Leo Quan's rubber HDD box - a modern approach

[Felger Carbon]

January 22, 2003 - wow, that's over 3 years ago. That's when Mike Chin described Leo Quan's rubber box for HDD quieting. A really interesting read, especially in view of the 20dBA quieting that was achieved (confirmed by measurement with SLM)!

Alas, there was also a substantial temperature rise, about 32C above ambient, assuming a normal room (the article didn't specify the room ambient). But then Leo added a quiet fan to somewhat reduce that rise, to about 23C above room ambient.

That 32C temperature rise (with the 3/8" rubber box) should be compared to the 16C rise of a naked suspended HDD (both figures with no fan). So about 16C is caused by the rubber box.

Hmm. My Samsung SP08x2Ns aren't all that noisy. Perhaps 3/16" rubber would give me ~10dBA with an extra ~8C rise?? Or 3/32" for 5dBA and 4C? When I went looking for sources of that particular rubber I found 1/4" but nothing thinner:

1/4" thick rubber 5 sizes from 4" x 6" $1.08 to 12" x 18" $6.48 here.

3/8" thick rubber 4" x 6" $3.59/2 or 9" x 12" $14.29/2 here.

3/8" thick rubber 6.75" x 11" $7.59 here.

Let's stop and think for a minute: Leo used lots and lots of thermal grease on the top and sides of the HDD to get good thermal contact. Is there an easier, less messy way? What if the rubber came with an adhesive backing?

Ahem. Check this catalog and "find" 8583K45. $9.65 for a square foot of 3/16" buna rubber with adhesive backing. You have to choose a hardness: Durometer 40A is equal to the above rubber blocks but I'd go for 30A, about as soft as a rubber band.

The adhesive backing greatly simplifys the construction project. Instead of building a close-fitting box (and using messy thermal grease), just stick rubber to the top, bottom, and sides of the HDD. On the bottom, you can "trim" the acrylic on the sticky side so the rubber doesn't stick to the electronic stuff. And I'd leave a hole over the biggest IC chip to let it breathe, even if that did allow a little sound to escape.

Want a different rubber thickness? Go to page 3328 (find 'page 3328'), and make sure you go to the second table, the one with the adhesive backing.

Some numbers: this buna rubber weighs about 76 lb/cu ft, or 1.2 lb/sq ft for 3/16". It takes about half a square foot to cover a 3.5" drive.

Other alternatives? Sure. The automotive-oriented damping material Dynamat Extreme (DE), which is .44 lb/sq ft adhesive-backed buna rubber, is what I've treated my #1 computer case with. Works great. I got lots left over. Autos have funny curves, so DE is very soft (uncured) buna, which means it comes with an aluminum foil backing so you can handle it. (For the curious, I've deliberately left #2 computer untreated so I can make the comparison.)

Anyhow, DE is almost 1/5 as thick as the 3/8" Leo used. Why not apply one layer of DE and check the temperature rise? From that data, it should be easy to project the rise with two layers, and if that rise looks OK to you it's very easy to add another layer of DE. As you continue to add layers of DE, the drive will continue to get quieter and the temperature will continue to increase.

In fact, I think this last approach is better than using 3/16" stuff from McMaster. Especially since you can also use the DE to line your case: see this report. Let me re-emphasize that you don't want plain Dynamat, which is asphalt and smelly. You want Dynamat Extreme, which is very soft buna rubber.

Ralf: some folk provide photographs but few numbers and no sources. Since I've always been backward, I provide no photographs but I do provide numbers including prices and also links to on-line sources instead of "go to your neighborhood retail outlet".

[Live]

Let's stop and think for a minute: Leo used lots and lots of thermal grease on the top and sides of the HDD to get good thermal contact. Is there an easier, less messy way? What if the rubber came with an adhesive backing?

When I stop to think for a minute I think about that thermal grease is good at transfer heat which is not the case for adhesive per se. There are of course so called thermal adhesives but I don’t see any indication that is being used in the material linked. But then again transferring heat to "rubber" isn’t very effective in the first place.

[Felger Carbon]

When I stop to think for a minute I think about that thermal grease is good at transfer heat which is not the case for adhesive per se.

