Choice of Material for HD suspension

postul8or · Post by **postul8or** » Thu Jul 29, 2004 11:04 am

I was just wondering if the material used for suspension of a hard drive mattered much.

People seem to prefer elastic type material and I'm not sure why that would matter. What would happen if I just used a thin industrial rope or something to suspend it. Is there something about that material that would conduct vibration to the case more than elastic would?

Thanks to anyone who can explain!

mrzed · Post by **mrzed** » Thu Jul 29, 2004 12:01 pm

Think of a violin string. Transfers vibration and sound extremely well. The tighter the string, the louder the sound. Hence the tiny little violin can fill a concert hall.

Nice soft elastic will not transfer any noticeable vibration.

hofffam · Post by **hofffam** » Thu Jul 29, 2004 2:20 pm

mrzed - I agree a soft elastic material would work better than something hard or stiff. A contrasting extreme might be steel wire. Steel wire would transmit vibrations very well. Soft materials dissipate vibration energy by convering it to heat.

But I think your violin string example is wrong. First a tighter violin string changes the pitch much more than the loudness. If the tension on the strings caused a big change in loudness it would be very difficult to play the instrument. The amplification of sound from a violin string occurs because the hollow body of the violin resonates with the string. Note the difference between a solid body electric guitar (with no amplifier hooked up) and a hollow acoustic guitar. The acoustic guitar with essentially similar strings plays much louder because of the body.

markjia · Post by **markjia** » Thu Jul 29, 2004 2:56 pm

I think that a more important part is how the string is used.

Take the violin example. The volume of the violin is not determined by the tightness of the string (within reason). It's volume has to obey the laws of conservation of energy. It can only be as loud (or have as much acoustic energy) as the kinetic energy you put in when playing it.

When it comes to hard drives, my guess is that string reduces the acceleration of the hard drive's vibration (by allowing the drive to move a longer distance). In effect, it lower the frequences produced by the vibration. Perhaps you end up with an inaudible low frequency sound as a result (but this is completely a guess).

lenny · Post by **lenny** » Thu Jul 29, 2004 3:09 pm

I prefer sorbothane to elastic suspension, but my (not necessarily valid) hypothesis is that an elastic material will absorb / dampen some of the vibration, and that may be healthier for the hard disk than having it vibrate while suspended.

postul8or · Post by **postul8or** » Thu Jul 29, 2004 8:34 pm

lenny wrote:I prefer sorbothane to elastic suspension, but my (not necessarily valid) hypothesis is that an elastic material will absorb / dampen some of the vibration, and that may be healthier for the hard disk than having it vibrate while suspended.

I'd have to read a little more about the method of using sorbothane. Material is one thing, how you use it is another. Are you just resting it on sorbothane at the bottom of the case, or surrounding all edges of the drivers in a tight box type of formation?

The "light at the end of the tunnel" that I see for suspension is that I can blow a fan on it and keep it cool if I need to. If I box the thing tightly I won't be able to draw heat away with a fan unless the box is metallic/conductive. Sorbothane strikes me as an ultra heat insulator type of material.

As for your assertion that vibration while suspended could be bad, I see your point. Maybe hard drives were designed to be attached to a bay tightly for a reason that is maybe something more than just heat discipation. A manufacturer should be able to shed light on this, who could we ask?

---------------------

As for my point about using elastic vs. plain industrial rope I'm not sure if there is really a proven difference from what I hear so far. I would almost think an elastic type material might be denser and transmit sound better than something more woven with air pockets (eg. maybe a shoelace or something!). Anyway, unless somebody has done a sound conduction experiment on the materials in different wires/cables/ropes we won't have too many scientific answers I guess. The idea that the tension might have an impact is interesting, although I agree the hollow nature near a guitar is the majority of the amplification. The inside of a computer case is probably a lot like an acoustic guitar.

I agree with the thought that metal cable would be a poor performer.

edited: replaced "wires" with "wires/cables/ropes" in that last big paragraph.

lenny · Post by **lenny** » Thu Jul 29, 2004 10:13 pm

postul8or wrote:I'd have to read a little more about the method of using sorbothane. Material is one thing, how you use it is another. Are you just resting it on sorbothane at the bottom of the case, or surrounding all edges of the drivers in a tight box type of formation?

