http://www.silentpcreview.com/goto.php?t=s&id=67&a=1. You will need to have some active cooling on it.

I saw your article. I still feel that I can't fully enclose the drive with out some sort of active cooling (50C max with 8.5 watts at idle). I looked at the sound specs and they should be quiter than they seem -- 3.9 to 5.0 bels. (can anyone tell me how to conver bels to dbA?) Perhaps my "pulsing" noise is from something else.
BTW- the drives were 30 bucks from TigerDirect, if anyone is interested.

it's very unlikely that teh sound you described is from a single source, a 'pulsing' noise usually means two motors of similar spec are slightly out of synch, so that for one the waves are coming at a slightly higher frequency. This causes positive and negative interferance is cause alternately (positive interferance is noise that combines to become louder, negative is two waveforms that cancel each other out) as the waves from one 'overtake' those from another.

this is usually caused by a pressure imbalance within the case so the intake fans have to try harder than the exhaust ones or vice versa (differences in static pressure mean same spec fans spin at different speeds). You could try balancing them more evenly by getting the pressure inside the case as close to outside as pos. Not sure exactly how you would do this though, other than by experimenting.

If you were determind you could count the pulses per second and work out the imbalance from that.

I don't have that many fans. A thermostatic exhaust fan (92 mms or so) plugged into the 3 pin motherboard. A small CPU fan, An Antec PP303X power supply with a "smart" fan. Thats it and I can hear when fans kick from silent mode to cooling mode. I get 66dbA with my RadioShack sound meter Shoot -- a 747 is quieter than this beast!

Mostly I think it's vibration converting to audibles. . .

Doesn't have to be fans. You have all those hard drives. The pulsing could easily be "intermodulation" noise caused by minor speed differences between the many drives you have. (like jhh said) If one is spinning at 7100, another at 7200 a third at 7300 -- and maybe 7150 & 7250. A 7200 rpm drive resonates at 120Hz. 7100 -- 118Hz; 7300 - 122Hz. So you probably have a bunch of noise sources at very slightly different frequencies... for sure you will hear some beating/bleating/pulsing effects.

MikeC - I agree with you that the "beating" is probably my two drive running out of sync. Especially since I am still running my two old drives, not the new ones. The are an IBM at 10K and the original Seagate at 7K. I need to get moving on quieting this boombox. . .

I don't think 7000Hz and 10kHz waveforms are close enough to intermodulate, most likely it's drives of the same spec running at slightly different speeds due to manufacturing tolerances. Such is life.

I checked the power consumption of my maxtors. They do 7.4 watts at idle/12.5 watts seek. Pretty close to yours. If you build a tight fitting rubber box and use a healthy amount of thermal paste, the rubber will acutally conduct the heat away from the hdd's. The only issue is that the rubber needs a fan to carry the heat away (but not at full voltage mind you). Look at my sig, those hdds temps are when my drives are seeking heavily and with a Panaflo @ 6V blowing on them.

EDIT: there's no formula that I know of to convert bels to dB/A, but my QAD method to get a general idea is to multiply bels by 10 to get dB.

you CAN convert bels to dB-A, but you need to know the frequency(ies) of the sound since the A scale is weighted to the human auditory scale. This'd be pretty hard to find out, I guess you could use a tuning fork but seems like a bit too much messing about for me.