A/C to Chiller conversion

[markdm]

hey drew, hows it going tonight, you getting anywhere on the OC'ing?

BTW did you have your mag inline when the volute cracked?

[madramper]

So how much money do you have invested in the chiller at this point?

When are you gonna start mass-producing these systems??


With the thermostat on there, how well could you control temps? I REALLY don't want to condensate proof my board, so sub zero temps are out of the question. Would you be able to keep a steady 10-15c by setting the thermostat?

[Drewmeister]

Originally Posted by markdm

hey drew, hows it going tonight, you getting anywhere on the OC'ing?

BTW did you have your mag inline when the volute cracked?

I was running it submerged for a week or two but noticed it was hurting the temps by 4-6c so I decided to go inline. When I was changing the ftgs around that's when I noticed the cracks. With my mag7 all I did was (lightly) bump the discharge line and the whole threaded part of the volute cracked off. They just become very brittle over time. I spent many hours patching them up but could never completely stop the leaking. Two of the mounting studs even cracked off(I have a fix for that). I have a complete photo shoot on how to fix the mags. I'm looking into encasing the volute in epoxy. I've found some that is rated down to around -70c. I'll keep you updated on that.

One thing to keep in mind when running a pump inline is to make sure you start it up pumping well above freezing to avoid stressing it from the rapid temperature change.

I've been trying different bios's but seem to be stuck at 230x13=2990 and 260x11=2860. I broke 7K on 2K3 at 260 FSB. The 12 and 12.5 multis just won't cooperate. I'm almost positive the board is limiting me but have no plans on investing anymore money into the socket A platform. Colder temps should get me 3GHz stable but I'm completely statisfied with 2990. It's just about time for the P4.

[markdm]

Cool, keep us posted on your progress.

Also, i was thinking about the submerged pump heat problem. what if you used a small platform, (or screwed the base plate to the wall of the tank) to set the pump on inside the tank so its not submerged and just ran the line with a filter on it to the bottom, set it in there in a way that you have the least amount of stress on the fittings, and insulate the hell out of the pump. That way it wouldn't be subjected to the liquid, just the cold temps, while the upper part would be slightly warmer (air wise), the bottom liquid should be able to hold the temps better and should help in not loosing the liquid temps to the heat from the pump. What do you think?

Im thinking the cracking on the inline use is due to the extreme temp difference from the liquid to the outside temp of the pump. with the pump inside at a lower temp, (but not in the liquid) it might help stop the cracking on the inlet, etc...

[Drewmeister]

Originally Posted by madramper

So how much money do you have invested in the chiller at this point?

When are you gonna start mass-producing these systems??


With the thermostat on there, how well could you control temps? I REALLY don't want to condensate proof my board, so sub zero temps are out of the question. Would you be able to keep a steady 10-15c by setting the thermostat?

Not sure how much I have into mine b/c of all the testing I've done with various waterblocks and pumps. Basically one could build a (no bells and whistles) complete system for $240-$300.

I will mass produce after you guys get a look at my next chiller.

If you only need 10-15c an A/C chiller would be way overkill. I can build you a micro chiller for that smaller and quieter than a prommy. I have digital thermostats adjustable from as low as -34c and up with a set point range differential of just a few degrees to keep you within 10-15c np.

[noremacyug]

Originally Posted by Drewmeister

If you only need 10-15c an A/C chiller would be way overkill. I can build you a micro chiller for that smaller and quieter than a prommy. I have digital thermostats adjustable from as low as -34c and up with a set point range differential of just a few degrees to keep you within 10-15c np.

also interested in that if the price is right. i dont want to have any condesation but would like to hold 10-15C. im just now fixing to build my first water cooled rig. so im gonna see how it does but let me know what comes of this.

[Drewmeister]

Originally Posted by markdm

Cool, keep us posted on your progress.

Also, i was thinking about the submerged pump heat problem. what if you used a small platform, (or screwed the base plate to the wall of the tank) to set the pump on inside the tank so its not submerged and just ran the line with a filter on it to the bottom, set it in there in a way that you have the least amount of stress on the fittings, and insulate the hell out of the pump. That way it wouldn't be subjected to the liquid, just the cold temps, while the upper part would be slightly warmer (air wise), the bottom liquid should be able to hold the temps better and should help in not loosing the liquid temps to the heat from the pump. What do you think?

Im thinking the cracking on the inline use is due to the extreme temp difference from the liquid to the outside temp of the pump. with the pump inside at a lower temp, (but not in the liquid) it might help stop the cracking on the inlet, etc...

Sounds good but you can't insulate the body of the pump. It gets way to hot and will damaged the pump and possibly start a fire. The heat would also eventually find it's way back into the fluid. Only insulate the volute and maybe 1/2"-1" onto the pump body. Do what you said and mount it to the wall but above the fluid level and seal off the pump from the fluid. That's kinda how I ran mine for a while and went from -22c submerged to -26 to -30c inline.

[markdm]

Ya, i just felt the mag3 in my water system and see what you mean about the heat, i didn't think it got that warm

So insulating the volute and maybe the face of the pump is the only way to do it. I was thinking that warm air rises and therefor would look for a way out thru the minute leaks in the lid of the tank. Still, not putting it in the liquid has to help some with the temps, as you said, it increased yours by 4-6 C

I have been playing with this pelt now to, testing different volts. I have read the meny articals about how you shouldn't under power a pelt, but so far this one seems to work its best at 9.2v @ ~14amps. with ambient @25C i get a Delta of about 36C. Im trying 11v now to see if i can do any better.