For me I'm resting the drive on blocks of Sorbothane. They're quite tacky and grips on pretty tight to the drive and the floor of the case. I'm a little nervous about having the Sorbothane come into direct contact with the PCB, and I'm not too keen about having the drive upside down (although it is probably harmless), so I attached those 3.5" to 5.25" adapters to the side of the drive, and put the Sorbothane on the adapter.

rperezlo · Post by **rperezlo** » Fri Jul 30, 2004 2:01 am

I have actually tried both 8 mm elastic cord and a thin (I would say 1 mm) packaging rope. I was lazy to go to the shop to buy the elastic rope so that I tried whatever I had at hand.

Well, the difference is quite big. The rope did almost nothing to improve the vibration, and in some sense it even made the noise of the disk more noticeable as it placed it closer to my ear. Meanwhile the elastic cord effect was the one you all know.

Resume: it does matter what kind of cord you use.

hofffam · Post by **hofffam** » Fri Jul 30, 2004 7:08 am

Postul8or - I think your idea of a woven material has merit. I understand that some portion of a vibration traveling through a material "reflects" back each time it encounters a new substance or barrier. That is the basic principle behind "sandwich" construction used in some audio equipment, especially turntables, to isolate a component from external vibration. A woven material is made of multiple strands and and I think a vibration would naturally be damped by the "friction" between the strands.

I believe this is similar to the effect of sand - also used in some turntable isolation stands. These stands hold a turntable on a bed of sand. I believe the vibration just "dies" as it goes from each grain of sand to another.

I bet a hard drive sitting on a small sand bag (like a bean bag) would be very quiet.

Spod · Post by **Spod** » Fri Jul 30, 2004 7:16 am

Suspension isn't bad for a hard drive, but may increase the average seek time. I think it was Seagate that actually released a white paper on this. It has to do with the calibration of the actuator being designed for the drive being hard mounted - if it is soft mounted, the head will tend to over/undershoot the desired track, and take a little more time to get back on track. Its an expected vs. experienced acceleration thing.

Post by **MikeC** » Fri Jul 30, 2004 9:12 am

The type of string material does matter a lot, and when using elastic, so does the tightness:

The softer and looser the elastic is, the more effective the damping & noise reduction; the harder and tighter, the less effective.

No way non-elastic string has any real merit for damping. Unless the HDD is just loosly sat upon it -- which makes for a precarious setup.

My favorite approach is to use a multi-mesh of stretchy thin elastic (for clothing, found in sewing shops, really cheap) so that the drive is not too tightly held yet really secure (because of the multiple criss-cross pattern. It makes the drive as well damped as when you hold it in your hand. Live human flesh is an excellent vibration dampener...

mrzed · Post by **mrzed** » Fri Jul 30, 2004 9:49 am

How very David Cronenberg of you Mike!

Perhaps it's time for an official SPCR guide to live flesh damping techniques. Perhaps in time for October 31?

postul8or · Post by **postul8or** » Fri Jul 30, 2004 9:44 pm

Responding to a bunch of you, thanks a bunch for the input I really appreciate the ideas.

lenny wrote:For me I'm resting the drive on blocks of Sorbothane. They're quite tacky and grips on pretty tight to the drive and the floor of the case. I'm a little nervous about having the Sorbothane come into direct contact with the PCB, and I'm not too keen about having the drive upside down (although it is probably harmless), so I attached those 3.5" to 5.25" adapters to the side of the drive, and put the Sorbothane on the adapter.

I am surprised that heat isn't an issue with this method. Are you monitoring your temps? Do you have a decent amount of air flow to draw the heat away?

I have actually tried both 8 mm elastic cord and a thin (I would say 1 mm) packaging rope. I was lazy to go to the shop to buy the elastic rope so that I tried whatever I had at hand.

rperezlo wrote:Well, the difference is quite big. The rope did almost nothing to improve the vibration, and in some sense it even made the noise of the disk more noticeable as it placed it closer to my ear. Meanwhile the elastic cord effect was the one you all know.

I could see the 1 mm packaging twine being too much of an extreme. With such little support the hard drive probably buzzes it's own outer case or something. As I said, I was thinking what if I used something like a shoe lace. Bigger than 1 mm (if it's flat it might be only 1 mm on 1 dimension I guess). Anyway, your experience is valuable and tells us that too thin is something to stay away from. There is a post I was reading where a guy used a pretty thin stretchy basically clear rope. He claimed it worked well, it was quite thin, not 1 mm thin though.

hofffam wrote:Postul8or - I think your idea of a woven material has merit. I understand that some portion of a vibration traveling through a material "reflects" back each time it encounters a new substance or barrier. That is the basic principle behind "sandwich" construction used in some audio equipment, especially turntables, to isolate a component from external vibration. A woven material is made of multiple strands and and I think a vibration would naturally be damped by the "friction" between the strands.