Thermal grease is terrific for transferring high CPU heat loads over a very small area. HDDs generate only 7-8W, and the area over which the heat is transferred is huge. We're talking a 100-1 difference in heat per unit area!

Leo used thermal grease to eliminate air pockets in his box. A layer of air is a terrific thermal insulator - exactly what you don't want!

[Aris]

from the reviews ive seen on here, the majority of noise from a HD is emitted from either the top or bottom. not so much the sides

how about this:

use somthing like this HD cooler - Zalman ZM-2HC2

and take the self adhisive backed rubber and apply it to the top/bottom/front/back of the hard drive, put the layer on top so its between the hard drive and heatpipes

that way you get the isolation and sound dampening, but still get great cooling cause the heatpipes are on the outside of the insulation

[piglickjf]

Leo used thermal grease to eliminate air pockets in his box. A layer of air is a terrific thermal insulator - exactly what you don't want!

Isn't a layer of rubber a pretty terrific thermal insulator as well?

PigLick

[Felger Carbon]

Leo used thermal grease to eliminate air pockets in his box. A layer of air is a terrific thermal insulator - exactly what you don't want!

Isn't a layer of rubber a pretty terrific thermal insulator as well?

Yes, but rubber also attenuates acoustic noise, while air does not, if I understand things correctly.

[Felger Carbon]

We'll get to the rubber box in a moment, but first:

I now have 2 HDDs in my ECE3505 chassis. I have one grommet-mounted in the top of the removable 3.5" cage, and one suspended directly below the first using Stretch Magic.

There are two 80mm Global Win NCB fans, each .7" to 1" away from the nearest HDD. The top fan is between the grommet HDD and the suspended HDD. Top to bottom: HDD, fan, HDD, fan.

Here are the HDD temperature rises above ambient for this arrangement for various fan voltages:

2 fans 0V susp 17C grom 14C
2 fans 7V susp 9C grom 9C
2 fans 12V susp 8C grom 8C

The two fans are very quiet with 0V applied , but the HDD temps go up. I'm a bit surprised that the 7V and 12V numbers are so close, but I measured twice, on different days, and that's what I got.

If I want really low vibration, I could replace the grommet-mounted HDD with a 2.5" drive in the same space, but suspended. This drive is for the swap file and for use during backup, so it's not performance-sensitive nor need it be large.

Now we can talk about the rubber box.

Leo got 32C rise by using a 3/8" rubber box. If 17C of that is just the HDD hanging in air, then the box added 15C, or 5C per 1/8". Some of the weight of Dynamat Extreme (DE) is the aluminum backing, so I'm now guessing that DE is about equal to 1/16" rubber in terms of sound attenuation and added temp rise. So:

0 layer DE = 0dBA improvement, 9C rise
1 layer DE ~ -3.3dBA, +2.5C
2 layers DE ~ -6.6dBA, +5C
3 layers DE ~ -10dBA, +7.5C

Above two layers, you might be better off buying the adhesive-backed buna from McMaster. 3/16" = 3 layers, etc.

For my experiment, the first step will be to wrap my suspended drive in one layer of DE and see if the resulting temperature is indeed 2-3C higher. If so, then I'd proceed to a second layer and, perhaps, a third layer just on the top of the drive.

DE is _much_ easier to apply to the bottom of the HDD than a single piece of thick rubber from McMaster. Using the thick piece, there's only a narrow strip outlining the drive where contact is possible. Not until the sides are stuck on does the bottom piece 'hold its place' well. Using DE, you cut the two sides and one end 1/2" longer in three directions (an inch wider overall), notch the corners, and fold the DE up against the sides and one end.

Don't forget not to remove the adhesive covering on the major part of the bottom cover.

Also, you don't want to close off the very small "breather hole" on the end of the HDD. Leave a small air passage between that hole and the outside world. I'm not sure how Leo handled that; Mike's writeup didn't mention it.

[IsaacKuo]

2 fans 7V susp 9C grom 9C
2 fans 12V susp 8C grom 8C
[...]
I'm a bit surprised that the 7V and 12V numbers are so close, but I measured twice, on different days, and that's what I got.

