How to cool a NANO tank ? Thread #2

[gr81]

Continue from:

http://www.shrimpnow.com/forums/showthread.php?t=486&page=20&pp=10



I did not read whole tread, but cant chiller poison shrimps by copper or other metals from peltier or "body of chiller"?

[silane]

no no no.

Small fans are noisier. They have to run faster to move the air and so create a higher pitch sound!

Big fans have big fins, so can be run at lower speeds to give a good amount of air movement.

Ask any PC Modder. :cool:

A spelling mistake, what I meant was to use a LARGE heat sink, with a lot of fins and small fan. The large surface area is offered by large number of fin. There are some fanless CPU heatsink around, but huge !

[silane]

I did not read whole tread, but cant chiller poison shrimps by copper or other metals from peltier or "body of chiller"?

Peltier is isolated from water, one cold is transfer to the water, so there is no chance of peltier material leak into the water.

[gr81]

I have two more question about chiller presented in library.
1. That is "cold fin" made of? It looks like aluminium, Will it not harm shrimps when the aluminium spread to watter?
2. What type of silicon is used? Is it classic aquaristic silicon? My does not glue plastic, so It will not seal well. Aren't other type of silicon harmful?

3. :) how much will it affect if the water flows only through the pipe inside of chiller. Chiller filled by watter or some other thermal compound good transferring heat and pipe only going through. Should it work? If chiller will be approx 40cm long?

[xtrekker]

Ok, you've got a 400W power supply and you've use a big heatsink with a fan on the peltier. How to you get rid of all this heat?

A peltier is very inefficiency, at least 50% of its power is wasted as heat so the heat from the Power Supply and the heat from the heatsink/fan will have to go into the air in the room with the tank.

So ultimately the temperature in the room will increase and this will in turn heat up the tank surely?

Then there is the noise aspect to all this. A powerful ATX Power supply may have 2 fans in it.
So with at least 2 or 3 fans running, won't this cooling system make a lot of noise?

Last summer I bought a cheap little peltier powered can cooler and simply fed 20 feet of airline tubing coiled up inside it to a small eheim water pump.
It took my 60L tank at 31C down to 26C after 12 hours.
But the fan in the cooler was pretty noisy for having on in the living room. So I opened it up and put in a low noise version of the fan.
However this just made the cooler useless as the heatsink was obviously heating up faster than the fan could dump the heat.

I would have had to double the size of the heatsink to offset the lower airflow.

I've now got a new 200L tank and I just know that next summer I'm going to have a tank at 31C again.

I think I'll be better off pumping the water off to the nearby kitchen into the Freezer or Fridge. :-)

i agreed that peltier is not really that efficient and has its limits. If you intent to cool anything more then 200L i would suggest getting a chiller. The only advantage is that it can turn into a heater in winter . The best C.O.P is Stirling cooler followed by phase change then at the bottom is peltier.

cheers
xtrekker

[xtrekker]

I was thinking to drive a 24V TEM using 15V. :)

Connect 2 12VDC TEM in series and bingo you get a 24V TEM.

cheers
xtrekker

[xtrekker]

I have two more question about chiller presented in library.
1. That is "cold fin" made of? It looks like aluminium, Will it not harm shrimps when the aluminium spread to watter?
2. What type of silicon is used? Is it classic aquaristic silicon? My does not glue plastic, so It will not seal well. Aren't other type of silicon harmful?

3. :) how much will it affect if the water flows only through the pipe inside of chiller. Chiller filled by watter or some other thermal compound good transferring heat and pipe only going through. Should it work? If chiller will be approx 40cm long?

1.For my case i used a spare P4 heatsink. Its made of anodised aluminium.

2. For the sealant , I used epoxy . So far nothing happended to my shrimps.

3.To reduce heat intake , the distance from the cold plate to the
tank should be kept to the minimum.

cheers
xtrekker

[gr81]

If I buy peltier 15.6V max. 10A max.
Can I use ATX PC PSU which gives at 12V 20A max? What will happen if max current (A) of peltier will be smaller than PSU can give? Will it work or will peltier burn. What other case if max A of peltier will be bigger than max PSU can give. will PSU collapse? I read "how to" but i'm not sure.

[xtrekker]

If I buy peltier 15.6V max. 10A max.
Can I use ATX PC PSU which gives at 12V 20A max? What will happen if max current (A) of peltier will be smaller than PSU can give? Will it work or will peltier burn. What other case if max A of peltier will be bigger than max PSU can give. will PSU collapse? I read "how to" but i'm not sure.

yes you can . dun worry the PSU will only supply the current needed. From my experience , PSU will "burn" out first. My 2 cheapo 480W PSU overheated after 10mins . Did some trouble shooting and the switching transistor had burn out.

here what i did to rectifiy the problem ,

1. By default the cooling fan is running at 5V . Run the fan at Vmax @ 12V

2. Orginal fan is rated at 0.14A . Change mine to a fan rated @ 0.24A

3. Change the fan orientation so that it blows inwards. It works for me cause the heatsink is just directly infront of the fan. There is alot of discussion/argument on "blowing" or "sucking" . As long as its work its a good method..

cheers
xtrekker

[gr81]

what type of Peltier do you have? How can PSU burn out if no max A is reached?

[bezz]

http://arofanatics.com/members/bezz/peltiercooler/image01.jpg?nocache=1134601373

Look what I found. Bought a few of these to play around with lol

http://arofanatics.com/members/bezz/peltiercooler/image02.jpg?nocache=1134601373

[bezz]

Cost $35 each :)

http://arofanatics.com/members/bezz/peltiercooler/image03.jpg?nocache=1134601373

http://arofanatics.com/members/bezz/peltiercooler/image04.jpg?nocache=1134601373

[NanoDave]

Nice find! :) where did you get it from?

[spinex]

Bezz,

Look very interesting. But it's a bit underpowered from what i see from the manual. 6v, 1.5A. Power is 9W .. our peltier normally already running at 30-50w range.

Of course nothing beat a practical experiment on it. Maybe you can test by fill it with water and see how fast the cooling work on that small amount of water 1st.

I wonder if we could dismantle the thing and see what peltier they using so we could replace it with a more powerful one :D

[tsia]

How much did u pay for the can cooler and where you got it from ? ;-)

[ianlim]

maybe just buy 2 or 3 of the can cooler and then loop them together in series, wonder whether it will works?

