DC drives & dampening mounts - OK?

sphinx99

Gawd
Joined
Dec 23, 2006
Messages
957
I have some DC drives I'm putting into a home storage server (mix of Exos x16 and WD HC530). The chassis I'm using uses trays with rubber grommets for drive isolation. I am perfectly fine using them, however I was not certain whether modern datacenter drives are happy with dampened mounts. My main concern was the use of grommets giving rise to resonances the drive would be better not seeing. It's possible this concern is non-existent or the effect is in the noise of what these drives can handle. Hoping an expert in the use of these drives @ scale can comment.

(My application is small, about 10 drives total.)
 

mwroobel

Supreme [H]ardness
Joined
Jul 24, 2008
Messages
5,085
Many data center drives, including the Exos have rotational vibration sensors to tune the movement and performance of the drives heads based upon the drives own vibration and the vibrations of its enclosure. You should have no issues unless you have an exceptionally intolerant environment where the drive is unable to properly accommodate the external influences and performance degrades because of it.
 

honegod

Supreme [H]ardness
Joined
Aug 31, 2000
Messages
7,854
I similarly wonder about rubber mounting my DC HC530 drives, my concern is that the head might need to be solidly mounted to accuratly accelerate and slow into position, the rubber flex could throw off the torque curve and add thrashing to find the right spot since the arm doesn't move the way it should because the case is absorbing some of the motor energy.

right now I have 4 drives (identical 14TB) each mounted on 4 rubber isolators used by drones to damp something that I forget atm.

could I be losing R & W speed because of inaccurate head positioning ?
 

mwroobel

Supreme [H]ardness
Joined
Jul 24, 2008
Messages
5,085
I similarly wonder about rubber mounting my DC HC530 drives, my concern is that the head might need to be solidly mounted to accuratly accelerate and slow into position, the rubber flex could throw off the torque curve and add thrashing to find the right spot since the arm doesn't move the way it should because the case is absorbing some of the motor energy.

right now I have 4 drives (identical 14TB) each mounted on 4 rubber isolators used by drones to damp something that I forget atm.

could I be losing R & W speed because of inaccurate head positioning ?
Unless you have a situation where you have significant external stress or sympathetic vibration from case to drive you should be ok with any data center drive IMHO.
 

honegod

Supreme [H]ardness
Joined
Aug 31, 2000
Messages
7,854
right.
since the drive case is floating on not very stiff rubber instead of being bolted to a slab of metal
the case can recoil from the action of the arm, which tells the case that the slab of metal it thinks it is bolted to is thrashing around
180 out of phase with the arm with equal energy to what moves the arm.
but only in the plane of the discs.
so when the arm motor yanks the arm across the disc, half of its energy is used to yank the entire drive the other way.
being much more massive than the arm the case will move less than the arm will, but it will move.
it is this motion that concerns me.
everything rotates around the axel of the arm which is well offset from the center of mass of the drive AND the axis of rotation of the disc stack.
I am using the bottom holes so the drive case is supported lying flat on the upright rubber cylinders.
all the energy fed into twisting the case is fed into the mounts in their most flexible direction.

writing this I realize that if I put a rubber mount in the side hole furthest from the arm pivot it would oppose that motion with its strong axis.
 

rhkcommander959

Limp Gawd
Joined
Mar 26, 2020
Messages
219
Interesting thoughts, those arms move extremely fast though. I'm not sure how much difference it makes either way, but I'm sure shook absorbent material is fine.

You get movement from other sources too, fans mostly so reducing outside motion is good.
 

daglesj

Supreme [H]ardness
Joined
May 7, 2005
Messages
5,582
Dunno about the rest but I place my NAS boxes on sorbothane feet and it cut down the noise from them a great deal. Decouples from what they sit on.
 
Last edited:

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
20,133
I have one enterprise drive that vibrates whatever its touching and the only way to completely silence it was fold a few large thick knit socks and place it on top electronics side up, in open space not a caddy, cooled with a fan blowing nearby.
Works a treat.
 
Last edited:

daglesj

Supreme [H]ardness
Joined
May 7, 2005
Messages
5,582
I have one enterprise drive that vibrates whatever its touching and the only way to completely silence it was fold a few large thick knit socks and place it on top electronics side up, in open space not a caddy, cooled with a fan blowing nearby.
Works a treat.
I have placed layers of electrical tape on the sides or where a drive comes into contact with the case if possible.
 

honegod

Supreme [H]ardness
Joined
Aug 31, 2000
Messages
7,854
thinking more about this, a primary point of rubber mounting drives is to reduce noise from whatever the drive is mounted to,
seems to be that the drive shakes the mount, which acts like a speaker diapragm creating sound in the air.
the rubber decouples the source, but allows more vibration of the source.
it also decouples the drive from any vibrations the mount might try to pass to the drive, like from other drives.

so what if the mount was solid and much more massive than the drive.
like directly bolting the drive to a 20 pound concrete paver block.

I would think the drive would be able to make very little sound happen, AND there would be little vibration pumped into the drive case to disturb head positioning.
 

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
20,133
Rubber only provides a certain amount of decoupling.
If the energy is strong it cant do enough.
How hard the rubber is also varies its resonant frequency.

If you lock the drive down solidly, my guess is the vibrations cant escape and there could be internal drive damage.
The energy will be expended somewhere, possibly with the head hitting a vibrating platter.
 

honegod

Supreme [H]ardness
Joined
Aug 31, 2000
Messages
7,854
If you lock the drive down solidly, my guess is the vibrations cant escape and there could be internal drive damage.
The energy will be expended somewhere, possibly with the head hitting a vibrating platter.
my visualization has the vibrations turning into heat, and since the outer shell cannot vibrate all the heat will happen inside the drive.
since the drive will be trying, and failing, to move the paver it will transfer energy into the paver, which will heat it, and so taking some energy from the inner bits, but not a lot.

How hard the rubber is also varies its resonant frequency.
yes.
I am assuming that the internal parts are designed to absorb the frequencies that the drive itself will produce in normal operation and turn that energy into heat that can be transferred to the outer bits to be dissipated into the airstream from the fan.

I would be concerned that the cement would not absorb part of the inner energy like the rubber mounts do.
like it is a trade, increasing the effeciency of the motors to position the components by increasing the heat generated inside the drive.
opposed to less heat inside because the rubber allows the positioning energy to wiggle the rubber around too, instead of only the head array.
 
Top