I believe this is similar to the effect of sand - also used in some turntable isolation stands. These stands hold a turntable on a bed of sand. I believe the vibration just "dies" as it goes from each grain of sand to another.

I bet a hard drive sitting on a small sand bag (like a bean bag) would be very quiet.

Yes, I'm thinking along the same lines as you are. Basically you want pockets to help deaden the sound. Conduction won't be solid because the material isn't totally compacted, but also the dead spaces can act as an area where the sound reflects in all directions rather than all of it in 1 direction down the string material. Something with strands, but still stretchy I think could be the ultimate choice. As MikeC says, the tension is a large part of it. A low tension elastic will damped sound but also hold you hard drive in place better than a material that doesn't flex.

In terms of using sand I nominate bluefront to try something with it! He makes birdhouses, now maybe he'll move on to sandboxes!

Spod wrote:Suspension isn't bad for a hard drive, but may increase the average seek time. I think it was Seagate that actually released a white paper on this. It has to do with the calibration of the actuator being designed for the drive being hard mounted - if it is soft mounted, the head will tend to over/undershoot the desired track, and take a little more time to get back on track. Its an expected vs. experienced acceleration thing.

Now that you say that, I think I have read that in another thread before. Thanks for the reminder. I wonder if the change in the seek time is noticable. I guess we'd rather hear the news that it might slow things down a little as opposed to greatly increasing probability of a crash. I suppose they could engineer a drive for suspensions that would not overshoot. I bet the production numbers wouldn't be high enough to be profitable.

MikeC wrote:The type of string material does matter a lot, and when using elastic, so does the tightness:

The softer and looser the elastic is, the more effective the damping & noise reduction; the harder and tighter, the less effective.

No way non-elastic string has any real merit for damping. Unless the HDD is just loosly sat upon it -- which makes for a precarious setup.

My favorite approach is to use a multi-mesh of stretchy thin elastic (for clothing, found in sewing shops, really cheap) so that the drive is not too tightly held yet really secure (because of the multiple criss-cross pattern. It makes the drive as well damped as when you hold it in your hand. Live human flesh is an excellent vibration dampener...

I think you make a good point about the tension. I think the higher tension will transmit higher frequencies better. I think eliminating high frequencies should be the focus for hard drives since a lot of the noise is high frequency. A second factor is that low frequency is hard to block, it requires more of a thick barrier approach as far as I know. As I said above, having a stretchy material allows lower tension, but still the ability for the hard drive to grab the material.

The main thing I noticed at crafty type shops is that their elastic was thin/flat and kind of wide on the other dimension (1mm thick, variety of wideness 5mm to 25mm), as opposed to circular. The closest thing I could think of is bungee cord, like the stuff you use for cars with the hooks on the end. That is pretty thick stuff though, would be too much drilling in my case to feed it through! Is having the round stuff important, or is that flat stuff just as good or better?

As far as human flesh being a great dampener, it's kind of strange. What I find interesting is that you get benefits of a solid mass, but filled with liquid. I believe the 2 different states of matter will block different waves, a pretty good combo to have. That's a guess. Ever notice when an antenna doesn't work (eg. the radio) if you touch the antenna with your hand it works a lot better! The human body: noise blocker and signal amplifier!

Post by **MikeC** » Fri Jul 30, 2004 10:38 pm

Forget "crafty" shops -- go to a good sewing shop. Hobnob w/ladies and ask them about the round cross-section elastic. It's been in all of the 3 sewing shops I've tried, in various diameters and stretchiness.

lenny · Post by **lenny** » Fri Jul 30, 2004 11:59 pm

postul8or wrote:
lenny wrote:For me I'm resting the drive on blocks of Sorbothane. They're quite tacky and grips on pretty tight to the drive and the floor of the case. I'm a little nervous about having the Sorbothane come into direct contact with the PCB, and I'm not too keen about having the drive upside down (although it is probably harmless), so I attached those 3.5" to 5.25" adapters to the side of the drive, and put the Sorbothane on the adapter.
I am surprised that heat isn't an issue with this method. Are you monitoring your temps? Do you have a decent amount of air flow to draw the heat away?

Why do you think there will be a big difference in temperature between a hard mount in the drive cage and resting on the base with Sorbothane?