[spinex]

maybe just buy 2 or 3 of the can cooler and then loop them together in series, wonder whether it will works?

Each cost $35 according to Bezz .. buying 3 = 3 x $35 = sgd$105.

Doesn't make economical sense :D in term of cooling rate and $$$ spend.

[silane]

Each cost $35 according to Bezz .. buying 3 = 3 x $35 = sgd$105.

Doesn't make economical sense :D in term of cooling rate and $$$ spend.


The rating of that TEM is too low, notice that it is good to cool a can of drink. Understand from someone who used another winer chiller which is specified to cool 4.5l of water, the result was also bad.

Best is to use direct cooling rather than passing through reservior as there is too much loss. I can get 22C with cold fine directly coupled into the tank.

http://www.shrimpnow.com/forums/attachment.php?attachmentid=316&stc=1

[spinex]

Thermal conductivity of common metal from Wikipedia

http://en.wikipedia.org/wiki/Thermal_conductivity

Thermal conductivity of other common materials:

Diamond 895-2300
Carbon Nanotubes 1400
Silver 429
Copper 386
Gold 317
Aluminium 237
Brass 120
Platinum 71.6
Iron 80.2
Lead 35.3
Mercury 8.514
Quartz (273K) 6.8-12
Ice (273K) 2.2
Glass 1.35
Wood 0.04 (balsa) - 0.35 (fir)
Styrofoam 0.033
Wool 0.04
Silica aerogel 0.017
Air (100 kPa) 0.0262
Water 0.6062

[xtrekker]

what type of Peltier do you have? How can PSU burn out if no max A is reached?

I am using a 350W TEC , Vmax @ 15V and Imax @30A.

Apparently , the cheapo ATX PSU comes only with one fan and its connected to 5V. This cause the switching transistors to overheat and its operating over Tmax.

cheers

[gr81]

I am using a 350W TEC , Vmax @ 15V and Imax @30A.

Apparently , the cheapo ATX PSU comes only with one fan and its connected to 5V. This cause the switching transistors to overheat and its operating over Tmax.

cheers
350W??? How big tank you have, how low can you cool it? Its quite big TEM :) I'm thinking about 100W for my 82L tank.

[gr81]

Shrimps aren't sensible to aluminium? If I use aluminium plate to contact peltier and water, I'm afraid aluminium will spread to water. I also saw that some pictures shows using old CPU passives which are coated with black/green/or other color. I have also some here. Is that safe? I'm really afraid of my minions:)

[bezz]

Bezz,
Of course nothing beat a practical experiment on it. Maybe you can test by fill it with water and see how fast the cooling work on that small amount of water 1st.
Can't fill with water as it is only meant as a can cooler. I tried cooling a can of drink and it took the can from 31 to 24 degrees in about 12mins, I think can go lower but i didn't wait around. Tried using it as a reservoir but got lousy results and also tried coiling airtube method also got lousy results.



I wonder if we could dismantle the thing and see what peltier they using so we could replace it with a more powerful one :D
I opened it up and will be trying a DIY setup in the next few days (if I have time to do it). The good thing about DIYing with this unit is I don't have to worry about powersupply.

[bezz]

Bezz,

Look very interesting. But it's a bit underpowered from what i see from the manual. 6v, 1.5A. Power is 9W .. our peltier normally already running at 30-50w range.


Spinex,

I went out shopping the last few days (actually my GF dragged me for X'mas shopping) and I was able to check out some TEC mini fridge and also compressor mini fridges. Seems most run on about 50-70w. If so, won't running a DIY peltier be too much hassle when one can just get a mini fridge and drill holes for DIY cooling tubes to be inserted?

Anyway, I thinking of doing this :D
Anyone knows the cost of running a mini fridge on 60w? I suspect the power rating is grossly understated... also how is it compared to switching on aircon 24hrs?

Thx

[xtrekker]

350W??? How big tank you have, how low can you cool it? Its quite big TEM :) I'm thinking about 100W for my 82L tank.


350W is the electrical power . The "cooling power" , Qc max is 280W. Did a simulation on the aztec software from Melcor Themoelectrics for my 2ft tank , 54L with the following parameters

Device Cooling with 6mm glass as insulation. Active load of 8W (ehiem 2013 filter). There are several assumptions made.

1. No heat intake from the hoses connecting the filter
2. No heat intake from Lithing equipment.
3. Perfect heat transfer from tank to Cold plate.
4. No heat transfer between hot side and cold side.

The approximate heatload Qc is about 100W . To be on the safe side i chose a TEM that can cool at least twice the heat load require. Dun want to waste my time and effort getting a TEC thats under power. If i find that 350W TEC is over powered is can reduce the cooling capacity by reducing the electrical power supply.

cheers

[xtrekker]

Shrimps aren't sensible to aluminium? If I use aluminium plate to contact peltier and water, I'm afraid aluminium will spread to water. I also saw that some pictures shows using old CPU passives which are coated with black/green/or other color. I have also some here. Is that safe? I'm really afraid of my minions:)

You can get marine grade stainless steel. For my case i am I dun have a bench drill. Working on aluminium is easier for me .
I am not sure whether shrimps are sensitive to AL .

cheers

[spinex]

Spinex,

I went out shopping the last few days (actually my GF dragged me for X'mas shopping) and I was able to check out some TEC mini fridge and also compressor mini fridges. Seems most run on about 50-70w. If so, won't running a DIY peltier be too much hassle when one can just get a mini fridge and drill holes for DIY cooling tubes to be inserted?

Anyway, I thinking of doing this :D
Anyone knows the cost of running a mini fridge on 60w? I suspect the power rating is grossly understated... also how is it compared to switching on aircon 24hrs?

Thx

Running the cooling tube into the mini fridge definitely will be ineffective. As few of us have tried and experience the best is get the cold fin in direct contact with the water you are cooling to be effective.

Running an aircon 24 x 7 .. think you looking to pay at least $70 extra in electricity a month.

[gr81]

Can I use aluminium heatsink with no coating? Just natural aluminium heatsink?
Or is better coated? Then i'm not sure what is coating made of.

[bezz]

Anyone of you have a lower power peltier that you don't want for sale?