Well, the amount of airflow doesn't change the total heat exchange area, so the only thing it can change is the temperature of the air in contact with the heat exchange area. How much of a temperature difference can this be? It starts off as the same cool fresh air; the only difference is how much the hard drive itself heats up the air along the way. I suppose the air is not getting heated up much at 7V already, so there's not much room for improvement.

Leo got 32C rise by using a 3/8" rubber box. If 17C of that is just the HDD hanging in air, then the box added 15C, or 5C per 1/8". Some of the weight of Dynamat Extreme (DE) is the aluminum backing, so I'm now guessing that DE is about equal to 1/16" rubber in terms of sound attenuation and added temp rise.

You might get better results than you expect. A composite approach with an absorbing layer within a reflecting layer can be more effective than a single homogenous layer. That's because the outer layer can bounce sound around within the enclosure to make it go through the absorbing layer with multiple passes. If it's just an absorbing layer, then the sound passes through only once--limiting the amount of sound reduction. If it's just the a reflective layer, then the sound repeatedly bounces around but ultimately much of it escapes anyway because there's not much actually absorbing the noise.

I find that a good outer layer of even thin metal makes a world of difference in noise reduction. There must be no gaps, though! You'd be shocked how much noise escapes out of small gaps.

[peteamer]

Hi Guys 'n' Girls...,
(Disclaimer: My memory is rubbish... kinda like a... er... you know... a... er... Sorry, where were we?... )


1/

I'm a bit surprised that the 7V and 12V numbers are so close, but I measured twice, on different days, and that's what I got.

2/

Well, the amount of airflow doesn't change the total heat exchange area, so the only thing it can change is the temperature of the air in contact with the heat exchange area. How much of a temperature difference can this be? It starts off as the same cool fresh air; the only difference is how much the hard drive itself heats up the air along the way.

To tie these two quotes together: (As far as my rubbish memory recalls...)

1:/ Don't forget 'Wind Chill'... It has it's greatest order of magnitude of change from 0-smallish numbers.
2:/ Fans follow a similar principle... i.e. Fans, when considering RPM, have a greater order of magnitude of change (in 'wind power' ) from 0- low numbers...

Couple that with the mechanical effect that dictates that it takes a disproportionate amount of growing power (volts/amps) to produce a 'doubling' of 'air pushing' power of a fan and you'll see that the whole thing makes more sense...


I.E. That a fan pushes a much smaller wind speed increase from 7-12V than from 0-7,... even given the difference in the difference... if you see what I mean.

Wind Chill does a similar-ish thing...


I think the two add up... Though what I would expect to see as a result I wouldn't even hazard a guess... ... Perhaps as above...



Now... If someone who actually knows what they are talking about could come along...




Regards
Pete

[Felger Carbon]

Disk seek noise

When starting up Win98SE, I can clearly hear the disk seeks, about 4 bunches of them, when starting my #1 or #2 system. All 4 drives are SP 80G units, veddy quiet. Each of the 4 drives has 10ea 2G partitions, or 20 partitions per computer.

When I empty the Recycle Bin, Win98 has to seek the FAT directory of each partition, 10 per drive for 20 partitions, to see if copies of the to-be-deleted file are on the other partitions. In the past I could clearly hear each individual seek as a "thunk"; back when my drives were all hard-mounted I could easily count the seek "thunks" and the last thunks - the ones from the drive in the 5.25" bay - were clearly louder.

I heard 10 "thunk"s and then 10 "THUNK"s!

System #1 has a case heavily (but incompletely) treated with Dynamat Extreme (DE). Drive C: is grommet-mounted, and D: is mounted in the top 5.25" drive slot, using Nexus Disktwin mounts with rubber dampening blocks.

When I empty the RB on computer #1, I hear nothing for as long as it takes to empty disk C:, and then 10 quiet thunks as it emptys the Twindisk-mounted drive in the 5.25" slot.

System #2's case is not yet treated with DE. Also, system #2 has that CAG opening on the left side toward my ear, not true of system #1. That's 2 reasons right there for #2 to be louder. However, drive C: is suspended and drive D: is grommet-mounted.

When I empty the RB on computer #2, I hear nothing at all.

So, booting is the only time I hear the hard drives on system #2. I infer that there is a great deal of big-time disk seeking going on during boot. Windows was once gonna fix that by arranging the sequence of files on the C: partition to minimize seeks, but that apparently went away.