On one system I have dtemp running all the time. The temperature is a couple of degrees higher than hard mount. It sometimes get above 40C so I have a fan blowing on it at 5V. I experimented with the Zalman HDD heat pipe thingie, but it doesn't seem to make any difference.

On another system, I saw temperature went up by 8C or so, which was quite surprising to me. This is a standard tower case and there isn't any intake fan. The temperature rise may have something to do with having the case dampened, and in the process some side intake holes sealed.

hofffam · Post by **hofffam** » Sat Jul 31, 2004 6:01 am

Postul8or - the vibration put out by a hard drive is not that high in frequency. If you're talking about a 7200 rpm drive, that is equivalent to just 120 rp(seconds) or 120 Hz. Most of us hear a hard drive spinning as a whine but a 120 Hz pure tone sounds higher than most people expect. The 27th key of a piano produces 123Hz. I expect there are other harmonics (higher frequency multiples) to the sound of a hard drive - and they might be magnified when the computer case resonates from the hard drive vibration.

postul8or · Post by **postul8or** » Sat Jul 31, 2004 10:26 am

MikeC wrote:Forget "crafty" shops -- go to a good sewing shop. Hobnob w/ladies and ask them about the round cross-section elastic. It's been in all of the 3 sewing shops I've tried, in various diameters and stretchiness.

Well it was a fairly decent sewing shop I would guess, it's not my area I've got to admit. They had a lot more ribbon than an average shop was my thinking which was why I thought they were a little crafty.

What is your thinking for going with the round shape as opposed to flat? Do you feel you get better mass with the round stuff and that improves performance of the suspension?

lenny wrote:Why do you think there will be a big difference in temperature between a hard mount in the drive cage and resting on the base with Sorbothane?

If your hard drive is in a normal bay it will discipate heat through the metal it's attached to and if the drive is near the air intake at the front of your case it can be cooled from the top and bottom as well.

I think of sorbothane as insulation in the walls of a house. The wind is blowing, insulation is stopping it from getting into your house keeping your house warmer (in winter). ie. An insulator is an insulator

Just for reference....with my case put together as normal if my hard drive is idle I think it's about 33 C. If I take a side panel off and there isn't containment for there to be pressure, the air flows in through the side and not so much through the front. This lack of air flowing in from the front puts the idle temp of the hard drive closer to 40 C. I imagine sorbothane would deflect access to most of the air even if my side panel was on. 8 C based on lost airflow is what I would expect to lose, not alone heat has trouble escaping because you are using an insulator.

If the sorbothane is only on the bottom of your drive it's probably ok. If it is also on top of it I'm surprised it's not cooking. It looks like you know what you are doing and are monitoring the situation and doing what is required to manage it though.

hofffam wrote:Postul8or - the vibration put out by a hard drive is not that high in frequency. If you're talking about a 7200 rpm drive, that is equivalent to just 120 rp(seconds) or 120 Hz. Most of us hear a hard drive spinning as a whine but a 120 Hz pure tone sounds higher than most people expect. The 27th key of a piano produces 123Hz. I expect there are other harmonics (higher frequency multiples) to the sound of a hard drive - and they might be magnified when the computer case resonates from the hard drive vibration.

I wasn't aware that the math/physics was this simple, but what you say does make sense. In a totally clear sound transmitting medium I think your calculation makes sense, my inkling is that sound travelling through metal objects bouncing around in an irregularly shaped box should change frequencies somewhat. Maybe it only changes it by 1 Hz or something so what I'm saying is irrelevant, I don't have the expertise to answer that.

I can hear the noise from quite far away and I'm pretty sure that low frequency sound attenuates slower than high frequency sound so maybe on that basis I should have known the sound was pretty low frequency.

I'm thinking that MikeC's "thick aluminum plate" approach would be required to fully block the amount of lower frequency noise coming from my drive (a WD 120 Gb drive has got to be one of the loudest on the market). Quite honestly, with a stock CPU fan, 2 low flow Panaflo fans @ 12 V, Zalman PSU, and a Geforce Video card with stock fan the noise level from this machine is acceptable (quiet, but not silent). My best bet might be to ditch this hard drive.