Looking for one to work on my 1.5ft tank :D

[silane]

The approximate heatload Qc is about 100W . To be on the safe side i chose a TEM that can cool at least twice the heat load require. Dun want to waste my time and effort getting a TEC thats under power. If i find that 350W TEC is over powered is can reduce the cooling capacity by reducing the electrical power supply.

cheers


Agree with that, with a overpowered, it is not necessary to keep the peltier running 24x7, it can be on/off using themostat or time base.

[gr81]

Agree with that, with a overpowered, it is not necessary to keep the peltier running 24x7, it can be on/off using themostat or time base.
Using thermostat should be better. I don't thing switch it on/off 2xmin of course. But thermostats can be configured to on/off when temperate goes of of range. eg. on at 27° and off at 25-24° then peltier will run always some time and the tetemperature shocks will be lower then using timer.

btw: How efeffective should be, if I but peltier directly on tank glass (more smaller peltiers) and on the other side heatsink and fan? should it work?

[NanoDave]

btw: How efeffective should be, if I but peltier directly on tank glass (more smaller peltiers) and on the other side heatsink and fan? should it work?

glass is a poor conductor of heat so I think you will lose a lot of effectiveness if you stick the peltier on the glass. If you touch the glass with your finger and compare it to the feeling when you touch the water directly, you will find that the water feels a lot more colder (maybe more than 50%!). Therefore, imagine how much the glass would be preventing the peltier from cooling the water. :)

[bezz]

As promised.
I took my can cooler apart and used the peltier to construct a nano cooler. And the results were more than what I expected :D

http://arofanatics.com/members/bezz/peltiercooler/image01.jpg?nocache=1134805655

http://arofanatics.com/members/bezz/peltiercooler/image05.jpg?nocache=1134805655

http://arofanatics.com/members/bezz/peltiercooler/image06.jpg?nocache=1134805655

http://arofanatics.com/members/bezz/peltiercooler/image07.jpg?nocache=1134805655

It got my 1.2ft tank from 30 degree celsius down to about 25 degree celsius, though the Peltier is only rated at 9watts!!! :afro:

I have another idea to lower the temperature by 1-2 degree celsius :D Will also put together a nicer casing for my nano chiller :)

[TEC Supplier]

Hello gents- This is Steve Sykes entering the discussion from Bangkok. I'm an American who runs two businesses that were briefly mentioned in the predecessor thread to this one - busineseses described at thermoelectricsupplier(dot)com, and thaisupercool(dot)com.

I found your forum by tracking back from website "hits" to my TEC supplier website. Your discussion somewhat mirrors similar diiscussions at some tropical fish forums - most notably at killies(dot)com.

I am a businessman - and I would like to earn my company a role providing solutions to your discussion group- which seems to be largely centered in Singapore. I already have a small TEC customer base in Singapore - beer home-brewers.

I would like to point out a number of things:

1) There area lot of different TEC module configurations - with different surfcae areas, different power consumption profiles, and different performance characteristics. In general, a 50 watt, 40 mm TEC1-12705 module is generally the "best" choice for most "consumer appliance" type applications. This is because the results you achieve on the cold side of a TEC are directly related to how cool you can keep the heat sink connected to the hot side of the TEC. Dissipating heat from a TEC module "more powerful than" a 50 watt module requires a quite large heat sink and cooling fan.

2) As I said, the heat dissipation sub-system of your TEC appliance is critical. I'd like to point out some of the differences between the components my company sells, and what you can purchase by "scrounging." First, let's look at cooling fans. In this thread, I saw images of a 12 VDC fan rated at 0.14 amps. That is a low airflow fan - optimized for silence. For cooling, we suggest a 12 VDC, 120 mm, 0.40 amp fan - much better cooling air flow. Always blow INTO the heat sink fins. Next - heat sinks. We use a particular aluminum alloy - AL6063T5 - to ensure optimal heat dissipation. All of the fins on our heat sinks feature a "grooved" ("ridged") surface - both to GREATLY increase heat exchange surface area (by more than 100%, compared to similar smooth fin heat sinks), and to allow cooling fans to be directly and securely mounted by screwing mounting bolts through holes in the fan housing, and threading right into the spaces between the fins.

3) In some of the "exploded" drawings, you will see a square aluminum "spacer block". Ours are 24 mm thick - allowing spacer to penetrate this same thickness of insulation. Along with heat dissipation, insulation is the second key to extracting good performance from TEC modules. To shill any space or thermal mass, you must first isolate and then THOROUGHLY insulate that target, and then use a spacer block to "suck" heat out of the target, through the insulation - TEC module positioned OUTSIDE the insulation, with cold side of TEC attached to spacer block such that this interfcae is EXACTLY flush with the plane where the spacer block exits the insulation. Now, apply power to the TEC - and it will start extracting heat from the insulated target. As the TEC works to pull heat out of the target, replacement heat will continuously seek to re-enter the target, by "leaking" through the insulation. To the extent that you can extract more heat than is re-entering, you thermal target will chill down. So - increasing your insulation will always improve your results.

I would be interested in developing a relationship with one contact in Singapore, to whom I could ship components in small bulk, for "retail" level distribution. As you will find out when you try to source individual items separately (at least internationally), the costs of items themselves are trivial - compartedto the costs of shipping, and of getting payment to the suppliers. Thus - it makes sense to purchase all items from a single source.

I may be contacted at steve(at)pacificsupercool(dort)com.

Cheers!
Steve Sykes
Managing Director
Pacific Supercool Ltd.
thermoelectricsupplier.com

[spinex]

Steve,

Thanks a lot for all the information. It will come handy for members who are interested to DIY their peltier unit for their small (NANO) tank.

[gr81]

question:
If I buy peltier rated Imax 10A, Umax 15.8V, Qmax (?T =0) 89W
and run it on 12V
will it have Imax=10/15,8*12=7,6A? & Qmax=89/15,8*12=67,6W

or is it nonlinear ?

[bezz]

Hello gents- This is Steve Sykes entering the discussion from Bangkok. I'm an American who runs two businesses that were briefly mentioned in the predecessor thread to this one - busineseses described at thermoelectricsupplier(dot)com, and thaisupercool(dot)com.

I found your forum by tracking back from website "hits" to my TEC supplier website. Your discussion somewhat mirrors similar diiscussions at some tropical fish forums - most notably at killies(dot)com.

I am a businessman - and I would like to earn my company a role providing solutions to your discussion group- which seems to be largely centered in Singapore. I already have a small TEC customer base in Singapore - beer home-brewers.