Boot noise doesn't bother me. It's like CD-R noise: you hear it briefly, infrequently and most important of all, predictably. Noise that you know is coming is far less intrusive than noise that jumps out surprisingly.

I don't think I really need to replace the second disk, now grommet-mounted, with a suspended quieter 2.5" drive.

In fact, I'm not sure this disk sound insulation project is actually worthwhile if the case is treated with DE.

[TomZ]

1:/ Don't forget 'Wind Chill'...

Wind chill?

From WikiPedia:

Wind chill is the apparent temperature felt on exposed skin due to the combination of air temperature and wind speed.

I'm not sure how this relates?

[peteamer]

I'm not sure how this relates?

Wind Chill:
If you go out on a cold day you will feel/be a certain temperature.
If the wind picks up, even though the air temperature remains the same, you will get colder.

If you have a CPU with a fan running on the heatsink at 5V it has a certain temperature. If you then apply 12V to the fan the temperature of the CPU will drop even though the air fed to it is the same temp.

Whilst this is not strictly wind chill, due the lack of moisture enhanced temp. loss, the similarities are close enough for me to 'Bastardise' the expression to help get my point across. Hence why I wrapped the words with ' .


Or have I missunderstood your question?
(I assume you weren't just picking me up on my usage of words, whislt ignoring "Wind Chill does a similar-ish thing... ")


Pete

[TomZ]

If you have a CPU with a fan running on the heatsink at 5V it has a certain temperature. If you then apply 12V to the fan the temperature of the CPU will drop even though the air fed to it is the same temp.

OK, I understand. What you're saying is that higher CFM means more air is passing by the warm surface thus cooling down the HDD more. I can agree with that!

(Let's just leave out the bit about wind chill, since that brings in the additional aspect of having a moist surface, with the moving air causing the moisture to evaporate, thus cooling down the surface a little more due to the phase change. Unless you want to further cool your HDD by putting in a misting device that keeps it wet!)

[peteamer]

(Let's just leave out the bit about wind chill, since that brings in the additional aspect of having a moist surface, with the moving air causing the moisture to evaporate, thus cooling down the surface a little more due to the phase change. )

Whilst this is not strictly wind chill, due the lack of moisture enhanced temp. loss,

I promise not to get upset if you read All of my posts, I wont mind .. Honest.



Pete

[Felger Carbon]

2 fans 0V susp 17C grom 14C
2 fans 5V susp 10C grom 10C <-- new data
2 fans 7V susp 9C grom 9C
2 fans 12V susp 8C grom 8C

As I implied when starting this topic, there are a million ways to quieten an HDD. I thought an updated version of Leo's rubber enclosure might be a good way of doing the job. I can now report that I was right; I've found a disk enclosure using rubber for sound insulation that works to even hide the noise my disk makes when booting Windows at 3AM. And, wonder of wonders, my HDD temperatures have not gone up at all!

Leo's box measured appx 7 x 4.75 x 1.75 inches. Mine measures 7.6" by a tad over 14" high and 16" deep. Yep, my quiet HDD enclosure is also my computer case. No, I didn't expect to wind up where I am now. But there are a million ways...

I had already, a while back, applied a layer of Dynamat Extreme (active ingredient buna rubber) to the side covers, the top cover, and the top and front side of the PSU. With the boot HDD grommet-mounted, the seeking during boot was clearly audible. Now, it isn't audible at all. What changed?

The main thing was my belated recognition that the mobo base plate is also the "spine" of the computer case. This is what the HDD cage attaches to and if it shakes, the whole computer shakes. So, yesterday I applied just one layer of DE to the mobo base plate on the side opposite the mobo and adjacent to the right side cover. I also applied some DE to the inside of the plastic escutcheon. Presto! No more HDD seek noise when booting! And that's with one layer of DE.

You can add layers of DE, and each added layer produces more sound attenuation. So I added a second layer to the front 9", a third layer to the front 6", and a fourth layer to the front 3" of the mobo base plate. I can't say the computer got any quieter because I already couldn't hear any disk noise. Let's say I have insurance against a future noisier HDD.

