[work-log] Plate HX water chiller (R-507/TXV)

[tiborrr]

Components:
- ACC Electrolux MP90TB 3/8HP MBP compressor
- 1.04kW condenser + 230mm fan @ 5W Papst 220VAC motor
- Alco TS-SW thermo-expansion valve /w 1 orifice (equivalent to 01 Danfoss)
- 16-plate stainless steel heat exchanger
- Dixell XR10C thermostat /w NTC temperature probe 1.5m
- refrigerant R-507
- Iwaki water pump running glycol/water mixture

Projected evaporating temperature: ~ 0 to -5°C
Projected cooling power: ~ 1.05-0.9kW

Build will start next weekend due to time constrain Only thing I managed to do is to braze the plate HX evaporator:

Purging / brazing station. Note that blue refrigerant bottle is there only to support Nitrogen high-pressure vessle

[rusty959]

awesome! I am tuned in. I am going to be starting my chiller soon with a plate heat exchanger as well, so I will be closely watching this.

(yea, so im not on xs as much as some other forums, so hello to all the people who don't know me )

[kiyka]

Nice parts.

Good luck with the process.

[tiborrr]

Thanks fellas!

Well, some progress has been made:


Pressure tested with nitrogen and appears to be leak-free. Tommorrow I have to assemble insulation compartment for this heat exchanger and manufacture the base plate. The most difficult and painstaking part is now behind me.

[tiborrr]

I have managed to complete the HX enclosure, now it only needs some bondo & colour:






Tommorrow frame welding then frame painting, the day after tommorrow -> brazing the last few joints, vacuuming & charging, perhaps first run

Best Regards,
Niko

[Jack]

Looks good, man

What kind of brazing rod did you use for the Plate HX?

[tiborrr]

Thanks Jack!

I used cheapest silver rods, I think they're rated 15% along with white flux.

[tiborrr]

Designing the water distributor (for three paralel waterblocks):


Heat-exchanger, insulated with pour-foam:

[tiborrr]

Final lay-out of the components (the vertical outlets will be tilted 45°):

[tiborrr]

Main work is done, waiting on steel frame (casing) to complete assembly (last few joints):

[DarkCow]

This looks awesome, I've always wanted to build one.

Whats that vertical tube that goes no where for? Is it to fill the loop? Where is the water flowing?

[Unseen]

Hi mate,

I think you have to change a little bit the design. I am sure you know that the water always choose the easiest way. In your case i think is the pipe with the smooth angle(the last of the 3).
It will be better if you do sth like that

Also the customer will have to use the same length pipes for connecting the blocks and same with same restriction.
Using blocks in parallel is a little bit difficult.


And at last I suggest to make a tank because your pump is really strong so with T it will be almost impossible to remove all the air from the loop.

[tiborrr]

Thank you for your input Unseen . I still think water distributor won't cause issues since you have to take suction in consideration. Water loop is closed so three equal blocks (it is meant to cool 3x VGAs) on three equal lenght tubing shouldn't cause issues, what do you think?

I'm off to my vacation tommorrow morning, so work will continue on the 1st of August. Until then frame will be assembled and painted aswell

Best Regards,
N.

[ilkkahy]

Im pretty confident that optimizing the distributor flow has zero effect in practice on this case. In theory, -ok. Hydraulics, exhaust manifolds and "bigger" applications are another thing.

Nice project

[Unseen]

Take a look those

[tiborrr]

Chiller is assembled and running. Will be throwing heat at it tommorrow!

Final assembly:


Initial charging:


Best Regards,
Niko

[tiborrr]

Chiller done, ready, tested:



Best Regards,
N.

[biglipzit]

OMG you tested without insulation and just paper towels under the lines over the board.. Very very brave hahaha. Awesome setup!

[tiborrr]

Hehe, yeah. Too bad the setup was limited by Striker II Extreme FSB clocking - this was a horrible sample I guess, only 550FSB, couldn't post at 551MHz no matter what. Should have chilled the NB aswell I guess

Thank you

[Sam__]

great chiller, very well put together.

[tiborrr]

Thank you Sam!

Here are more photos of re-done water stage, made by my customer's buddy who is also a professional plumber:






In action:


Best Regards,
Niko

[tiborrr]

My buddy already did some benching with insulated unit:
HWBot


Hardware:
- Sapphire ATI Radeon HD3850 AGP 512MB DDR3 (Vgpu, Vmem mods)
- Motherboard ASRock 4CoreDual SATA2 (all mods imaginable done),
- Processor Intel QX9650 (ambient water cooled)
- 2x1GB DDR2 Crucial Balistix PC8500
- HDD SATA2

Photos:

Graphics card preparation:



... some photos of the system:



...GPU voltage (default = 1,24V) & water return temperature (ambient temperature = 31°C)



This is by far the fastest AGP system. Next run will be with single-stage cooled CPU

[Naja002]

great chiller, very well put together.

Definitely Agreed!

[DangerMouseDom]

Awesome work here m8,,,

It seems so much like the project I Have still un-started in bits needing attention.

I have a large 20 plate exchanger, various valves and spare copper, and old rotary compressor with large condenser and fan still attatched, it was originally using R407C in an AC unit.

I am now desperately seeking someone in the UK, who has the tools, gasses and skills to complete my project for a good price.

I am sure I have most of what is required...

How did you select your TEV, and would you recomend going this way using a plate HX, or would simple capilliary tubing suffice, as was my original plan.

Props for your unit, looks great..

Any ideas or recomendation to get my non starter off the ground appreciated..

[tiborrr]

Thank you guys!

@DangerMouseDom:
I selected the 01 orifice as It is rated for 1kW heat-load. I have looked up what cooling power this compressor offers at prefered temperatures and then selected the appropriate orifice for the TXV. Go with Plate HX, it has a fast pull-down in comparison to coax HXs.

TXV or TEV is selected by the gas you are intending to use and then you choose your nozzle / orifice by the cooling power or desired temperatures. For every gas there is a different TXV as the gases have different pressure-temperature corelation curve.