I would like to point out a number of things:

1) There area lot of different TEC module configurations - with different surfcae areas, different power consumption profiles, and different performance characteristics. In general, a 50 watt, 40 mm TEC1-12705 module is generally the "best" choice for most "consumer appliance" type applications. This is because the results you achieve on the cold side of a TEC are directly related to how cool you can keep the heat sink connected to the hot side of the TEC. Dissipating heat from a TEC module "more powerful than" a 50 watt module requires a quite large heat sink and cooling fan.

2) As I said, the heat dissipation sub-system of your TEC appliance is critical. I'd like to point out some of the differences between the components my company sells, and what you can purchase by "scrounging." First, let's look at cooling fans. In this thread, I saw images of a 12 VDC fan rated at 0.14 amps. That is a low airflow fan - optimized for silence. For cooling, we suggest a 12 VDC, 120 mm, 0.40 amp fan - much better cooling air flow. Always blow INTO the heat sink fins. Next - heat sinks. We use a particular aluminum alloy - AL6063T5 - to ensure optimal heat dissipation. All of the fins on our heat sinks feature a "grooved" ("ridged") surface - both to GREATLY increase heat exchange surface area (by more than 100%, compared to similar smooth fin heat sinks), and to allow cooling fans to be directly and securely mounted by screwing mounting bolts through holes in the fan housing, and threading right into the spaces between the fins.

3) In some of the "exploded" drawings, you will see a square aluminum "spacer block". Ours are 24 mm thick - allowing spacer to penetrate this same thickness of insulation. Along with heat dissipation, insulation is the second key to extracting good performance from TEC modules. To shill any space or thermal mass, you must first isolate and then THOROUGHLY insulate that target, and then use a spacer block to "suck" heat out of the target, through the insulation - TEC module positioned OUTSIDE the insulation, with cold side of TEC attached to spacer block such that this interfcae is EXACTLY flush with the plane where the spacer block exits the insulation. Now, apply power to the TEC - and it will start extracting heat from the insulated target. As the TEC works to pull heat out of the target, replacement heat will continuously seek to re-enter the target, by "leaking" through the insulation. To the extent that you can extract more heat than is re-entering, you thermal target will chill down. So - increasing your insulation will always improve your results.

I would be interested in developing a relationship with one contact in Singapore, to whom I could ship components in small bulk, for "retail" level distribution. As you will find out when you try to source individual items separately (at least internationally), the costs of items themselves are trivial - compartedto the costs of shipping, and of getting payment to the suppliers. Thus - it makes sense to purchase all items from a single source.

I may be contacted at steve(at)pacificsupercool(dort)com.

Cheers!
Steve Sykes
Managing Director
Pacific Supercool Ltd.


Hi Steve,

Your website looks very interesting :cool:
To DIY a cooler for a fish tank, is the Compact Cold Plate Assembly the best assembly for the job and does the cold plate corrode in water?

Thank you.

Henry

[silane]

2) As I said, the heat dissipation sub-system of your TEC appliance is critical. I'd like to point out some of the differences between the components my company sells, and what you can purchase by "scrounging." First, let's look at cooling fans. In this thread, I saw images of a 12 VDC fan rated at 0.14 amps. That is a low airflow fan - optimized for silence. For cooling, we suggest a 12 VDC, 120 mm, 0.40 amp fan - much better cooling air flow. Always blow INTO the heat sink fins. Next - heat sinks. We use a particular aluminum alloy - AL6063T5 - to ensure optimal heat dissipation. All of the fins on our heat sinks feature a "grooved" ("ridged") surface - both to GREATLY increase heat exchange surface area (by more than 100%, compared to similar smooth fin heat sinks), and to allow cooling fans to be directly and securely mounted by screwing mounting bolts through holes in the fan housing, and threading right into the spaces between the fins.


Hi Steve,

Understand that radiation is a mode of how heat conducts, polished aluminium like the heatsink has emissivity of 0.05, and Black anodised aluminium has emissivity of 0.7 to 0.9. Would you like to comment on your heatsink emissivity?

[silane]

Thermal conductivity of common metal from Wikipedia

http://en.wikipedia.org/wiki/Thermal_conductivity

Thermal conductivity of other common materials:

Diamond 895-2300
Carbon Nanotubes 1400
Silver 429
Copper 386
Gold 317
Aluminium 237
Brass 120
Platinum 71.6
Iron 80.2
Lead 35.3
Mercury 8.514
Quartz (273K) 6.8-12
Ice (273K) 2.2
Glass 1.35
Wood 0.04 (balsa) - 0.35 (fir)
Styrofoam 0.033
Wool 0.04
Silica aerogel 0.017
Air (100 kPa) 0.0262
Water 0.6062

http://rfcafe.com/references/general/thermal_conductivity.htm

More material in the list. :)

[gr81]

seems to diamond is best. Let's make diamond chiller :)

to your previous:

Understand that radiation is a mode of how heat conducts, polished aluminium like the heatsink has emissivity of 0.05, and Black anodised aluminium has emissivity of 0.7 to 0.9. Would you like to comment on your heatsink emissivity?
how is it possible that anodized aluminium has so much better heat transfer than "natural" AL? Is it possible?

[silane]

how is it possible that anodized aluminium has so much better heat transfer than "natural" AL? Is it possible?


Heat is lost to the air by two mechanisms

Conduction/Convection
Radiation

Maximum radiation of heat occurrs when the surface has maximum emissivity, and that happen when color is dark and rough. Not only anodised Al has good emissivity, dark varnish (thin layer) Al heatsink work better than one that is shiny. Matte black work well too.

Conduction/Convection is not being affected by color. :)

[gr81]

Heat is lost to the air by two mechanisms

Conduction/Convection
Radiation

Maximum radiation of heat occurrs when the surface has maximum emissivity, and that happen when color is dark and rough. Not only anodised Al has good emissivity, dark varnish (thin layer) Al heatsink work better than one that is shiny. Matte black work well too.

Conduction/Convection is not being affected by color. :)
strange, why than all CPU coolers are made of "natural" AL and not "colored". even those for more than $us100. I never heard before.

[silane]

You should have learn that in primary school science class, color matters when come to heat. :)

My guess is, with forced air cooling of heatsink, heat dissipated by radiation is no longer an major contributor.

