StillDragon® Community Forum

Welcome!

Be part of our community & join our international next generation forum now!

In this Discussion

Water Cooler

245

Comments

  • Would be more interesting if they cut out the evaporator coil and braized in a plate exchanger

  • That is the most picture rich post i have ever seen. I got to the bit where he took a picture of evrey tool he used and quit. Fuck me.

    I'm following this with interest though.

    StillDragon Australia & New Zealand - Your StillDragon® Distributor for Australia & New Zealand

  • I didn't look at the link in in detail but the computer guys are overclocking so the lower the temp the harder they can push their chips before frying them. Their actual power requirements would tiny I recon. Ice banks are are good way to make use of off peak power but it's a lot of mucking around for not much benefit in my case I think but it's another good option for someone to look into. Ambient should be fine for me most of the time. I'm starting to wonder if I even need the fan?

  • So I hacked up the coil, did some testing and got some good data. Turns out I really do need a fan so I just put a floor fan in front of it with some towels to direct the air flow past the fins. I tried the different speeds and they made bugger all difference so just enough air to flush the hot air out of the fins is fine.

    I made three cuts on the tubing.
    First, one cut where the gas used to feed in, then another one where the liquid came out. This separated compressor and all the extra crap so that came out.
    Then the last cut where the gas turned into liquid between the liquid single coil and the two gas ones in parallel (below the purple tee in the first post which is the orange one in this post).
    With a bit of hose I changed the flow direction through the gas coils.

    Sounds a bit complicated but all I really did was change it so I could fed the water in the very bottom and have it rise up through the lot.
    First a precool in the bottom coil where the temp difference was highest then the flow split and the flow rate halved for the final cooling pass.

    I was a bit concerned about how much pressure I would need to get though so I tested that first without heat.

    image

    The chugger was able to get it through the lot plus 10m of hose and up a head height of 2.5 meters and still give a flow rate of 1l/min. At 1 meter head height it was 2l/min.
    So far so good as the pump I’m using in the finished system should have a bit more balls than that.

    After that I put a condenser in between the pump and the unit to test how much power it could handle.
    I set the controller so the condenser itself would have about 4kW hitting it.
    First I just tried single pass from a drum with 18.5° water in it straight to drain.
    The SD 2” shotgun needed 1l/min to knock down 4kW. I couldn’t get the temp of the water coming out of it but it would have been pretty close to 100° (hot enough to burn). Ambient air was 22° so I guess the real test is how close could it get to that.
    I tried all the fan speeds and the temp of the water after the unit was between 23.5° and 24° so I just stuck with the low speed after that.

    Next I wanted to see what would happen if I doubled the flow rate through it. The outlet temp rose 1 degree.
    Finally I put the outlet hose back into the drum to let in recirculate and see what the temp would settle out to after a while.

    So recirculating at 2l/min and with about 5 litres of water in the entire system, 4kW heating the water leveled out at 28.5°.
    Just 6° above ambient, which is pretty good I recon.
    I just need to get the best solution for the fan locked in.

    image

    1.jpg
    600 x 800 - 89K
    2.jpg
    800 x 600 - 87K
  • Excellent.

    StillDragon Australia & New Zealand - Your StillDragon® Distributor for Australia & New Zealand

  • Question - why is everyone using hx's from ac units as opposed to car/truck radiators?

    Also, I remember from a colleague in college that vertical mounting of the hx is less efficient than horizontal. Anyone tried that already?

  • They look like they are supposed to be there would be my guess.

    StillDragon Australia & New Zealand - Your StillDragon® Distributor for Australia & New Zealand

  • Suppose the camouflage the fact that it's already packaged in a chassis made for outdoors is worth a bit.

  • @Unsensibel said: Question - why is everyone using hx's from ac units as opposed to car/truck radiators?

    Also, I remember from a colleague in college that vertical mounting of the hx is less efficient than horizontal. Anyone tried that already?

    Horizontal means you get more of an advantage from convention currents.
    I tried it to see if I could get way with just that and no fan but it didn't really work.
    I'm using the AC unit because looks nice and tidy, is self contained and free standing (and it was free).
    There's also a nice space where the compressor used to sit where I could stick some more stuff too. My flow switch will go there.
    I may even be able to get the reservoir in there too.

  • Man my typing is shisen, convection was the word I was obviously after there but can't edit now. Maybe a mod can fix it.

  • edited July 2015

    @jacksonbrown said: I tried it to see if I could get way with just that and no fan but it didn't really work.

    No way, we all wish it was that easy to cool. Companies are spending small fortunes on cooling towers, adiabatic coolers, chillers, etc etc to cool liquid. Not because there is money to burn, but because it's pretty difficult to cool liquids that aren't very far from ambient temp/dew points.