Post by **MikeC** » Sat Jul 31, 2004 10:35 am

Forget trying to block the low freq noise -- it's just not possible. The best you can do is minimize the transmission of those low freq vibes to the rest of the case. This is where the suspension excels. With a properly suspended drive, you simply cannot feel any vibes when touching the case. Try it with a normally mounted HDD PC -- you can feel the whole box vibrating anywhere you touch it. On my systems, there is ZERO vibration to be felt. It removes a huge haze of low level low freq buzzing that yoy might not even notice right away, but put this noise back and man-o-man, you want scream to kick the thing after a while!

- edited for typos

josephclemente · Post by **josephclemente** » Sat Jul 31, 2004 3:18 pm

I use 1.8mm Stretch Magic. It works very well.

Stretch Magic suspension in more detail:
http://forums.silentpcreview.com/viewtopic.php?t=11674

Post by **Ralf Hutter** » Sun Aug 01, 2004 6:12 am

postul8or wrote:
I think of sorbothane as insulation in the walls of a house. The wind is blowing, insulation is stopping it from getting into your house keeping your house warmer (in winter). ie. An insulator is an insulator

My application of Sorbothane to prevent it from insulating the drive. Works fan-freaking-tastic:

jmpsmash · Post by **jmpsmash** » Sun Aug 01, 2004 3:05 pm

have anyone tried metal springs?

elastic band and sorbothane works great to isolate vibration but they are very poor heat conductors.

a metal spring will provide for both, the metal can conductor heat to the case and the spring will help isolate vibration.

i understand that the metal spring alone will be too springy, some dampening is probably needed. otherwise the HD will be dancing to it own tune in there...

postul8or · Post by **postul8or** » Mon Aug 02, 2004 8:48 am

Ralf Hutter wrote:
My application of Sorbothane to prevent it from insulating the drive. Works fan-freaking-tastic:

That's a good idea...only the bottom is covered, but you've also got it in one of those quick swap cases to give the bottom some room to disperse heat.

jmpsmash wrote:have anyone tried metal springs?

elastic band and sorbothane works great to isolate vibration but they are very poor heat conductors.

a metal spring will provide for both, the metal can conductor heat to the case and the spring will help isolate vibration.

i understand that the metal spring alone will be too springy, some dampening is probably needed. otherwise the HD will be dancing to it own tune in there...

I don't think metal as a suspension vehicle is a good plan. I would think it would transfer some vibration. Look for a softer material that isn't such a conductor is my feeling. The best thing I can think of is taking a metal spring, but putting a coating of rubber on it for extra dampening. After going to all that work I'm thinking the elastic type material people are using is probably more straightforward.

lenny · Post by **lenny** » Mon Aug 02, 2004 12:33 pm

postul8or wrote:If the sorbothane is only on the bottom of your drive it's probably ok. If it is also on top of it I'm surprised it's not cooking. It looks like you know what you are doing and are monitoring the situation and doing what is required to manage it though.

A picture is worth 2K bytes, or so they say. I apologize for not explaining it better. Ralf's picture is similar to my setup, except I added the 3.5" to 5.25" metal adapter arms to the side of the hard disk.

It's not to add additional surface area for heat dissipation - I don't think they'll work very well in that capacity anyway - but rather to provide a greater surface area for the Sorbothane to adhere to.

I tried placing it the way Ralf did (without the tray) but the hard disk (horror of horrors) slid off the Sorbothane. I didn't want Sorbothane touching the PCB of the hard drive.

So, in my case, the top and bottom are completely exposed to air, and the side is in direct contact with metal, but the whole assembly is resting on two strips of Sorbothane at the bottom of the case.

And I'm still nervous about temperatures.

pangit · Post by **pangit** » Mon Aug 02, 2004 4:40 pm

I was worried about rising temps with suspension mounting too, so this is my solution:

(You can see more pics and info at the page in my sig)

I made a home made hard drive cage/duct for 2 hard drives, then suspended the whole thing, thereby not losing the heatsink effect of hard mounting.

It also acts as a duct with the intake air flowing through it, with the angled fan directing it towards the GPU and Northbridge. So the HDDs get excellent airflow without impeding it to the rest of the case. And removing hard drives is just as easy as a normal hard mount. I've actually put a filter on the front (a la Bluefront) recently, I must update the website with new photos.

HDD temps are in the low 30s, and noise is pretty low too. Although of course I'm transmitting noise to the cage, but not the whole case. I haven't done a direct noise comparison between this setup or suspending the HDDs without the cage, but I guess the difference is pretty slight.

The elastic cord I used is not the softest, but the tension can be adjusted and it's pretty secure for transporting. I might replace it with stretchier thinner cord if I find any and see if it makes a difference.