Beginning to test.

[Stewie007]

7F.... That is the lowest value of the day.

There are some spots I need to insulate a bit better. However, flow gets a bit low when it gets down that far. I suspect a bit of slush may have been forming during the off cycle because I turned the pump off for a bit. Either that or it formed at some point during operation. Or it might just be the viscosity of the solution. Perhaps it was just because the liquid was so thick that it didn't move much once discharged... LOL

[Buckeye]

What is normal pull down time for most? I'm on 2 hours from 65F. Its down to ~15F now. 25 degrees per hour. Not bad? I just don't have a comparison to benchmark against, hehehe.

I am at a loss as to what you will be using this for if it takes 2 hours to get this rolling. Yes I get it's for a PC but do you plan on leaving this on for very long periods of time ?

[Stewie007]

Well, it might be a flow issue. I ran it through a pretty restrictive coil to test. It took longer than usual. Guess you arent so eager to see it work anymore huh hehe

Still working out the kinks. Flow was dreadfully low. Have to find out why. It pulled down to 8 degrees far faster the first time.

[Buckeye]

Of course I am Stewie, I love these things

[n00b 0f l337]

Flow was probably from restriction and fluid becoming very viscous.

[Stewie007]

Partly, but it wasnt as big of a problem the first time around. So, there was something I did a bit differently. At 20 degrees viscosity should be around 50 cps... at 8F it would be around 65 to 70. Olive oil is around 85 and this pump would handle that,

So, I agree that restriction is the likely culprit. But I am unsure why it differed the first time around.

[Stewie007]

This sort of changes the whole dimension of the project. Restriction is proving to be a problem. Water blocks are, by design, are restrictive.

What type of block,then, do we utilize for this to be effective?

[Stewie007]

http://www.enzotechnology.com/vga_ib.htm#

This one was recommended by someone a while ago, and I also like it. Looks like it isn't very restrictive same with their CPU blocks. The problem is, do you think it will fit a 7900 Radeon or a 290/280?

http://www.enzotechnology.com/scw_reva.htm#

I like this as well. Looks like I might be able to back order it... Its AM2... but isn't that pretty much the same socket retention mechanism as the AM3/3+?

[n00b 0f l337]

How little flow are you getting?

[Stewie007]

Alarmingly little. The pump suction appears adequate, and the outlet is pushing pressure out. However its so low that the reservoir appeared to be still. It screams blockage to me. I am going to negate the coil next week and see if that makes a difference.

[n00b 0f l337]

Coil could have frozen up?

[Stewie007]

It shouldn't have. Its a 50/50 Glycol mix. But if it got slushy, that would plug the restrictive coil. I am going to test with a wide open system next week, but when its finally running with a computer... I'll have a strainer inline.

The problem is, I don't know if I turned off the pump when it was below freezing or not. I think that I did. And that probably caused my problem. But I can't verify it. We'll see next week.

[Stewie007]

I wonder if I overcharged the system. What is the normal amount of frost that should form on the accumulator? Can I assume that with a return gas temp of 0F (even with superheat) that its pretty normal for the accumulator to have a nice layer of snow on it??

The biggest indicator to me, though, is that if I have the TXV adjusted up just slightly, the start up pressure is ridiculous and I have to shut it off to protect the compressor. If I adjust the TXV down a bit, it alleviates the condition at start up and superheat appeared to be normal. Later in system operation, the pressure will fall to a more acceptable level.

Is this just because of the higher load at start up? OR do you think I'm overcharged?

Edit: I'm just gonna have to reset the TXV setting and see. Its obviously off because I didn't know what I was doing at the start.

Also, I wonder if this GPU block will fit most modern cards? Seems to be a perfect choice.

http://koolance.com/gpu-200-video-ca...et-water-block

[Stewie007]

Got some results to post later

[Stewie007]

Apparently my favorite phrase of the day is ".....to be expected...." LOL. I should make the videos more time neutral. Doesn't make much sense to make a video and then post it 5 hours later as if I just made it

Anyways... I've done some testing and I think the system is pretty much ready for the enclosure. It will be a few weeks before I can take care of that given the holidays. Still not where it needs to be reservoir temp wise. Given the difficulty keeping flow rates up that is the main reason why the temperature change slows to a virtual crawl. Its still pumping pretty fast at that point, I confirmed that by shutting the valve and watched the bubbles flow. Needless to say they were moving fast. So maybe not such a big deal. Anyways, first video... then second one in a few.

