CCD water cooling advise

[rogerg]

Hi all,

I am wanting to fit a closed loop water cooling system to my new SBIG when it arrives. Anyone have experience and recommendations?

This is the knd of thing I a looking to build:

Http://www.tuvaclub.org/cooler.html

Suggestions of parts suppliers and kits?

Thanks,
Roger.

[mswhin63]

I am currently working on cooling device for my DSLR. I considered water cooling but decided to stick with Peltier.
The reason is so much gear on a moving device (telescope mount etc), the weight is just too much and installing the water tank elsewhere requires some very flexible and long pipes. As the pipes extend it requires more pressure to keep the flow of water going. Overall power consumption for water cooling and Peltier is close.

I have already selected a heat-sink but haven't purchased it yet. Too many other projects trying to complete now.

[mswhin63]

If you really need to try water cooling then the best place are CPU coolers. Quite readily available from most computer stores.

[Barrykgerdes]

Simple water cooling a chip would be a lot more expensive and bothersome to implement rather than a peltier chip. However cooling a peltier with water might be a way to get lower temperatures.

Barry

[rogerg]

In an attempt to keep the thread on-track....

Sorry if I wasn't clear in my initial post, but I am talking about an SBIG which already has peltier cooling and water cooling capability but for which I need pump, radiator, etc. I'm not interested in attaching additional peltiers to the SBIG or such.

In summer my current SBIG gets down to -5 or zero at best. I'm looking to extend that.

Thanks,
Roger.

PS. Malcolm, you always have too many projects on the go it seems... don't know what you're complaining about :)

[pmrid]

I'm with Malcolm on this. I have a Thermaltake Pro Water 850i Liquid CPU cooling kit waiting to get installed on my big ATIK11002M. It contains all the items in that article referenced by the original poster (pump, reservoir, radiator, fan etc, plus what looks like reasonably long tuning etc. It cost A$157 from UMart but it won't require any additonal fittings, adapters, messing around etc. I quickly priced the main items that article rererred to - and there would have been no change out of at least A$200 plus - and that's without the 5 adapters he mentioned, or the tubing. So all-in-all, I think the integrated CPU cooler kit wins hands down.
Peter

[mswhin63]

If you need extra cooling, I was even considering running Peltier on Peltier. I was going to experiment on spacings and varying heat-sink methods including a simple 5mm aluminium plate separating the Peltier wafers.
But I had considered water cooling and if the Peltier to Peltier is not efficient enough then I would consider water cooling and subsequently would only go for a after-market system. CPU's run close to if not higher heat so it should work well. I would still be concern over extending cooling pipe to take in mount movement. An experience was a pond pump i set-up when I had a pond, I experienced some issues with pond pumps and flow. I am not fully experience with fluid dynamic calculations though so there may not be a major issue :shrug:


:lol: can be a bit overwhelming.

[marki]

Roger SBIG can supply pumps for their camera's I have one sitting around here that I don't use. Water cooling is PITA in terms of extra lines and gear spread all over the place. You also need to be very very careful to keep the water temp above the dew point or you could cause condensation to form inside the camera and electronics have been known to fry under these conditions. The advantage gained is quite small and when compared to all the rest of the hassles that come with it and I am not a big fan. Better off buying a camera with a decent peltier and use air cooling. You have not mentioned which model camera you have bought so it is difficult to say anything of real value.

Mark

[rogerg]

I'm not keen on water sitting around the observatory in a bucket or such, or long lengths of water tubing. The SBIG pump solution seems like a bare minimum in that respect really, not a neat solution. I much prefer the idea of a closed loop system where an air cooled radiator keeps the water at/near ambient (but doesn't go below ambient, because at best its being cooled by ambient air). It's this kind of solution I was thinking of.

There are stats out there suggesting a cooling improvement of 10-15 degrees which sounds worthwhile when the camera is only at zero degrees. I'm interested to try it, but I completely see yours (and others who have PM'd me) point that it ends up being more trouble than it's worth and wouldn't be surprised I put it in that category too.

