Grunty chiller

[{.bLanK} GoD]

I'm thinking i might put a loop in the discharge line to take away some vibration to that top condensor joint. What do ya reckon?

[SexyMF]

I think people should stop guessing the reservior capacitance. Since you have spent megabucks on brand new gear then spend the few tens of dollars more on capacitors which will do the job.

25% voltage ripple is unacceptable. 5% is high. Remember that the 15.5V going to the pumps and fans is already 29% overrating them.

Two diodes in series would knock down the voltage to around 12V.They will need to be heatsinked.

[SexyMF]

// I think the build looks great. Like the look of building the HX into a box. Nice and tidy.

Look forwards to it running.

[{.bLanK} GoD]

25% voltage ripple is unacceptable. 5% is high. Remember that the 15.5V going to the pumps and fans is already 29% overrating them.

Two diodes in series would knock down the voltage to around 12V.They will need to be heatsinked.

My contact @ south island componants assures me the voltage will drop to around 12v once I pump a few Amps through her, I will monitor everything once it's all going. If the voltage is still to high I will place a voltage regulator in just for the fans, the pumps are rated to 24VDC.

Anyway, been a few days since i last posted, fired her up for the first time lastnight, got some frosty pipes and the temp from 15c to -20c in around 45mins. (approx 8 litres)
Still haven't connected the heat load, or tuned the cap tube yet. Thats this weeks mission. Along with wiring her up once the electrical box is finihed being made.

[Xeon th MG Pony]

yes you should all ways add three loops to dampen vibration.

[Nosfer@tu]

Subscribed :d

[symphy]

Awesome work so far (on a chiller)... I like your evap very much... Lookin' forward to the results. How many W does your "loadtester" have?

[{.bLanK} GoD]

Awesome work so far (on a chiller)... I like your evap very much... Lookin' forward to the results. How many W does your "loadtester" have?

Cheers dude
The load tester will produce approximatly 520 watts.
The electrical box has arrived, I spent most of last night wiring it up. I will have pictures later today. The G-froster is now holding me up

[{.bLanK} GoD]

I modified the load tester because it was going to be difficult submerging it in the res and sealing the res properly. I't will now be inline with the pump. (the ehiem is mine, just for testing purposes).



Copy and pasted from another forum....

How do you make a 500watt heat load to test this puppy with i hear you all say?
First borrow mums good jug...
Next, butcher the old jug thats been rolling around the back of the van for the last year, seal the top about a third of the way down with a piece of tin i prepeared earlier. Seal two 1/2" pipes through the new lid and secure them with some galv strap so you don't upset the seal when connecting hoses. (drill holes thru lid before sealing in place)

WARNING!!! Mains electricity must only be tampered with by a qualified or competent person. Incorrect or substandard wiring can lead to serious injury or death!!

Now, this jug is about 4x more power than we need so to run it at the wattage we want we do this.
(Being a simple resistive ac circuit, Ohms law still applys)
The elemnt has a resistance of 27 Ohms
The voltage suppling my garage is 237 VAC
P = Watts

P = V squared / R
P = 237 x 237 / 27
P = 2090.3

Now here is where mums good jug comes into play.
Wire both jugs in series, being just two resistors in a series circuit, add the resistors together and recalculate.

P = V squared / R
P = 237 x 237 / 54
P = 1040.16

Half that (1040 / 2 = 520) 520 watts is approximatly what each jug will use. Close enough. Of corse the jug with the coolant running through it will need to be insulated, the other jug will just sit there and boil. But it will take it about an hour or so to boil a full jug.

[{.bLanK} GoD]

Here is the electrical box so far, have a spot there ready for the G-froster already, still need to cut holes for the cables, put gromets in too, and the mains plug socket in the back.

[{.bLanK} GoD]

I have her running atm. I let the temperature drop to -27c before i was ready to turn on the load tester (which by the way looks a bit different now )
While away having tea the temp raised to a balmy -1c So more gas was added until we get a little frost on the suction pipe about 4-6 inch's back from the compressor. Temp has dropped now to -15c, and with a lot more colder gas getting back to the compressor, the compressor has cooled down to around 80c lol
Will have pics and more results a little later tonight.

[{.bLanK} GoD]

Well after spending 6 hours test running it today, i could get it holding -17c under 520 watt load. after turning the heat load off, it took just over 30 mins to drop to -30c. A CPU and two GPU's idling would level the temp somewhere around -23 -25c I suspect.
But I'm afraid -17 loaded just isn't good enough. I will cut the cap tube back and redo the tests, probable a few times to get a result i'm happy with.
The last pic is the load tester, I placed that sealed jug in a box and filled the box with expanding foam. God i love that stuff It's insulated, and it sealed the tiny leak the jug had, even covered in silicon.

[-Acid-]

subscribed

great thread

[Revv23]

wow really clean work especially witrh everything you have going on.

[{.bLanK} GoD]

Cheers guys
gave it a second run tonight before i cut the cap tube back, and yielded similar results. Slightly colder ambiant temp than lastnight and tonight was holding -18c.
Compressor RLA is 3.13A and she's running 2.4A at the mo.
Pressures are 12 and 170psi
3.5m of 1.3mm cap tube

[Modzilla]

Very nice work indeed! Impressive. Is that insulation on the pump motor housing? I'm not sure you want that. You want the heat generated by the pump motor to dissipate into the air, as much as possible.