[Stewie007]

[Stewie007]

I used my temperature probe to get a shot of the outlet temp from the chiller. This is the coldest temp I took, and it was right before I had to close up shop for the day. That is about an 7* Delta T. If I ran it for long enough, I could hit 5F, but considering that I'm hitting 10 in a little less than 1.5 hours time from 65F or so that's not bad. I'm going to do some additional tweaking the week after next (won't be here next weekend). However, in all likelihood I can probably add a booster pump to help increase the head pressure capabilities in order to boost the flow through the evaporator.

That said, it doesn't take as long to get the temps down after first run. so I'm not too concerned.

One concern that I have is that for some reason, the solenoid remains party energized after the system is deactivated. It causes a pressure bleed that won't allow the system to stay off. Very concerning....

[Buckeye]

Are you still using that copper coil being heated by a camp stove as a load to test ?

That's a fair bit hotter then a CPU I would think so it should cool down much faster with a normal 300watt load.

[Stewie007]

No load, just pull down. The stove was for a known heat load and it served its purpose hehe

[Stewie007]

I sm going to do alot of folding type stuff though, so that will produce some heat. But thats no worry... I am getting 100 degree condensing temps. So we're talking more capacity than the chips can put out.

Rough estimate at 1500+ watts based on specs of the compressor. Actual unknown, but I ran some calculations on that data and that gave me a general idea.

Edit: I will probably have to remove some charge from the unit. I may have overdone it. Still, with temps down to ~5F that isn't that bad. Of course, that's after the chiller and not in the reservoir. The viscosity of PPG is just not going to allow the big flow rates. Ethylene Glycol would have been better, but I chose PPG for a reason. It does put out high pressure.

[NEOAethyr]

I was thinking of these chilled water loops as well the last few days.

I just thought about it again today, wondering if 180 proof vodka would work well as the fluid.
Instead of water and other things to keep ec down.
One can't legally distill, but you could always buy something close enough to 100% .
Thought about acetone and methanol but both are a hazard, while ethanol isn't so much (still flammable though).
Better then probably hydrogen peroxide or rubbing alcohol.
Hydrogen peroxide would oxidize stuff as soon as it hits silver probably.

I found tubing that handles sub zero, but pumps are a problem.
I don't really wanna use a metal pump either.


Edit:

I just found this it was interesting:
http://www.overclockers.com/data-on-...-watercooling/

Sorry I don't mean to derail your thread dude.
It's just something I just became interested in too .

Windshield wiper fliud I think uses 10% methanol, I would never use methanol though.

[Stewie007]

Well, I was figuring that I might have this thing around for a while and if a little munchkin comes into the picture, I don't want any accidents to turn fatal. PPG is a pretty reliable fluid, though.

I got a good idea of how well the flow was going when my time delay kicked in and the slightly warmer water in the coil above discharged into the reservoir. It was by chance that I took the makeshift cover off to see a wave of warmer water hit the cold solution. It dispersed into the water quite quickly. So I think my pump is getting decent enough flow. Its just tough to get that temperature change once you're temperature differential sinks so low. I may have to recover some charge so that I can close the valve up a little bit to get the pressure, and thus the temperature inside the evaporator down.

I have been dancing around the option of adding a booster pump after the evaporator. I have to do some research on that first, though. It is a no no to run two dissimilar pumps in parallel, but not so much to run them in series. I just have to figure out what type of difference is acceptable. It would help to have a second pump adding flow to help turn over the temp in the reservoir more quickly. The system would be allowed to continue down to the 0F in a more expedient manner.

One thing I can say about these Ryton volutes... they insulate well. Won't have to add much insulation to the pump for condensation. It barely condenses moisture at 10F fluid temperature.

[Stewie007]

Well, some major changes to report. Although I haven't done much lately, I have learned that Polypropylene will be a superior choice to Ryton. Therefore, I will be acquiring a Polypropylene Volute for the existing pump, and the second pump (should it be needed) will have the Polypropylene housing. Big Flub-up on my part. It'll be fine for testing, but once that's complete, the volute will be replaced.

The engineer I spoke to with the manufacturer advised that Ryton becomes brittle at low temperatures. I could use it, but I would have to be absolutely sure that the plumbing doesn't apply stress to the volute.

I was able to find a TE-5-MD Pump motor on Ebay with the motor drive magnet. All I need is the pump head. The MD-SC variety is about $140 bucks. So if the second pump is needed, I've got a source for an identical pump.