Thanks,
Roger.

[rogerg]

Thanks for the product recommendation - it looks exactly like the kind of kit I'm looking for. Closed loop with radiator, and the same price as SBIG's pump just on its own.

The only thing I wonder about that kit is if the pump will be strong enough to pump the water around a large loop of tube, say 3-4m, to get from pier up to OTA and back.

[mswhin63]

This will depend on the specs of the motor. I have seen not with cooling but with other systems that computer hardware is engineered so tight that just the smallest change in the system would destroy the system. This is a market that needs to save money. The pump could be replaced but with what I am not sure (maybe a pond pump). Most water pumps use a pulsing compression which create vibration in the water, this could transfer to the camera. There are some dampening systems out there but the cost of the solutions would be very high.

The concept of water cooling or even Freon cooling is flawless but the practicalities I feel are concerning and I personally could not spend the money to find out.

[marki]

Thats your other problem. The only way I could get a pump powerful enough to push the water the required distance was to use a pond pump I bought from a garden and Koi fish shop in Balcatta. It was meant to power a water fall/fountain in fish ponds. It was quite large and even this struggled with the task as I had to reduce the line size to make the pipes small enough to attach to the camera. This meant sleeving the outlet down from about 1 inch dia to 1/4 inch. Not good practice as most of the water simply cavitates around the impeller . The sbig pump could not lift the water from ground level to the top of my pier (it is too high, need to shorten the pier) and I had to lift the water reservoir up on a stand to make it work, very messy. In the end I used to attach the hose to the pond pump and run it at a low pressure with the spent water being diverted to the garden. I was never able to achieve a 10 degree drop in temp, best I could do was about 4 - 5 degrees. This may have been due to the heat added by the big pump but I was very disappointed with the results vs money spent and ease of use. I did consider using a peltier cooled heat exchanger between the pump and camera and made a crude model radiator using ice in place of the peltier as a test bed. Everything just fogged up so I forgot the whole idea. I think if I found myself in the same position again I would work to improve the heatsink (make it bigger) so I could attach a larger more effiecent fan and a second peltier or even experiment with chilled air or gas rather then liquids.

Mark

[pmrid]

My instinct (without having tried it yet) is that you might begin with the pump etc close to the height of the CCD but when the flow is established, the syphon effect would tend to keep it going so you could then lower the pump etc to somewhere on the mount of whatever. That's my untried hope. Note I didn't say 'expectation'. As with all things DIY, a certain amount of 'suck it and see' is usually involved.
Peter

[Terry B]

This exactly what happens. I use a sbig pump with a 20l drum of water. The tubing drapes from the camera over the finder scope mount( to stop any tension on the camera) back to the drum at the pier base.
I connect it up and turn it on. I have to lift the drum to establish flow but can then leave it on the ground and it will flow with no problems.
It works very well and gives me much better cooling than no water. I change the water about once a year.
I don't find it difficult to use at all.

[Bassnut]

Aqaurium/pond pumps are crap, ive tried em all. Peristaltic pumps work.

I cool my SBIG till its dripping wet, stuff it, suffer for art, with a bucket of iced water. Bad idea according to this thread , read it, its good http://www.iceinspace.com.au/forum/showthread.php?t=81592

Its still going , so ill keep torturing it.

The tubes etc arnt a prob with a bit of management.

[Peter Ward]

You know they have tropical fish tank coolers.... small capacity ones for around $400-500

;)

[DavidTrap]

+1 Terry,

The siphon effect works as you and Peter describe. I bough a small 12volt impeller pump off eBay for $~20 and a $~10 variable power controller to turn down the pump speed to a trickle. I needed a 12v setup to run out a Leyburn.

I started using water cooling this summer. I can get -15 degrees with my QSI from ~+25 ambient, with the cooler operating ~90%. I use 1/4" tubing and just drag the two lines up and over the mount. There's enough other data and power cables hanging off the mount it already, so why worry about a couple more.

The flow rates are so low, you could probably get away with finer tubing if you're worried about the weight of tubing hanging on the mount.

DT