    Not saying there aren't benefits, but unless you have small rig, unless you have active cooling I don't see how you can get away without a reservoir.

  • edited July 2015

    Another interesting option is to use the heat exchanger to reduce the amount of municipal water supply needed, with no reservoir. Call it a recirculation hybrid approach.

    You could plumb this up using a pump and 3-way mixing valve. Bonus is that you can push high flow rates though the condenser at all times, which will actually improve your heat exchanger capacity. Higher flow rates provide better heat exchange as well as higher turbulence through the exchanger, all good.

    Another benefit of this approach is that it still works when ambient temperatures are high. The lower the ambient temperature, the lower your water usage will be. You could use this year round without worrying that you won't have the capacity, and you don't need to worry about a reservoir or chilling a reservoir in higher ambient temps.

    (Forgot to label the recirculation pump, but you all get the idea)

    image

    Exch_mixer.jpg
    702 x 406 - 22K
  • edited July 2015

    You could actually do this with solenoids too, probably a whole lot cheaper than a proportional 3 way.

  • My convection test was just leaning the coil back 45°. It was still cooling the water just not enough for my uses.
    I think I could get a convection only rig to work quite well but it's a bit more mucking around finding a good place to mount it. Even better would be to mount a big one in the roof space of a shed and pair it with a whirlybird and shroud. I think I could easily smash 5kW with a few liters of tepid water doing that.
    I was killing 4kW with just enough water to charge the system. Barley 5 liters (you can see how much water is in the drum for the recirc test in the last pic). The reservoir was just a buffer to stop the system running dry. The temp was stable in it for 20 mins so I don't think it wouldn't have mattered how much was in there.

    That drawing's not to far off what the ultimate plan is but I'm not bothering with three ways, just a two way on the feed.
    If that three way wasn't there and the supply was turned on and off it would still circulate fine. If your worried about your chilled going straight to drain just make sure the pump suction is lower pressure than the branch and you shouldn't loose any (or throw in a check valve).

    A simpler description of what I'm doing is hooking up the drain to the supply moving the HEX there. A large buffer tank isn't really required.

    To get your sketch to work you would need to do a lot of de-tuning of the HEX. Even with no fan I think it would it would take quite a lot of power to overcome it. Fine for a VM but not so good if you're using it on a CM as you'd struggle to get any take off. You'd pretty much only need the supply to fill the system on start up.

    Com'on Grim, you must have some ideas to make that brushless dc live again.

  • @jacksonbrown said: To you think it can be easily hacked?

    No idea. Maybe. You got to follow the traces on the circuit board and identify the parts. Eventually there must be a speed signal to control the speed. If you are lucky it's an analog signal. Please be careful, there is no isolation transformer so everything is live.

  • @CothermanDistilling said: yes, that would be the ideal way, but a complete mini-split is a lot more than a just the outdoor part of a conventional unit, especially when you go to 3-4 tons..

    If you mean 400000 BTU thats 117 kW. That's a 2000 kg chiller. :-)

  • Problem with getting the existing kit to run the fan is that it probably has some logic built in for when to run the fan, I'm sure if you hooked it up, it would be throwing all sorts of fault codes, so it probably wouldn't trigger. Means you'd probably need to modify the existing circuitry to get it to bypass all of the brains and just spin the fan.

    You can buy a dc brushless controller, but honestly, replacing with a similarly sized motor is probably much cheaper.

    I'd need to spend some time looking at that schematic to figure it out. You still have all the boards?

  • I've got all the stuff on the circuit diagram in the first post.

    Here's the inverter PCB (IC932 is in the flat heat sink below the three small caps in the top right) The fan is the pluged on the top (power on one plug and sensors on ther other).

    image

    image

    and the back (IC932 is where the white goop is)

    image

    image

    I realise it’s a long shot but I was hoping with fingers crossed that someone would have a light bulb moment i.e. “just do x” and its working.
    I don't really want to power the board up without a good reason but I can dig up any more info if someone wants it.
    Unfortunately I think this one might have to go in the too hard basket.

    1.jpg
    800 x 600 - 129K
    2.jpg
    800 x 600 - 151K
    3.jpg
    800 x 600 - 152K
    4.jpg
    800 x 600 - 144K
  • Here's some more that may help.
    The IC without the heat sink.

    image

    The tracks around it.

    image

    The three rotation sensor wires seem to head here.

    image

    1.jpg
    800 x 600 - 83K
    2.jpg
    800 x 600 - 136K
    3.jpg
    800 x 600 - 118K
  • edited July 2015

    Speaking of coolant and reservoirs, construction never ends in the shed. Whoever said putting your coolant reservoir underground was a great idea ... probably didn't need to dig the pit by hand.

