How do I match HD rotational speeds exactly?

[redshadow]

From my physics classes I know that if there are two sound sources with very close frequencies, then you will hear a pulsation that is the exact difference of the two sources. In the case of my two hard drives, the rotational speed is specified to be 10000 RPM. If that is the case, there should not be any noise beyond that of the heads and the normal rotational noise.

The weird part is that half the time I hear the difference in the rotational speeds, and that difference is not constant. This difference can be anywhere from about 6Hz to 0.5Hz and is extremely annoying at times, even at a few meters distance.

Does anyone know what if anything electrical can be done to eliminate this noise? (yes, I can soundproof the case, but that wouldn't solve the underlying problem and would be waaaay too simple)

any ideas you may have are appreciated

 

[MisterChief]

Dude, seriously, is it really that noisey?

 

[Peter]

That's an interference vibration caused by tiny differences in rotational speed.

Back in the days, drives used to have an auxiliary connector where, amongst other stuff, there also was a "spindle sync" signal that you would connect to all drives.

 

[ReiAyanami]

yes, the resonance can be wall piercing. years ago when i had a maxtor and a wdc, it was atrocious. now even with the "ultra quiet" seagate barracuda's i put a 7200.7 V 160gb and 7200 IV 80gb next to each other and they still generate an annoying resonance pitch. however when i use two 7200.7 V's it stops.

 

[redshadow]

That's the problem...I have two identical drives, and the problem is intermittent...and a drive sync connector would be nice...

any other ideas?

 

[bobsmith1492]

Crack 'em open and hook the spindles together with a pulley?

ok, no.....

Does anyone know if the motors that drive the spindles are stepper motors or not? If they are, you could maybe connect the timing circuitry together so that one timer drives both of them. That could do the trick. Finding the circuitry would be the tough part though.

 

[PottedMeat]

Originally posted by: bobsmith1492
Crack 'em open and hook the spindles together with a pulley?

ok, no.....

Does anyone know if the motors that drive the spindles are stepper motors or not? If they are, you could maybe connect the timing circuitry together so that one timer drives both of them. That could do the trick. Finding the circuitry would be the tough part though.

sounds like an interesting project, on a bunch of old HDDs (IDE ~500 MB) you could easily identify the three pairs of mosfets for the three phase servomotor ( i think thats what they're called ) - as long as you could find the logic inputs and sever them from one controller IC, and then split one control output into both harddrives it may work. But then theres probably some crazy feedback stuff goin on where the rigged harddrive may get some data error and try to adjust the spindle speed for correction. heh you end up with identically spinning drives but can only access data from one.

 

[Howard]

Just insulate one of the hard drives with something that's relatively thermally-conductive.

EDIT: And sound-deadening, obviously.

 

[RichPLS]

Brilliant, a silencer drive encloser, they make them per drive or cages that hold up to 4 drives in a three bay slot.