Mac Pro RAID Enclosure - Need Swivel Mounting Solution

[dsa420]

So my Mac Pro RAID enclosure project is almost complete, just working through the last few steps. Specifically, I need a solution for mounting the 2x 5 Bay backplanes to the bottom of the "shelve" where the Optical Drive(s) and PSU is typically mounted.

I have 2 of these:
http://www.newegg.com/Product/Product.aspx?Item=N82E16816215001

sitting freely on the floor of my Mac Pro case and I want to mount them to the underside of the shelve that generally houses the optical drives and PSU in a MP. Ideally, I will find something that can swivel so I can turn the backplanes 360 degrees when the side door of the case is removed.

Any ideas?

 

[nanofrog]

A small swivel plate ~ 4 - 5" square used in a Lazy Susan (examples) from something like Home Depot might work for you. It should be small enough to mount to the backplane and the internal area available to you.

Hope this helps.

 

[hyram]

Could you provide a pic or 2??? Might help us see exactly what you're trying to do.

 

[Tesselator]

Nano's idea might be alright, I dunno, they're ball-barrings and pretty loose typically, plus they're meant to bare a load not be pulled on by something hanging - so I dunno how smooth they would be.

It's easy to make your own tho. It's just two pieces of plastic (like your mom uses for a cutting-board in the kitchen) with a hole drilled through them and a countersink on one hole for the bolt head. If you have room on the upper part of the "shelf" you can use a wing-nut on the shaft that sticks up through so you can easily adjust tension.

I think I would rather use slides myself. More like the MP itself incorporates for the drive sleds. Again easy to make out of wood or plastic.

 

[nanofrog]

Nano's idea might be alright, I dunno, they're ball-barrings and pretty loose typically, plus they're meant to bare a load not be pulled on by something hanging - so I dunno how smooth they would be.

No, they're not meant to have a load pulling on them, but I'd think it could withstand the backplane and drives though.

Another issue that could come up that I didn't mention, even if the plates are tight, it could still vibrate enough to make noise loud enough to be a nuisance. The reason being, is that the swivel plate is the only thing it's mounted to (no secondary mounting points to prevent oscillations).

A DIY solution could help to mitigate this, but would still likely remain an issue (the backplane and drives are fairly heavy once assembled together), assuming it is a swivel plate method.

 

[dsa420]

Thanks for the ideas guys, I tried the lazy susan idea and after hours of attempts, I just couldnt seem to make it fit properly. Back to the drawing board regarding options for my set up...

The fans on the backplanes are louder than I anticipated, would it be possible to employ a water cooling solution for the 2 backplanes? There is no mother board, and it appears all stock solutions require a the mother board as a prerequisite.

 

[nanofrog]

The fans on the backplanes are louder than I anticipated, would it be possible to employ a water cooling solution for the 2 backplanes? There is no mother board, and it appears all stock solutions require a the mother board as a prerequisite.

Those small fans are always on the loud side. Few are quiet, and if they are, have larger fans in them. Some enclosures have opted to do their own cooling with typically an 80mm size used, though I've seen at least one that used a 120mm. Even a couple of backplane units have opted for larger fans, but they're not nearly as common.

I'm not sure what you want to do with water cooling, as I don't think you'll be able to apply it to the backplane enclosure. Just not enough physical space I'd imagine, as they're made tight to get the max drive quantity in a small space (less wasteful than 1-3.5" drive per 5.25" bay).

Nor am I sure as to what you mean by a Mother board. If you could explain, that would help.

BTW, the data connections on the back of the backplane units are typically attached to a RAID card via what ever cable connector you need, usually an SFF-8087 to 4i*SATA for internal ports on the card, or an SFF-8088 end if the card's port is external (harder to find, but not impossible, as they typically they have to be ordered). But they don't have to be connected directly to a motherboard.

 

[dsa420]

Thanks for the advice as always.... Clearly I am not very well versed in the world of water cooling.

My set up is already up and running, although I am not happy with the layout inside the machine. I have 2 external miniSaS to 4 SATA fanout cables running from my Mac Pro directly into the backplane.

Would love to find a plate for the back of the "enclosure" the can accept 4 SATA connections that has 4 SATA outputs on the other end.

Additionally, I really want to silence the fan noise if at all possible as well as obviously keeping the drives cool. The installed fan is an 80mm fan, are those "silent" fans worthwhile (actually quiet and keep the drives cool)?

 

[Tesselator]

No, they're not meant to have a load pulling on them, but I'd think it could withstand the backplane and drives though.

Yeah, I didn't mean they couldn't withstand the forces applied. If I understand the typical design of those things they work because each metal plate has a race that is in contact with the opposing sides of a number of metal balls (or rollers). This design needs force (usually gravity) for the plate-races to remain in contact with the balls. So I questioned how well it would work with that design intent unemployed.

 

[nanofrog]

My set up is already up and running, although I am not happy with the layout inside the machine. I have 2 external miniSaS to 4 SATA fanout cables running from my Mac Pro directly into the backplane.

It may not be pretty, but it's the best way to go with SATA. Seriously.

Would love to find a plate for the back of the "enclosure" the can accept 4 SATA connections that has 4 SATA outputs on the other end.

You don't want to use these. SATA has a short cable length (2.0 for active signals, and 1.0m for passive signals). Your setup is passive BTW, as there's no active circuits between the card and drives, such as a Port Multiplier board which has a power connection to the enclosure's PSU (hope this makes sense).

There's also the issue with the adapters creating issues themselves, even if it's all kept within 1.0m, due to contact resistance, and the introduction of additional noise (poorer shielding than a direct cable as you have now, as there's additional points for entry = data instability). In short, you end up with drop outs.

Additionally, I really want to silence the fan noise if at all possible as well as obviously keeping the drives cool. The installed fan is an 80mm fan, are those "silent" fans worthwhile (actually quiet and keep the drives cool)?

You can swap the fan, assuming you can find one with sufficient specs (meets the airflow requirements with a lower noise rating). It's been done before, but in the enterprise environement, noise is a distant consideration to cooling. Rack rooms are noisy as heck. Fans, HVAC system, inverters (if the backup system resides in the same section as the racks), generators running when running in a power out occurs,....

Yeah, I didn't mean they couldn't withstand the forces applied. If I understand the typical design of those things they work because each metal plate has a race that is in contact with the opposing sides of a number of metal balls (or rollers). This design needs force (usually gravity) for the plate-races to remain in contact with the balls. So I questioned how well it would work with that design intent unemployed.

I understand. Some are beyond cheaply made, while others would seem to be able to take some sort of load (though deminished) in an inverted installation.

Think cheap barstools vs. the really heavy duty casters used in portable racks, workbenches,... As I think about it, even some of the montior mounts have used an inverted mounting IIRC (i.e. ceiling mounts for CRT's that supported over 100lbs, and the pin didn't go through the bottom plate <pass through> keeping the load on the bearings by gravity on the upper plate).