And my boot drive is still grommet-mounted. Let me make this clear: Mike sez suspension is better than grommet-mounting, and he's right. I proved that to myself by suspending a drive in my #2 computer. But grommet-mounting is easy and produces a safe, sturdy, transportable computer. The difference can be made up by agressive application of damping material such as DE.

How agressive? Until yesterday, I had added 3 sq feet of DE at .44lb/sq ft to my computer. Yesterday, I added another 3 sq ft. My computer is now ~3lb heavier because of the added DE.

The cost? About the same as just one HDD "quieting" enclosure ($40 or so), only my version doesn't heat up the HDD.

Now that I've identified the mobo base plate as critical to proper damping of a computer case, I'm thinking of experimenting with adhesive-backed buna rubber sheets from McMaster ($9.67 for 1 sq ft of 3/16" buna, roughly equal to 3 layers of DE) on my #2 computer case mobo base plate, and perhaps another 1/8" on the front side of the plate on the 6" in front of the mobo itself. This will make for a heavy, and I believe quiet, computer.

This will make for a neater installation than multi-layer DE. Did I mention that my new approach doesn't result in an HDD temperature rise? That is a significant improvement over Leo's 3-yr-old rubber box.

[makutaku]

I am considering using your approach in my htpc case but I am having problem visualizing what you have done. Would you mind to post some pictures ?

[Felger Carbon]

I am considering using your approach in my htpc case but I am having problem visualizing what you have done. Would you mind to post some pictures ?

I can't get my W98SE-Asus K8S-MX system to work with USB drives - so no pics, sorry. Odd, a USB mouse works great. But not drives.

[makutaku]

I don't understand exactly what you mean by "mobo base plate". I am quite fluent in English but I am not a native English speaker. Could you please elaborate on that ?

I am trying to figure out the critical areas to proper damping in my desktop htpc case (SilverStone LC13). Here are some pictures of it:

http://www.pcalchemy.com/product_info.p ... htpc-cases

[ultrachrome]

Has anyone considered thermal interface rubber? I don't know much about it. I first discovered it while servicing an early Mac Powerbook.

Several logic chips was covered this squishy light green rubber pad that was cool to the touch. Googe turned up stockwell.com but I'm sure there are others.

I don't think it's available very thick but you could use multiple layers to achieve your goals while still allowing the drive to dissipate some heat.

[Felger Carbon]

I don't understand exactly what you mean by "mobo base plate". I am quite fluent in English but I am not a native English speaker. Could you please elaborate on that ?

The sheet metal slab that the mobo is mounted on. On most tower cases, it's close to, and parallel with, the right side cover.

[jaganath]

I don't understand exactly what you mean by "mobo base plate".

I think he is referring to this, right FC?

[Felger Carbon]

I think he is referring to this, right FC?

Yep.

[Bill Dunsmore]

Felger, I have a work in progress that you may be interested in. It's the basic carving rubber box - at least, on the top, bottom, front and rear. The sides consist of two 1 x 1 x 1/8 inch pieces of aluminum "L" angle stock bolted together to form a "C". The aluminum "C" is attached directly to the sides of the hard drive and holds the rubber pieces in place. Testing is in progress with my Barracuda IV enclosed. So far, temperatures have remained below 40 degrees in still 25 degree ambient conditions. The enclosed Barracuda IV is just barely audible outside the PC case. Will report further if any interest is expressed.

[Felger Carbon]

Bill, as I understand it you're blocking sound and heat from escaping except from the sides. My question is, what is the aluminum contacting aside from air? In other words, is there anything helping the aluminum dissipate the heat?

You're showing a 15C rise above ambient. I get a 17C rise above ambient when I suspend a plain SP08x2N, no rubber and no aluminum.

[Bill Dunsmore]

Your assumption is correct. Nothing but air. No fan either. Apparently, the thickness of the aluminum is helpful. However, now I notice a very low level but high frequency sound when seeking?

[Bill Dunsmore]

Looks like my little project is a success! The enclosure is sitting on bubble wrap in the bottom of the PC case and is SILENT! The little bit of high frequency noise turned out to be resonance in the top aluminum L pieces. I superglued a 3/4 inch square piece of carving rubber to each L and the resonance was damped out. A bit of Dynamat would probably also have worked well.

Continuing to monitor temperatures. Looks good so far.