Jump to content
Cultivated Reef

DIY Chiller for $60.


taekwondodo

Recommended Posts

taekwondodo

OK. Holy Grail time.

 

We've all looked at ways to continue our passion of DIY for many of the things we do. One thread that seems to be consistant is that no matter what you do, it's just cheaper to buy an ICA or an iceprobe and be done with it (heck, I'm on the edge... looking seriously at the iceprobe for my 20G).

 

The obvious choice would be using a Peltier, as does the ICA and the Iceprobe. The two problems we as DIYers have that are not yet solved -

 

1) an effective heat (cold) transfer method that is safe and non-toxic to our reef.... and

 

2) A temperature controller for the Peltier.

 

 

So - setting 2 aside for now, as one can have a timer turn it on and off and have the system's heater fight to keep the tank warm if the cold is working, and keep the heater set at 78o.... and diddle with the timer until the battle between the heater and chiller is mimimal. Not the best choice, but there are other ways to skin this cat and it does have a work-around for now....

 

But people have been racking their brains out trying to figure out how to effectively solve #1 without killing the reef. The IceProbe looks like it uses a metal rod with some type of coating, and places the (short) rod directly into the water flow.

 

The ICA uses a modifid (apparently) VGA adapter used for watercooling PCs that appears to be coated on the inside of the heatsink with something saltwater safe. Something like this may or may not be out of our reach (the coating part), but a little VGA WC-HS isn't really going to scale well (an assumption).

 

So - how do we build this safely, and better yet, cheaper (by ~50%, or only pride becomes the DIYers motivator) than an ICA/Iceprobe?

 

Well, for $25, you can get a TEC at Frys,

http://shop2.outpost.com/product/1968938?s...CH:MAIN_RSLT_PG

 

Another $15 for the HSF

http://shop2.outpost.com/product/4175033?s...CH:MAIN_RSLT_PG

 

and an 8x8x.125 piece of aluminum as the Heat Transfer device.... $2.50

 

Cheap Power Supply - $15.

 

So we are up to about $55-60. Close to the ICA price... so it's getting to be a pride thing.

 

Put the heat exchanger in the tank? No.

Coat it? No.

 

Because glass transfers heat fairly well (not great, but well enough) making a good contact of the heat transfer device (the aluminum plate) to the glass will essentually transfer the cold to the wall of the fuge (see picture) which in turn is "radiated" (OK, so I'm speaking in reverse) into the water as it passes through the fuge.

 

Any thoughts?

 

- Jeff

Link to comment
taekwondodo

I thought about that - still need to find the metal plate (copper or aluminum)... not sure what the price difference is, but you are correct.

 

(edit) hundreds of Peltiers on EBAY:

 

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewI...7514776526&rd=1

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewI...7511763636&rd=1

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewI...7514778625&rd=1

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewI...5772196714&rd=1

 

and a copper sheet (I'd like a little thicker)

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewI...8192429460&rd=1

 

as

Link to comment

This could be a bad idea do to the temperature differential between the peltier and the glass. if the thermal transfer of the internal temps of the tank and the aluminum plate are to great, there is a good chance of cracking the glass. however this plate will spread the cooling over a larger area so cracking "may" not be an issue as would for sure if the peltier was directly compounded to the glass. how would you keep the device in contact with the tank short of propping it? and would a single peltier spread over a aluminum plate be sufficient

to cool the internal temps. interested to hear other ideas.

Link to comment
taekwondodo
Originally posted by dga

This could be a bad idea do to the temperature differential between the peltier and the glass. if the thermal transfer of the internal temps of the tank and the aluminum plate are to great, there is a good chance of cracking the glass.

 

that assumes two things:

 

1) that the peltier can get cold enough - I am sure it can, but the bigger the plate the more the cold gets distributed, and

 

2) that the water will not carry away/echange the cold fast enough to there the temperature difference is big-enough to have any harm to the glass.