    Gave up around nine pm, took about 5 hours to get the concrete out and dig down 2-3 feet or so. Floor was nearly 7" of fiber reinforced cement. Insane. Hole is about 6' by 12'.

    Hopefully we can use some of the cold ground temps to our advantage. That back wall is approximately 23 feet underground already ( the building is partially underground, set into a hillside. The tank will sink close to 6 feet below that (29 feet deep) so average soil temps around the tank once set should be about 52-55f.

    image

    image.jpg
    600 x 800 - 82K
  • @jacksonbrown - Looking at the simplified circuit diagram, it seems possible (you can see the rectifier circuitry leading into the driver clearly laid out there) - but looking at the board itself, there is quite a bit going on.

    It's probably possible, but you'll likely be cutting traces and making jumpers to do it. I'm sure there is enough circuitry throwing fault codes to not spin the fan otherwise. And even then, not sure how the fan speed is controlled.

    Replacing the fan seems a whole lot easier after seeing the pcb. It's probably doable, but is going to take access to the full schematics and a bit of trial/error. I'd try to find something similarly sized with the same shaft diameter as the fan itself.

  • That's kinda what I was thinking too.
    Good luck with ya hole :)

  • Hoping if we keep digging, a 3000l copper boiler might pop out the other end.

  • @jacksonbrown Nice plumbing job there, I have been running something similar for a while but your pipework and fittings put it to shame. Definitely time to upgrade I think, where do the small cam-lock fittings come from?

    I like to keep my cooler reasonably portable so it uses a ~12l PVC pipe resivour in the space that held the compressor, so far this has worked fine at ambient temps up to 32C but it only has to deal with a maximum of around 4Kw so others mileage may vary.

    As for the fan Maybe plug in 12v and see what happens, if the smoke comes out just get one of the cheep oscillating box fans, I used one on a different cooler for a while, it was nice and quiet and worked fine.

    @grim Cool hole, but some things are best left to hydraulics. Are you sure you couldn't squeeze one of those nice little mini diggers in there;-)

  • Brushless dc, don't bother connecting any power to it that doesn't come out of a controller. The controller delivers pulsed output to the coils to control speed. It's not a conventional motor with the windings on the rotor.

    You can probably find a controller, but it will cost double or triple what a motor would cost.

  • @grim said: Hoping if we keep digging, a 3000l copper boiler might pop out the other end.

    +1

  • edited July 2015

    what are the motor dimensions? I wonder if you can squeeze a 42 or 48 frame blower motor in that

  • @badbird said: where do the small cam-lock fittings come from?

    I like to keep my cooler reasonably portable so it uses a ~12l PVC pipe resivour in the space that held the compressor, so far this has worked fine at ambient temps up to 32C but it only has to deal with a maximum of around 4Kw so others mileage may vary.

    If I can't find a more elegant solution I'll just be sticking a 30 buck exhaust fan in there.
    Those fittings worked out better than I thought they would. I pulled the old ones out then opened up the hole in the sheet with a die grinder then bent a thread into it. It's just a male 1/2" BSP hosebarb on the inside and a female 1/2" bsp cam and groove on the other.
    The Camlocks have come from all over the place. Be aware that some of the cheap Chinese Ebay ones don't really fit well on the better quality ones I have. They either crush the seals or simply don't fit or are too loose. Pick a supplier like these guys and stick to them IMO.
    If I run into issues at higher ambient temps I might just drape some wet hessian over it, put a cracked bucket of water on top and turn it into a fan forced Coolgardie :D .

    I was going to put in a PVC tube too. I have some large diameter stuff sitting there I just need to make sure I can mount my supply pump so it stays flooded. It's quite different to the Chugger.
    Got some picks of your setup?

    @grim said: what are the motor dimensions?

    See here:

    image

    image

    1.jpg
    434 x 409 - 16K
    2.jpg
    775 x 363 - 18K
  • Mostly done, just need to pour the floor.

    Underground cistern for cooling water tied into the site storm water system for catchment.

    There is a spring on the property that feeds into the storm water drainage system, provides approximately 20 gallons a minute of cool water to the cistern. Runs 24/7, never stops.

    The cistern itself is about 500 gallons and uses a shallow well pump to feed the 500 gallon still coolant reservoir that sits on top of the boiler room.

    Hoping I'll never need to turn on the chiller.

  • edited August 2015

    image

    image.jpg
    600 x 800 - 108K
Sign In or Register to comment.