Need help w/ a bit of Market Research...

[Gooserider]

This may be a bit off topic, though it does deal with waterblocks. I have noticed a significant lack of good info on watercooling hard drives, or hardware to do it with. I know there is some level of debate about whether it is better to cool a drive by putting a plate on the top or bottom of the drive vs. putting blocks on the sides of the drive housing. However there seems to be little doubt that cooling is needed on the new 10 and 15K RPM drives, especially when one tries to quiet their noise by putting sound insulation around them.

I have become convinced that the optimal approach to drive cooling is to put a cooling plate on either the top or the bottom of the drive, ideally both.

In evaluating the demands for cooling a drive, I found that there isn't a huge amount of heat involved, so there isn't the need for the level of intensity and perfection that a CPU demands, so it should be possible to use simpler construction methods than for a CPU waterblock.

I am perfecting a design consisting of one or two peices of 1/8" copper plate, with a cooling loop made of 3/8" OD, 1/4" ID, copper tubing soldered to it. I put a plate on both sides of the copper loop when making a cooler for two drives, or put a plate on one side only if I'm making a cooler for just one drive.(I could also use smaller tubing.)

I plan to be doing an article which I'll be posting on how I make these, and certainly wouldn't have a problem with someone making their own. However I am well aware that not everyone is either mechanically inclined or as much of a tool freak as I am; so at least some folks might prefer to buy a block rather than make their own.

I could stand to pick up some spare money making blocks, but if I was going to do very many, would want to make some fancier jig fixturing to make building them easier.

So I need to get a feel for the market - How many of you would be interested in purchasing drive blocks if I was to make them, and how much would you be willing to pay? I figure that to be worth while, I'd need to get at least $35 for the single drive type and $50 for the double drive type.

PLEASE NOTE: I AM NOT SELLING THESE AT THIS TIME, JUST TRYING TO FIGURE OUT IF I SHOULD!

Gooserider

[bigben2k]

I'm making mine:
http://forums.procooling.com/vbb/sho...ight=hdd+block

As a marketing survey, you'd want a sample of your target audience. The problem with a Forum inquiry, is that you're not allowed to advertise a commercial product in them, unless it's to discuss the working principle. (hint hint). So ideally, your sampling should be on your target audience, the one that's going to be the recipient of your marketing campaign.

[Gooserider]

as these aren't that much fun to make, just trying to find out if there would be enough people wanting them to justify spending the money on better jig fixtures than what I'm currently using. (which would also improve the quality of the product)

bigben2k: I'm making mine: http://forums.procooling.com/vbb/sh...light=hdd+block

Thanks for the pointer - I may do a repeat of my question there, with more technical info.

As a marketing survey, you'd want a sample of your target audience. The problem with a Forum inquiry, is that you're not allowed to advertise a commercial product in them, unless it's to discuss the working principle. (hint hint).

Agreed, that's why I tried to make clear that I wasn't attempting to sell in that venue, just finding out if there would be any interest.

My thought was that people reading a forum on WC'ing would be as likely a target audience as any. If I got several enthusiastic responses then that would be a pretty good sign of it being worth while to push things further. If I got no response then it would be a sign that there isn't enough demand.

I tried to give a bit of the working principle in my post but didn't want to make it TO long, or get away from the main subject of finding out if there was any interest.

So ideally, your sampling should be on your target audience, the one that's going to be the recipient of your marketing campaign.

Well, I would hope to sell at least some of these through the FS/FT forum if I had interest - If the people here aren't likely customers, I don't know who would be!

Gooserider

[sevisehda]

50$ is kinda steep. My dual HD water cooler cost me less than 10$ in parts.

[bigben2k]

Originally posted by Gooserider
Well, I would hope to sell at least some of these through the FS/FT forum if I had interest - If the people here aren't likely customers, I don't know who would be!
Gooserider

A lot of Forums don't allow commercialisation of products through the FS/FT forums, but I think that as long as the quantity is small, you should be able to get away with it.

[Blackeagle]

The 10-15K drives need more cooling for sure, but I'm not sure they require water cooling. A simple finned sink and steady air flow should be enough.

And at much lower cost.

To me the only one I've seen (water cooler for drives) of interest to me would be the german made Evolution I which covers both of the sides and the top as well. And it's strongest point of interest isn't it's cooling of the drive, but it's very fine eye candy loods. It's about the finest eye candy I've seen. But it's cost makes your ideas look like a VERY low cost option.

[Gooserider]

sevisehda: 50$ is kinda steep. My dual HD water cooler cost me less than 10$ in parts.

I agree it isn't cheap, but then neither are commercial CPU blocks...

I haven't worked out my exact material costs, but the material isn't that expensive. What is expensive is that it takes me a couple of hours to make each one, and involves a significant amount of use of expensive tools. (And if you didn't have the tools it would take much longer... For instance I use a $100 SawzAll to rough cut the plates, which takes about 5 - 10 minutes per plate. You could do it with a $5 hacksaw, but it would probably take about 2 hours per plate)

A mechanically skilled person with a good shop full of tools (or access to same) like you and I, are not the people I see as potential customers. We make our own stuff for the pride of it, and because we CAN... I'm more looking for the person who buys a cooling kit of commercial parts because he either doesn't have the tools to build his own, or because he wouldn't know how even if he had them.

My girlfriend would be an extreme example of this. She is a Principle Software Engineer (one notch less than Architect) for a very large high tech company. She can write code in just about any language, on any platform, using any O/S. But give her a screwdriver, and she's a danger to herself and others. (I won't let her near my power tools) The only way she'd ever get a water-cooled PC is if she could purchase one, or have one built for her.

The potential customer would be a person with enough clue to build their own system, but not enough tools / know how to build their own cooling parts for it.