 

Remember - if we apply this same argument (heat differential of the glass, the element and the water) to el-chepo glass HEATERS, then we would have cracking heaters happen all of the time... but we dont. Becuase the water moves the heat around/away from the heater the temperature effect of the heater-glass-water interface never changes so drastically all at once to cause the glass to expand/contract fast enough to break. HOWEVER, pull that sucker out of the water while it is on, and then put it back in and CRACK. And I believe that a glowing heater provides a much LARGER heat differential than any peltier might.

 

Now for the reality - I had to buy the iceprobe because it got so hot this last weekend I had to put ice in the fuge a couple of times to keep the temp under 84 and its going to be another hot weekend and I just don't have time to experiment around (yet) to get this working AND fight the coming hot weather all in the same day (that, and I'm going skydiving saturday :P and that's #1 priority)...

 

I will get back to this, however I think my next project will be an LED array for my 5.5AGA, and then get back to this.

 

But the more I look around and think about it, I believe it can be easily made to work for under/around $40.

 

 

wasn't there already a long thread about how this method was a bad idea?

 

Yes, but I think the premise was wrong (or following the same logic we would have broken heaters all over the place).

 

walmart mini fridge and 20 ft of tubing w/ a small low gph pump

 

Thanks, but my wife doesn't want to look at the mini fridge. But again, a very workable idea (and in both cases, I still need to overcome the controller problem).

 

how would you keep the device in contact with the tank short of propping it? and would a single peltier spread over a aluminum plate be sufficient

 

There is thermal grease you can get at Frys (A local electronics chain) that one would have to make sure the coating was sufficient enough to be "sticky" enough to maintain contact. A few dolops of silicone on the corners should keep it in place (and allow for removal later if you really had to).

 

And I _believe_ the spread would be fine... and actually more efficient. But that's just a guess.

 

Now dammit, you all got me thinking about it again, here after I've gone and bought that iceprobe. Now I want to get back to it because, as an engineer, to prove or disprove my theory. But I'll no longer have a tank that needs it.

 

Oh bother. Wait! I hear a King Air calling my name....

 

 

(jeeefffff.... jeeeffffff....)

Link to comment

I had been thinking about this back here, but came up against two major barriers. A: Heat conduction, and B: Current draw. Glass and acrylic are both (relatively) lousy conductors of heat. Copper, which we cant use in the tank, is one of the best. The other option is inconel, which is a pretty good conductor of heat, but a hair expensive, and hard to find. (I had a quote from a local machine shop for about $100 for a 2x2inch inconel heat sink... )

 

Id love to hear from someone with an engineering backround what kind of heat transfer we could get through say, a 1/4" sheet of acrylic or glass...

 

The current draw on a peltier is pretty high as well. Not insurmountable, but the powersupply is going to be rather large and expensive, if you dont already have one laying around.

Link to comment
Originally posted by MrYenko

I had been thinking about this back here, but came up against two major barriers. A: Heat conduction, and B: Current draw. Glass and acrylic are both (relatively) lousy conductors of heat. Copper, which we cant use in the tank, is one of the best. The other option is inconel, which is a pretty good conductor of heat, but a hair expensive, and hard to find. (I had a quote from a local machine shop for about $100 for a 2x2inch inconel heat sink... )

 

Id love to hear from someone with an engineering backround what kind of heat transfer we could get through say, a 1/4" sheet of acrylic or glass...

 

The current draw on a peltier is pretty high as well. Not insurmountable, but the powersupply is going to be rather large and expensive, if you dont already have one laying around.

 

Can we just paint aluminum with marine epoxy instead and use it just like iceprobe?

Link to comment

Most epoxys are pretty good heat insulators themselves...I mean, you'd get some heat transfer, but not nearly what you would with a bare chunk of metal in the water. I would imagine that a VERY thin coat of epoxy or corrosion-resistant paint would be the best, but it would still be a comprimise.

???

Link to comment
taekwondodo

Think about it - Glass is pretty good at heat transfer. I think we're all trying to make this a tougher problem than it really is...

 

Glass is used on the heaters we use for our tanks, and many states now have laws that require dual pane w/Argon gas between panes to provide insulation and save energy - because single pane windows are so crappy at keeping heat from transfering through them (heat in winder, cold in summer).