[TEC Supplier]

RE: Effect of Anodization - We are talking about forced air cooling, using a cooling fan. I cannot yet post messages with URL's - so, see

aavidthermalloy.Xom/products/extrusion/anodize

and add htt....ww. ahead, and .shtml at end.

The main corrolary of anodization is increasing surface area - we do this by adding extruded ridges (ripples) on all fin surfaces.

[xtrekker]

Hello gents- This is Steve Sykes entering the discussion from Bangkok. I'm an American who runs two businesses that were briefly mentioned in the predecessor thread to this one - busineseses described at thermoelectricsupplier(dot)com, and thaisupercool(dot)com.

I found your forum by tracking back from website "hits" to my TEC supplier website. Your discussion somewhat mirrors similar diiscussions at some tropical fish forums - most notably at killies(dot)com.

I am a businessman - and I would like to earn my company a role providing solutions to your discussion group- which seems to be largely centered in Singapore. I already have a small TEC customer base in Singapore - beer home-brewers.

I would like to point out a number of things:

1) There area lot of different TEC module configurations - with different surfcae areas, different power consumption profiles, and different performance characteristics. In general, a 50 watt, 40 mm TEC1-12705 module is generally the "best" choice for most "consumer appliance" type applications. This is because the results you achieve on the cold side of a TEC are directly related to how cool you can keep the heat sink connected to the hot side of the TEC. Dissipating heat from a TEC module "more powerful than" a 50 watt module requires a quite large heat sink and cooling fan.

2) As I said, the heat dissipation sub-system of your TEC appliance is critical. I'd like to point out some of the differences between the components my company sells, and what you can purchase by "scrounging." First, let's look at cooling fans. In this thread, I saw images of a 12 VDC fan rated at 0.14 amps. That is a low airflow fan - optimized for silence. For cooling, we suggest a 12 VDC, 120 mm, 0.40 amp fan - much better cooling air flow. Always blow INTO the heat sink fins. Next - heat sinks. We use a particular aluminum alloy - AL6063T5 - to ensure optimal heat dissipation. All of the fins on our heat sinks feature a "grooved" ("ridged") surface - both to GREATLY increase heat exchange surface area (by more than 100%, compared to similar smooth fin heat sinks), and to allow cooling fans to be directly and securely mounted by screwing mounting bolts through holes in the fan housing, and threading right into the spaces between the fins.

3) In some of the "exploded" drawings, you will see a square aluminum "spacer block". Ours are 24 mm thick - allowing spacer to penetrate this same thickness of insulation. Along with heat dissipation, insulation is the second key to extracting good performance from TEC modules. To shill any space or thermal mass, you must first isolate and then THOROUGHLY insulate that target, and then use a spacer block to "suck" heat out of the target, through the insulation - TEC module positioned OUTSIDE the insulation, with cold side of TEC attached to spacer block such that this interfcae is EXACTLY flush with the plane where the spacer block exits the insulation. Now, apply power to the TEC - and it will start extracting heat from the insulated target. As the TEC works to pull heat out of the target, replacement heat will continuously seek to re-enter the target, by "leaking" through the insulation. To the extent that you can extract more heat than is re-entering, you thermal target will chill down. So - increasing your insulation will always improve your results.

I would be interested in developing a relationship with one contact in Singapore, to whom I could ship components in small bulk, for "retail" level distribution. As you will find out when you try to source individual items separately (at least internationally), the costs of items themselves are trivial - compartedto the costs of shipping, and of getting payment to the suppliers. Thus - it makes sense to purchase all items from a single source.

I may be contacted at steve(at)pacificsupercool(dort)com.

Cheers!
Steve Sykes
Managing Director
Pacific Supercool Ltd.
thermoelectricsupplier.com


hi steve welcome to this discussion . your information is pretty useful .

I agreed with you .... big heat sinks , powerful fans and Good insulation is the key to success.

which TEC from your company would you recommend if i were to cool a heatload of 100W from 29C degrees to 24 degrees ?

perharps you could attach some datasheet showing the COP , Qc vs I . Uv vs I or DeltaT for fellow forumers that wish to design their chiller.

cheers
xtrekker

[gr81]

So finally I have all parts and lets start :)

One more question: I will use small 6L tank as cooling reservoir. Height of tank is about 15cm and height of cooler part in watter is 5cm. Is better lower height of tank so watter will flow through cooler fins directly. Or is better let bigger space for watter, watter will be longer in cooler and maybe cool down more.

[xtrekker]

So finally I have all parts and lets start :)

One more question: I will use small 6L tank as cooling reservoir. Height of tank is about 15cm and height of cooler part in watter is 5cm. Is better lower height of tank so watter will flow through cooler fins directly. Or is better let bigger space for watter, watter will be longer in cooler and maybe cool down more.

why go through the trouble of using a reservior ? Your are cooling extra 6L of water which is not a very good idea.

cheers
xtrekker

[bezz]

Joejoe,

I noticed your very nicely designed acrylic box in an earlier post, please tell how you managed to cut the acrylic so perfectly?

Thank you

[silane]

As promised.
I took my can cooler apart and used the peltier to construct a nano cooler. And the results were more than what I expected :D

It got my 1.2ft tank from 30 degree celsius down to about 25 degree celsius, though the Peltier is only rated at 9watts!!! :afro:

I have another idea to lower the temperature by 1-2 degree celsius :D Will also put together a nicer casing for my nano chiller :)


Hi Bezz,

I did a simulation using software, the cooling power (Qc) required by a glass tank of 12x8x8" requires 31.50W to bring from 30 to 25. Your peltier is driven by 9W, I wonder if it is because the fan has aided in cooling the water by evaporating it?

[bezz]

Hi Bezz,

I did a simulation using software, the cooling power (Qc) required by a glass tank of 12x8x8" requires 31.50W to bring from 30 to 25. Your peltier is driven by 9W, I wonder if it is because the fan has aided in cooling the water by evaporating it?

Actually, I have no idea, but the water level is holding steady the pass 4-5 days of using it. Temperature is also holding at about 25C.

Also my tank is 14x8x8.

[gr81]

Hi Bezz,

I did a simulation using software, the cooling power (Qc) required by a glass tank of 12x8x8" requires 31.50W to bring from 30 to 25. Your peltier is driven by 9W, I wonder if it is because the fan has aided in cooling the water by evaporating it?
I'm not sure if simulation software can give serious results. Efficiency of cooling depends on so many factors (volume of heatsings, cooling of heatsinks, "warming" those in watter) does software take it?