Blackeagle: The 10-15K drives need more cooling for sure, but I'm not sure they require water cooling. A simple finned sink and steady air flow should be enough. And at much lower cost.

Depends on how you define it. I would agree they don't REQUIRE water cooling, but then nothing else in our PC's does either if you get right down to it. (Certainly there's no need for pelts, vapor-chills, and other extreme cooling!) I would never consider putting in a WC system just for the drives.

But to do the finned sink and airflow method implies more fans. Many of us went to WC in order to get rid of fans and make our systems quiet; putting fans back to cool drives isn't where we are at! Putting a WC block on the drive gives cooling without fan noises.

In addition, the drives themselves are reportedly noisy, adding heat sinks will amplify the noise (resonance and 'fin ringing') and having to stick the drives in the airflow means the noise has lots of escape routes to get out of the case and into your ears. Liquid cooling allows one to tuck the drives away where they can't be heard, and possibly even totally seal them in sound deadening foam...

In short Blackeagle, you are partially right, and partly wrong; The drive doesn't require watercooling, *I* require the drive to BE watercooled!

ART

[MadDogMe]

HD's don't realy get that hot AFAIK. It's the chips on the small PCB that reach a temp that makes the case air around the HD warm. I'd concentrate on them before even thinking about the HD body. The only other heatsource is the motor/spindle area which does'nt get any warmer on my drive than the rest of the body, and that's probably from the heat radiated from the PCB! ...

[Pritorian]

Koolance had a good cooler for the HDD, but was alu, and very restrictive.

It gooped over all the components ( bye bye, warranty ) and you attached a cooler to the same side..

Hmmm, linky maybe...

here : linky

[bigben2k]

Interesting.

GMat posted this bit of info, about "top cooling" an HDD:

Sides are flat, and closer to bottom (and are the same metal part as the bottom).
Cooling a PCB with various components with *one* waterblock (or heatsink), without shorting anything but touching the most things at the same time is ... tough.
So indeed the sides are the hottest *usable* surfaces for cooling.

With top cooling, I also understood that the motor spindle's heat would have to travel up through the PCB, to be cooled by a top cooler, where a side cooler will capture it.

No HDD "requires" water cooling, and in fact, if the water temp is too high, it'll make things worse. But those 10-15K scsi units sure can benefit from it.

Anyone can get an idea about how hot their HDD's are going to run, by checking out www.storagereview.com .

[Gooserider]

Per the very long, and somewhat dated but still informative article on all about the innards of hard drives, on that site; by FAR the biggest single source of heat on a drive is the motor and platter assembly!

The heat comes from spinning the platters up against the friction of the air in the drive housing, and the heat is dissipated to all the metal surfaces in contact with that air. This includes the sides, the lid, the body under the PCB, etc. Since the heat is spread over such a large area, none of the areas feel like they are getting all that hot, but the total is still pretty high.

Note also that most of the heat is spread by convection, not conduction. This should be fairly intuitive given that big air-mixer inside the drive housing.

Supposedly, putting a cold spot anywhere on the drive will cool the entire drive effectively. The heat that is being generated and transferred to the case by convection will mostly transfer to the cooled area (by heat flow laws) and the hotter parts of the case will eventually xfer their heat to the cooler area by conduction. I suspect that if the heat coming off the drive body is removed, the chips on the PC board will come back down to reasonable temps.

Another interesting data point - because drives can be face mounted on the PCB side, it is a design requirement that the drive body ledge around the outside of the drive be HIGHER than the tallest component on the PCB. (some drives also leave the motor housing at the same level as the ledge, it's easier to mill it that way, and MIGHT provide extra cooling) My former employer used drive arrays with hot-swappable housings holding the drives. The drives were face mounted, with NO air circulation on the PCB side, but forced air flow over the lid (mostly) and sides (a bit). This is a fairly standard design for SCA based hot pluggable arrays, one of the prime users of the high RPM drives, and they don't overheat in that application...

Gooserider

[Brad]

if you could make one that would work as a 5.25" to 3.5" converter, not take up any more bays, and actually cool the drive ok, then you'd have a very good market

[Gooserider]

Brad: if you could make one that would work as a 5.25" to 3.5" converter, not take up any more bays, and actually cool the drive ok, then you'd have a very good market

Well that would be possible, but not necessarily with top/bottom cooling. There is considerable variation in the height of a 3.5" HDD. some low profile drives could have a top/bottom cooler attached and still be less than a bay high. Some of the higher profile drives push bay heights as it is, so no way. I wanted to stick with the ability to use 3.5" bays and wasn't worried about height that much, so I concentrated on a design that didn't try to do the conversion.

If one wanted to do conversion, the simplest approach would be to go with side cooling, and make two blocks w/ lids per drive, each the height and length of a bay, The thickness of each block and lid would be equal to half the difference between the 3.5" & 5.25" bay widths. I would drill counterbored holes at the proper spacing to fit the drive, so as to put the drive bottom even with the bottom of the block. Then I would mill a passage into the block, using most of its volume. Probably a 'U' shape, about 1/4" wide, with the points of the 'U' at the back where I would put two barbs. One barb would crossconnect the two blocks, the other would be for I/O. Alternatively, I might mill a single straight passage (possibly with some wiggles) and put the crossover at the front. I/O would have to be at the back with the other cabling.

This is just off the top of my head, I would need to sit down and work out the details, however I'm certain it would be no problem to meet the stated design objective. Of course, unless I came up with some sort of bracket, or did some other fancier construction techniques it would be pretty hefty if I made it in copper, and expensive just on account of the materials cost, not to mention what a pain it is to mill copper. But all such problems are solvable with enough effort.

Gooserider