[{.bLanK} GoD]

Very nice work indeed! Impressive. Is that insulation on the pump motor housing? I'm not sure you want that. You want the heat generated by the pump motor to dissipate into the air, as much as possible.

Yeah it is. thats the pump off my chiller, sticky back closed cell foam was wrapped around it to prevent condensation, but you make a valid point, do any of you guys insulate your pumps? do you get much if any condensation? or is the heat from the pump generaly enough to prevent condensation?

Wee update.
The electrical box is nearly finished, added a fuse to the 12v rail, wired up a cable for the pumps and put jug plug sockets in the back, a normal male socket for 230v in and a female socket to power the pc.

[Modzilla]

My only experience with chillers is the chiller I recently finished, as seen in the Proper A/C Chilling thread in this forum. The pump is a danner mag supreme 3, which consumes 35w. You can see from the pics that the pump motor housing is free of frost while the impeller chamber is frosted up, at about -32C coolant with about 20C ambient. While running, the motor housing is slightly warm to the touch. I previously used a danner mag model 2, which consumes 24w, and that also was enough to keep the motor housing warm enough to prevent condensation. I saw a thread somewhere with pics were somebody completely and thoroughly insulated the entire danner mag model 3, including the motor housing, with thick neoprene, and the plastic parts of the pump melted around the pump motor. Thats looks like an ehiem, I'm not familiar with them, but I'd think you'd find the same: that the impeller chamber needs insulating and the motor housing does not. Only one way to know for sure... Keep up the good work!

[{.bLanK} GoD]

Spent most of this past weekend pulling my hair out because the chiller didn’t want to play the game. Long story short, I started with 3.5m of 1.3mm cap tube assuming it was slightly long. Three days, a couple of experiments and five times charging and vac’ing it, I found that 3.5m was slightly short. I extended the cap tube to 4m and it was holding about the same crappy -15c. I’m now at 3.9m and holding -17c under 520 watts.
So our sweet spot is somewhere between 3.5 and 3.9m. Hopefully this week I will find that mmmm sweet spot.
Controller is in ozzy I have been informed, should be here “touch wood” just in time to give her a final testing run.
I still have to break it all down again to extend that coolant return pipe in the res, replace the cap tube with a continuous piece and I’ll replace the filter/dryer since they cost me less than it does for a cherry coke. Lmao
All going well, it’s looking like sometime around the end of next week when I will be able to freight it out.

[n00b 0f l337]

Blank with capillary so long, why not change to a wider ID capillary line.

[{.bLanK} GoD]

Blank with capillary so long, why not change to a wider ID capillary line.

The fridgie that worked it out the cap length for me roughly worked it out to about 2.8 -3m i think it was of 1.2mm cap tube. We can only get (easily) 1.1mm or 1.3mm cap tube, so we decided to go a little bigger and a little longer. He said 3.5m would be a good starting point and i'd need to cut a bit back a little.
And after i have spent so much time so far tuning it with 1.3mm, I'm not going to change the diameter now.
Look on the bright side, it just means more cap tubing to wrap around the evap for that tiny bit more sub cooling

[{.bLanK} GoD]

Dudes, I wish I had days like today more often. Apart from an upset customer at work, (which doesn’t really phase me much anyway) my day went sweet.
First off my annual pay review was today and I’m now getting an extra $1.50 per hour, stopped off at the metal finishers to see if the could acid dip the evap to remove the oxidization on the outside. The guy I dealt with was really interested in what I am doing and dipped the evap then and there, took about 5 minutes and he didn’t charge me a cent I’ll be going to that firm next time I want anything chromed, electroplated, acid dipped etc.
Then when I got home, the G-Froster was sitting there smiling at me. Hehe

The final cap tube length I ended up with is 3.65 meters. It was holding -18.5c.
The temperature under testing is measured from a brass hose joiner in the pipe between the coolant out connection of the reservoir and the pump. So that will be as good as an average temp between all three water blocks, (CPU + x2 GPU) under max load. I’m curious to know what the temp of the coolant itself is so I borrowed a submersible temp probe. Mine is sealed inside my chillers reservoir.
Gave the reservoir hard coating a lick of black paint and it’s looking pretty smart.
So I have a few nights work to keep me busy this week.

[{.bLanK} GoD]

Finished!
Finally completed all testing today and she is ready to be packed and shipped out.
The G-Froster controller controls the whole unit including the pc psu.
Once wired in correctly, (which I have made particularly easy), the G-Froster is powered first, by the press of your normal pc on button the controller comes to life and goes into it's 5 minute start delay for the compressor.
Once the countdown has reached zero the phase unit is powered on as well as the coolant pumps which are powered directly from the electrical box.
When the temp sensor reads your desired pc start up temp (which is configurable from the windows interface software), the controller then starts the pc automatically.











Hows that?

[mike047]

I have watched this project for some time now, VERY NICE work.

Who is the recipient of this creation??? I am sure they will be pleased.

[Modzilla]

Very nice work. Nice startup delay and startup temp select via software.