[silane]

I'm not sure if simulation software can give serious results. Efficiency of cooling depends on so many factors (volume of heatsings, cooling of heatsinks, "warming" those in watter) does software take it?


That software calculates the power need to cool down a tank of liquid with consideration of tank surface losses. The figure you get is Qc, and with Qc, you select the right peltier that can provide and heatsink with proper thermal resistance.

The previous calculation shows that a tank of that size, even 6 sided glass would require 35W even without active load (pumps) heat consideration. So make me query how can a 9W electrical power cool that tank at 30C? Note that Qc is lower than electrical power, which mean Qc of that peltier is less than 9W.

[gr81]

That software calculates the power need to cool down a tank of liquid with consideration of tank surface losses. The figure you get is Qc, and with Qc, you select the right peltier that can provide and heatsink with proper thermal resistance.

The previous calculation shows that a tank of that size, even 6 sided glass would require 35W even without active load (pumps) heat consideration. So make me query how can a 9W electrical power cool that tank at 30C? Note that Qc is lower than electrical power, which mean Qc of that peltier is less than 9W.
It's really strange that 9W can cool it. By previous post I mean, that software can calculate eg. you need 50W, but with improper cooling of heatsinks and other parameters, can be even 100W too small power. Due to peltiers will simply not work.

[silane]

It's really strange that 9W can cool it. By previous post I mean, that software can calculate eg. you need 50W, but with improper cooling of heatsinks and other parameters, can be even 100W too small power. Due to peltiers will simply not work.


True, improper cooling screw things up. The software just tell how much cooling power you need, and not how to cool it.

[xtrekker]

True, improper cooling screw things up. The software just tell how much cooling power you need, and not how to cool it.

Agreed with fruitepie , its just a estimated heatload. I would not go into the details of Themodynamics. In my previous post , my estimated heatload (Qc) is about 100W. I decided that the simple way is to trial and error by getting a TEC that can support twice the heatload of my tank. Adjusting the input power later to get my desired temperature.

The software is useful to esitmate the heatload so that you wouldn't get a under power TEM.

cheers
xtrekker

[gr81]

Question related to chiller:)
To power my peltiers I need 40A@15.4V.
1) Is there a chance to get more that 12V using PC power supply? Max is 12V, can it be somehow connected to get 15.4? I mean 12+3.3v e.g.
2) I have 3 supplies, 15A@12V each. Can I put 12V paralelly together and get 45A?

[silane]

gr81,

1) I think the grounds are common, so you cannot connect 12V and 3.3V trail in series. Use a multimeter to do a check, I could be wrong. You can drive it using 12V, the max current drawn is (12/15.4)*40.

2) Yes.

[xtrekker]

Question related to chiller:)
To power my peltiers I need 40A@15.4V.
1) Is there a chance to get more that 12V using PC power supply? Max is 12V, can it be somehow connected to get 15.4? I mean 12+3.3v e.g.
2) I have 3 supplies, 15A@12V each. Can I put 12V paralelly together and get 45A?

yup you can . But make sure that you PSU have the same ratings. Or alternative just add a diode at each 12V terminal. Bear in mind that diode has a forward volatage drop of 0.7V ..


cheer
xtrekker

[gr81]

yup you can . But make sure that you PSU have the same ratings. Or alternative just add a diode at each 12V terminal. Bear in mind that diode has a forward volatage drop of 0.7V ..


cheer
xtrekker
Why do that? (diode), & why PSU have to be same?

[xtrekker]

Why do that? (diode), & why PSU have to be same?

if PSU 1 is rated at 12V 10A , PSU 2 rated at 12V 15A . With PSU 2 higher current rating it will attempt to charge psu 1. With diodes in series with the 12V line , current would only flow toward the TEC . A more efficient way is to implement a load sharing circuit.

I had "fried" a couple of PSU before i found out what happen.

ps: always Get a PSU rated at least 125% more then your TEC required.

cheers
xtrekker

[gr81]

if PSU 1 is rated at 12V 10A , PSU 2 rated at 12V 15A . With PSU 2 higher current rating it will attempt to charge psu 1. With diodes in series with the 12V line , current would only flow toward the TEC . A more efficient way is to implement a load sharing circuit.

I had "fried" a couple of PSU before i found out what happen.

ps: always Get a PSU rated at least 125% more then your TEC required.

cheers
xtrekker
Great info, thanks a lot
Will try to get one big PSU :)

[ianlim]

Hi Spinex,

Have you ripped apart your peltier reservior tank and tried with a bigger peltier? I have the big monster peltier fromEbay, 60mmx60mm, 320W! you wanna to try it? Ian Lim

[gr81]

Hi Spinex,

Have you ripped apart your peltier reservior tank and tried with a bigger peltier? I have the big monster peltier fromEbay, 60mmx60mm, 320W! you wanna to try it? Ian Lim
I'm pretty sure, that using of so big peltier has no sense. It's nearlly impossible cool down this creature. U will need very big and noisy fan.
Much more better i using of more smaller petliers. This way you risk, that it will run on low W. Never mind it has theoretical 320W,with improper cooling you can get less than 100W

[spinex]

I'm pretty sure, that using of so big peltier has no sense. It's nearlly impossible cool down this creature. U will need very big and noisy fan.
Much more better i using of more smaller petliers. This way you risk, that it will run on low W. Never mind it has theoretical 320W,with improper cooling you can get less than 100W

What Ian can do is just power the peltier to be 100w-150w to achieve the desired cooling. 320w is really an overkill and will be difficult to cool down like you mentioned.

A big heat sink and powerful fan is needed now :D

[ianlim]

What Ian can do is just power the peltier to be 100w-150w to achieve the desired cooling. 320w is really an overkill and will be difficult to cool down like you mentioned.
A big heat sink and powerful fan is needed now :D

hi Spinex,

I am going for 100w-150w, tried 240w but the heat generated is too hot, I had got the heatsink and over the weekend will try to finish the full assembly.
thank you for clearing my doubt yesterday.

Ian Lim

[xtrekker]

hi Spinex,

I am going for 100w-150w, tried 240w but the heat generated is too hot, I had got the heatsink and over the weekend will try to finish the full assembly.
thank you for clearing my doubt yesterday.