 

I had to buy some window this weekend for a shed I am building and needed something inexpensive - its an outbuilding and there really isn't any need for insulation - and I can't buy el-cheapo single panes in CA anymore.

 

- Jeff

Link to comment
Originally posted by taekwondodo

Think about it - Glass is pretty good at heat transfer. I think we're all trying to make this a tougher problem than it really is...

 

A distinct possibility.

 

Originally posted by taekwondodo

...and I can't buy el-cheapo single panes in CA anymore.

 

Gotta love the Peoples Republik eh?

Link to comment
PeTe PoLyP

the glass in your 10 gallon is designed to shatter (tempered) and with a direct change in extreme temps it will.......but lets face it, you arent going to go from 150f to 50f. the only flaw that i could see is that the product you are looking at isnt designed to reduce a temp in such a big volume....and even in its designed application its probable that it does not reduce the temp to an ideal reef tank temp. lgreen has a good point. A mini fridge can work if you drill a couple small holes and run airline tubing with a small pump. Lets face it.....how many times have you been admiring your reef and needed one of your favorite cold beverages? Or had to run all the way into the kitchen to get frozen mysis?

Link to comment

how about a thin plastic box that will float on the water,in the bottom of the box we attach the cold side of the Peltier.

 

can it do the job?

 

the same thing we can do with a box thats made from other material thats made of a more heat transfer material...any suggestions?

 

some material will not float but we can attach it to somthing.

 

the idea is doing the effect of attaching the peltier in a way that it wont harm the tank glass...or even have a more effective solution

for the heat transfer instead of the tanks glass

 

what do you think?:happy:

Link to comment

i would be wary of attaching the unit to the rear pane of my tank....any spillage and i'm kinda done for.....

secondly, i don't like hte mini fridge idea.......

last, using a floating plastic box would shade my corals.........

iceprobe it is......

Link to comment

I'm thinking about doing this. I have a question though, how would you attach the thermoelectric cooler to the copper/aluminum plate and the copper/aluminum plate to the tank/refugeum?

Link to comment
taekwondodo

I would use some arctic silver between the thermo and the plate and epoxy the edges, and then use Arctic/silver between the plate and the glass and GOOP along the edges (cause I can cut-away the goop if I had to).

 

Unfortunately - it will be a while before I get to try this theory - I'm so overbooked on projects right now I just went ahead and ordered the Iceprobe.

 

- Jeff

Link to comment

Hey, check out Nanocustom's job on a TEC chiller:

http://nanotuners.nanocustoms.com/nanotune...?products_id=79

All you'd need is

 

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewI...ssPageName=WDVW

and

http://shop2.outpost.com/product/3600895?s...CH:MAIN_RSLT_PG

and

http://nanotuners.nanocustoms.com/nanotune...opping_cart.php

 

 

The total including shipping comes out to $77.24

 

Who knows, it may be more effective than the Iceprobe. I am going to save up enought to do it.

Link to comment
Originally posted by halfpint1187

Hey, check out Nanocustom's job on a TEC chiller:

http://nanotuners.nanocustoms.com/nanotune...?products_id=79

All you'd need is  

 

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewI...ssPageName=WDVW

and

http://shop2.outpost.com/product/3600895?s...CH:MAIN_RSLT_PG

and

http://nanotuners.nanocustoms.com/nanotune...opping_cart.php

 

 

The total including shipping comes out to $77.24

 

Who knows, it may be more effective than the Iceprobe. I am going to save up enought to do it.

 

Where are you going to get a 15v 30 amp power supply. 30 amps is a lot.

Link to comment

I have an old model train that i was going to use to power the peltier; that way i would be able to monitor how cold the peltier gets. I'm not sure if it is a 30-amp transformer or not, but ill bet its pretty close.

Link to comment

Do you guys think i should use arctic silver on the waterhead directly in contact with the porcelin? or should i put a copper plate in between?

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Recommended Discussions

×
×
  • Create New...