[Drewmeister]

Originally Posted by markdm

Ya, i just felt the mag3 in my water system and see what you mean about the heat, i didn't think it got that warm

So insulating the volute and maybe the face of the pump is the only way to do it. I was thinking that warm air rises and therefor would look for a way out thru the minute leaks in the lid of the tank. Still, not putting it in the liquid has to help some with the temps, as you said, it increased yours by 4-6 C

I have been playing with this pelt now to, testing different volts. I have read the meny articals about how you shouldn't under power a pelt, but so far this one seems to work its best at 9.2v @ ~14amps. with ambient @25C i get a Delta of about 36C. Im trying 11v now to see if i can do any better.

Some of the heat will escape but most of it will get sucked up by the extremely cold liquid. Maybe still try to seal it under the pump a little to reduce the heat absorbtion some.

[Drewmeister]

These are the heat exchangers that I will be using on the next generation chillers. I had these specifically designed by engineers for some extreme liquid chilling at low flowrates. They are small but pack a big punch. The design refrigerent was r507 so with a saturated suction temp of -60c/flowrate 2GPM of methanol these guys will cool 500watts(1.7kbtu) to -50c. That's 500watts(1700btu's). 200-250watts will be cooled to within 2-3 degrees of the SST. Keep in mind too if one were to increase flowrate and or thermal properties the efficiency will be sick.

As you can expect I have shelled out a few bucks on all of these but don't worry the total cost of the chillers will be very reasonable. My first prototype is going to be a machine that can produce close to the above results. The heat exchangers will work fine in lesser capacity chillers too. If you guys are serious about a micro chiller I can start production on that first.

[markdm]

Those are sweet looking units. but i think i missed AC/101 class on what there for and how there used.

BTW do you have a pic of one of your micro chillers, and a price?

[Kalivar]

Drew,
If you can give me a price tag of a micro chiller that could hold around 10C for an athlon 2400 @ (hopefully around 3Ghz) proabably around 2.0v, 9800pro and NB.

More important than temps is size. How big and heavy are we talking for the micro chiller. I want to go sub-ambient, but want my system to be semi/portable. I don't want to deal with the condensation because of the changing environment for the computer. Maybe it would be better to go with a less powerful pelt attached to my watercooling system?? Maybe 200Watt or so?? It think the pelt would be tricky to stabilize temps.

I will be able to buy one (if the price is right) in a month or two.

Let me know what you have in mind.

[madramper]

I guess "how big will it be" is the main question. I have two rigs, a monitor, and a tv on my desk already, so space is at a premium.

What do you have in mind as far as the design? Will it be one box with everything inside, or external pump/rad etc? Any sketches? And of course, this is America! Whats the price for us po folk

[markdm]

Update on my video card: they have refunded me the price of the old card, as soon as it posts back to my account i will be ordering the new Albatron 6800GT

[Melodis]

I have the Albatron 6800 GT and it's a workhorse. Good choice!

[Iliketofrag]

K after I make my computer I will start this as I just got a 6000btu a/c for free because my dad works in the hvac buisness. he gets unused a/c and converts or uses the parts.!!!!!

[markdm]

Originally Posted by Melodis

I have the Albatron 6800 GT and it's a workhorse. Good choice!

Cool, I looked around and it seemed to me that it was the best one i found with the best features for the price. $399.00 + 0.99 fedex shipping.

[Drewmeister]

Originally Posted by markdm

Those are sweet looking units. but i think i missed AC/101 class on what there for and how there used.

BTW do you have a pic of one of your micro chillers, and a price?

Simply put....Those are the evaporators. That is where the liquid refrigerant absorbs the latent heat(heat) from the process fluid(coolant) and phase-change(evaporation) takes place. To give you a good idea of where they are positioned...one of these plate heat exchangers can take the place of the bulky evap in reservoir on an A/C chiller for much better efficiency, shorter pulldown time and like 20 times more compact.

The sub-ambient micro chiller is officially in production as we speak. As soon as I gather a few more details and component prices I will quote you a price through pm.

[markdm]

Originally Posted by Drewmeister

Simply put....Those are the evaporators. That is where the liquid refrigerant absorbs the latent heat(heat) from the process fluid(coolant) and phase-change(evaporation) takes place. To give you a good idea of where they are positioned...one of these plate heat exchangers can take the place of the bulky evap in reservoir on an A/C chiller for much better efficiency, shorter pulldown time and like 20 times more compact.

The sub-ambient micro chiller is officially in production as we speak. As soon as I gather a few more details and component prices I will quote you a price through pm.

Nice, i understand now that would be great to be able to pull down faster and be more efficient at the same time. Im sure they were not cheap though.

Oh, and now that i understand what they are, the wheels are turnning on how to put one in the 12 can cooler, or maybe even a 6 can cooler.

[Drewmeister]

Ok guys... There is quite abit more engineering involved with the design of this one b/c of the size and temp control requirements. One of the main problems is not having a reservoir of any useful volume to buffer the temperature so the compressor doesn't have to cycle every few minutes. I'm working out all the details and pricing of components for two or three different type systems ranging in size and price to meet everyones needs.

The goal is to fit it all in an enclosure that measures 8.5"-10.5"H X 8"W X 18.75"L. That would sit perfectly under a Chieftec mid-tower. http://www.chieftecusa.com/dragon.htm (of course it will be painted to closely match your case) A 21' length on that would allow for a decent size tank. Worst case size would be approx 12.5"H X 8"W X 18.75"-21"L but would also be the cheapest of the bunch. These are approximate sizes but should be pretty close.

I have calculated the heaviest load to be around 300watts with a 3.4EE oc'd 4GHz @1.8v and an oc'd 6800 Ultra. I will build in a little more capacity on top of that for good measure.

Again...Once I have all the details and pricing worked out I'll pm quotes to all those who have inquired.