Ian Lim

Conventional Heatsink you get from SLT dun really work for big TEC. I am using a 320W Water cooled TEC. ie using pumped water to transfer the heat from the hot side to a 30cm x 12cm x 5cm radiator.Driving my TEC at 250W , i managed to cool my 60cm x 30cm x 30cm tank to 19.7C overnight .

cheers
xtrekker

[spinex]

Conventional Heatsink you get from SLT dun really work for big TEC. I am using a 320W Water cooled TEC. ie using pumped water to transfer the heat from the hot side to a 30cm x 12cm x 5cm radiator.Driving my TEC at 250W , i managed to cool my 60cm x 30cm x 30cm tank to 19.7C overnight .

cheers
xtrekker

Wow .. impressed with your setup. Really extreme :shocked: and looked very professionally done.

Any photo of your setup with the radiator ?

So 19.7C is the constant temp you will achieve using the TEC ? Suppose you step down to 200W and get a desired 23-25 degree won't it cost more for your electricity bill since you need to run it 24x7 ?

Or i assume if you use a thermostat to control the TEC to kick in, i read that the best way to control temperature using TEC is regular the power being supplied to the TEC rather than just on/off it because this will cause the TEC to be spolit easily. How true is that ?

Whatever is it this is really a very impressive DIY TEC chiller. If you have a article on the setup definitely deserve to be in the "LIBRARY" section.

[gr81]

I don't know why thermostat can burn out peltier. Of course if it will on/off twice a min. there is some risk. But if it will run eg 30min. and then 60min. will off, it can't be problem.

[ianlim]

Conventional Heatsink you get from SLT dun really work for big TEC. I am using a 320W Water cooled TEC. ie using pumped water to transfer the heat from the hot side to a 30cm x 12cm x 5cm radiator.Driving my TEC at 250W , i managed to cool my 60cm x 30cm x 30cm tank to 19.7C overnight .

cheers
xtrekker

wow Xtrekker, really impressed with the DIY chiller kit you did, where u bought the cold plate and water block, did u diy the 4 holding screw for the cold plate?

is it cost effective to run a 100w or 200w peltier for 24hrs daily, how much will it cost us monthly? is it cheaper in the long run if we get a Resun CL280?

[silane]

Conventional Heatsink you get from SLT dun really work for big TEC. I am using a 320W Water cooled TEC. ie using pumped water to transfer the heat from the hot side to a 30cm x 12cm x 5cm radiator.Driving my TEC at 250W , i managed to cool my 60cm x 30cm x 30cm tank to 19.7C overnight .

cheers
xtrekker


That's very impressive to manage that large amount of heat.

Do you soak the cold plate into the tank of water?

Writing an article on it? I will put it in Library.

[xtrekker]

Any photo of your setup with the radiator ?
>>None at the moment . Will try to get pic of the whole setup . My Digtal cam just broke down.

So 19.7C is the constant temp you will achieve using the TEC ? Suppose you step down to 200W and get a desired 23-25 degree won't it cost more for your electricity bill since you need to run it 24x7 ?

>>19.7C was achieved during my first testing.I wwas driving the TEC at 12V. I belive lower temp can be reached if i drive the TEC at Vmax of 15V. Have to look at your TEC data sheet for proper powering up of your TEC. I had burn up numerous computer PSU before i realise the mistake. Electrical bill is not a concern at the moment. Will implement a PID temperature controller with setpoint of 24C when i have the time.

Or i assume if you use a thermostat to control the TEC to kick in, i read that the best way to control temperature using TEC is regular the power being supplied to the TEC rather than just on/off it because this will cause the TEC to be spolit easily. How true is that ?

>>Ya. True TEC is not very torelant of Switching. Check the Data sheet of the switching frequency. Implementing a PWM Driver is better.. its becoming a more like final year project aredi...hahaha. Dun worry about switching the TEC on and off . Your are cooling water. Water have a high specific heat capacity so it won't lost or gain heat fast. ie your tec will not be switching too.

cheers
xtrekker

[xtrekker]

I don't know why thermostat can burn out peltier. Of course if it will on/off twice a min. there is some risk. But if it will run eg 30min. and then 60min. will off, it can't be problem.

totally agreed with gr81. Switching frequency in this aquarium chiller is too low to be of a concern.

cheers
xtrekker

[xtrekker]

wow Xtrekker, really impressed with the DIY chiller kit you did, where u bought the cold plate and water block, did u diy the 4 holding screw for the cold plate?

>>DIY it myself. Screws , cold plate from hardware shop . arcylic left overs from other diy projects. Bulkhead from SLT.


is it cost effective to run a 100w or 200w peltier for 24hrs daily, how much will it cost us monthly? is it cheaper in the long run if we get a Resun CL280?[/QUOTE]

>> NO . TEC is the least efficient comparing to stirling coolers and phase change cooling. Google it

Assuming the no energy lost, cost of running 200W TEC (always on) is ,

0.2KW x 24 x 7 x 4 x $0.2 = $26.88

cheers
xtrekker

[xtrekker]

That's very impressive to manage that large amount of heat.

Do you soak the cold plate into the tank of water?

>>Yup .suspended the cold plate in the tank during my trial run.

Writing an article on it? I will put it in Library.

>> OK. Will try to do it .

[ianlim]

Actually, I have no idea, but the water level is holding steady the pass 4-5 days of using it. Temperature is also holding at about 25C.

Also my tank is 14x8x8.

hi Bezz, u still running the modded setup? can it run for 24/7? any problem u encountered so far eg power socket overheating etc.. do the temp. fluctuate much? tia for the update.

[bezz]

hi Bezz, u still running the modded setup? can it run for 24/7? any problem u encountered so far eg power socket overheating etc.. do the temp. fluctuate much? tia for the update.

On hot days it goes to 26C.
Cold days, drop to 24C.
Otherwise at around 25C.

Running 24hrs in my room. I changed the fan to a lower power one to reduce the noise.

[ianlim]

On hot days it goes to 26C.
Cold days, drop to 24C.
Otherwise at around 25C.

Running 24hrs in my room. I changed the fan to a lower power one to reduce the noise.

hi bezz,

u changed the original fan and added another rite?

are the dc supply for the fan 'tap from the original power adaptor or it is from a separate source?

what are the ambient temp. for the 3 readings quoted above. tia

[bezz]

hi bezz,

u changed the original fan and added another rite?

are the dc supply for the fan 'tap from the original power adaptor or it is from a separate source?

what are the ambient temp. for the 3 readings quoted above. tia

I have another 2 tanks beside this one, so I just tap the power from the adaptor I use to power the fans for the 2 tanks.

I'm not sure about ambient temperature, but my 2 ft tank cooled with fan is at 28C constant.


Thx

[CandieG]

Are there no new improvements or experiments on DIY chiller these days? :D

[xtrekker]

Are there no new improvements or experiments on DIY chiller these days? :D


Its in the shrimpnow library .

http://www.shrimpnow.com/modules.php?name=Content&file=viewarticle&id=43

cheers
xtrekker

[Mraz]

Very informative write-ups on TEM xtrekker. Be waiting for Part2.

where can I purchase a TEM model CP1-12705, heatsinks and PSU? And you mentioned
"choosing a higher rated TEM at 77W to give a some excess cooling power to play with. TEM can easily turn off when the set point temperature of 24C is reached." Do I need a temp controller to achieve this?

Thanks alot.

[gr81]

I just finished new/first chiller. when I switch it on, hot plate runs hot and fan cools it to mid. hot (approx 30°c). if I switch fan off, heatsink goes untouchable hot in short time. In test tank with pump (6 L only) I have constantly 22°c (same as room temperature). 2m pipes are connecting this two parts. It runs more then a hour and still no temp. lost. what should be wrong? All seems fine.
there runs 2 TEM 68W each.

[gr81]

I just finished new/first chiller. when I switch it on, hot plate runs hot and fan cools it to mid. hot (approx 30°c). if I switch fan off, heatsink goes untouchable hot in short time. In test tank with pump (6 L only) I have constantly 22°c (same as room temperature). 2m pipes are connecting this two parts. It runs more then a hour and still no temp. lost. what should be wrong? All seems fine.
there runs 2 TEM 68W/8A each.
tems works fine and are correctly connected, PSU 25A@12v.

[xtrekker]

Very informative write-ups on TEM xtrekker. Be waiting for Part2.

where can I purchase a TEM model CP1-12705, heatsinks and PSU? And you mentioned
"choosing a higher rated TEM at 77W to give a some excess cooling power to play with. TEM can easily turn off when the set point temperature of 24C is reached." Do I need a temp controller to achieve this?

Thanks alot.

sorry for the late reply i was away for the weekend. ...

some introduction on reading of the TEM part number ...
for example model XXX- 12705

XXX is the manufactuer's code.
127 means that the TEM has 127 peltier Junctions in that module
05 means its max current of 05 Amperes

another Example model XXX-12830

XXX is the manufactuer's code.
128 means that the TEM has 128 peltier Junctions in that module
30 means its max current of 30 Amperes

>>Do I need a temp controller to achieve this?

yup . a simple temperature control circuit with a solid state or solonoid relay can achieve this.

>>where can I purchase a TEM model CP1-12705, heatsinks and PSU?
In singapore , you can got to Sim Lim Towers to get them. You can get XXX-12705 and XXX-12708 there. I order my CP1-12730 from ebay merchant called thermal entreprise.

cheers
xtrekker

[xtrekker]

I just finished new/first chiller. when I switch it on, hot plate runs hot and fan cools it to mid. hot (approx 30°c). if I switch fan off, heatsink goes untouchable hot in short time. In test tank with pump (6 L only) I have constantly 22°c (same as room temperature). 2m pipes are connecting this two parts. It runs more then a hour and still no temp. lost. what should be wrong? All seems fine.
there runs 2 TEM 68W each.

emmm ..... your 2m pipe might be the culprit . if you work out the surface area of the pipe its quite significant. Try reducing the length of the pipe or insulated it. I must stress Insulation is the key issue when building this mini chiller.

cheers
xtrekker

[Mraz]

sorry for the late reply i was away for the weekend. ...

some introduction on reading of the TEM part number ...
for example model XXX- 12705

XXX is the manufactuer's code.
127 means that the TEM has 127 peltier Junctions in that module
05 means its max current of 05 Amperes

another Example model XXX-12830

XXX is the manufactuer's code.
128 means that the TEM has 128 peltier Junctions in that module
30 means its max current of 30 Amperes

>>Do I need a temp controller to achieve this?

yup . a simple temperature control circuit with a solid state or solonoid relay can achieve this.

>>where can I purchase a TEM model CP1-12705, heatsinks and PSU?
In singapore , you can got to Sim Lim Towers to get them. You can get XXX-12705 and XXX-12708 there. I order my CP1-12730 from ebay merchant called thermal entreprise.

cheers
xtrekker

Great thanks! :)

[NULL]

Hi, I got some experience in industrial cooling/heating at work, and in computercooling as a hobby.
I don't say I can everything about cooling, you still have to experiment, but I can likely help when someone are theoretical wrong.
So it's a pity I didn't joined this forum earlier when I read about all the research and work you have put in to this.

The site Heatsink Guide (http://www.heatsink-guide.com/peltierinfo.shtml) has this article about TEC's.

But there's some more stuff to think of:
- The TEC, heatsink, water, thermal compounds and air are just transportmedia or exchangesurfaces for that indestructible heat. It will not just disappear, and have to go through the weakest link.
- Thermal conductivity(transportmedia), heat transfer(exchangesurfaces) and specific heat capacity(transportmedia) are not the same thing in designing.
Thermal conductivity is how heat is flowing through the material.
(You can transport heat much further in a copper fin, than in a steel fin)
Specific heat capacity is how much energy a material can hold.
(A gallon water can hold much more energy than a gallon air)
Heat transfer is how heat is flowing from one material to another.
(It’s easier for heat to escape from an aluminium fin into air, than from a copper fin into air)

- You can move the heat, or you can move the media.
- It's great difference in values between water and air, but they work the same.
- If done wrong, TEC just like air-condition will just produce heat, in the same way.

[NULL]

gr81: is your chiller still non-working?
In the spirit of "To be able to help in troubleshooting, some symptoms are required.", maybe you have some picture to show the system?

[Redza]

that gave me an idea... of having cooling fins at the back of the tank with a an ontop.
will it work?

[craigthor]

Great read I jsut spent the day reading both of these threads. Lots of great info for DIY. I think I'm going to break down and just buy a peltier based Resun though as time is short with a month old baby.

Craig