just had a quick read of this
http://www.gigabyte.com/MicroSite/185/on-off-charge.htm
Is that what your talking about? Do you want to stop it charging when system is off or just stop it changing to quick charge when system is on?
can you uninstall the driver, it says at the bottom of that link that you need a driver to use it so i say uninstall it if you don't need it?
nah that didn't work :-(Originally Posted by Liam_G
Are these of any interest or help to you for your controller:
http://www.meerstetter.ch/products/TEC-1122/index.html
http://www.meerstetter.ch/products/T....Datasheet.pdf
http://signalllc.com/products/Si5HyUdPTC1-30V-20A.html
Both types have a max of 20amps so could be good for 12v. The 245w Qmax TEC I'm looking at http://www.customthermoelectric.com/...W_spec_sht.pdf needs about 18amps with 12v for ~40*dt, so could use both outputs in parallel on the dual TEC controller or just use the single TEC controller to run 6 tecs to cool a 420watt heat load @ 40*dT. Could it handle 6 TEC's? Not sure how that works with TEC's in series, If it needs 18amps at 12v does that mean it can only run 216watts worth of TEC's?
err no
thats 18 amps per tec so you would need 6 of those controllers and your TEC's would be drawing a combined total of 1300watts. Also the 40 Delta is in perfect conditions and is never actually possible
ok, i'm a bit confused about the amps thing. i understand i will be running 6 tecs at 12v and current is 18amps so 216watts per TEC, so does that mean i need a PSU that can supply 1296watts and 108amps? I thought that with the TEC's in series it might work by supplying 12v and each TEC would use 216watts at 18amps but that the current would remain at 18amps because they were in series? Clearly i'm over my head but thats the only way i could explain how you were running 4 x 245Qmax TEC's on your controller, because i thought you were using mosfets that could only handle 30amps? I must have missed something in the thread, I'm learning here so don't laugh to hard at me, I knew next to nothing about TEC's or electrics a few days ago, still know next to nothing :P
EDIT: Also if i revise it to 30*dT could I use 4 TEC's at 12v 18amps on your controller? I'm not really sure where your up to with the low pass filter converting PWM to analog, because that would be ideal for controlling voltages if it can handle 72amps.
Also 2 rx480's radiators can dissipate 1600watts at a 10* water to air delta with 8x1850rpm GT's, thats why i was going with 6 tecs 40*dT.
Last edited by Liam_G; 05-19-2012 at 02:15 AM.
it's all good man we all have to start some where
To overly simplify things we will assume the resistance of the TEC is constant at 0.6666ohms
6 tec in parallel will mean each TEC will receive 12 volts and therefore pull 18 amps each. so a total of 1300watts but will achieve a delta of 40c
6 Tec in series will mean each TEC will receive 2 volts and therefore for pull 6 watts for a total of 36watts but obviously the delta they will achieve will be very low AND the volume of heat will be small too
This is ohms law. however never use it when it comes to TEC
I have does some else to the above sort of
Last edited by Ultrasonic2; 05-19-2012 at 02:01 AM.
ok, maybe i got confused in my terminology, how many amps do 6 tecs at 12v/1296watts in parallel use? 108?
so were you running 4 tecs and controlling 24 volts, so supplying each tec with 6 volts and it would pull about 8 amps and use 48watts? and with a 20*dT dissipate 60watts per tec?
Also i edited my previous post so have a read of that too.
Last edited by Liam_G; 05-19-2012 at 02:14 AM.
ok, maybe i got confused in my terminology, how many amps do 6 tecs at 12v/1296watts in parallel use? 108?
yes 108 amps
so were you running 4 tecs and controlling 24 volts, so supplying each tec with 6 volts and it would pull about 8 amps and use 48watts? and with a 20*dT dissipate 60watts per tec?
im currently applying 27volts but this controller will apply less than that
but yes i will be applying 6 volts so assuming the resistance is still 0.666 ohms (which it's not) then EACH TEC will draw 9 amps for 53watts so a total of 212 will need to be supplied to by the PSU each tec can move about 60 watts so thats 240watts
and for the record this is awesome efficiency ( actually COP not efficiency ) can move 240 watts while only using 212watts of electricity
EDIT: Also if i revise it to 30*dT could I use 4 TEC's at 12v 18amps on your controller? I'm not really sure where your up to with the low pass filter converting PWM to analog, because that would be ideal for controlling voltages if it can handle 72amps.
No mine couldn't but with some changes it could. however You'd be a crazy person to do that.
What you'd want to do is apply 48volts so the current was 1/4 that of a system running 12 volts. this is what i do well i run 24 -27volts
why would it be crazy to run 72amps. Heat/fire risk? Power consumption? If its a heat or fire risk i'd add the resistor or whatever gets hot into my water cooling loop and if its a power consumption thing, well the system would rarely be at 100% load so the tecs would rarely be pulling 72amps, and I don;t leave my computer on 24/7. Just want to know why its crazy so i can discard the idea.
I need to figure out a way to get at least 30*dT and dissipate ~400watts, so whatever gets me there is what i will do.
Alright, here is some more realistic numbers that might work with your controller while still using 12v.
255.3w TEC x 8 http://www.customthermoelectric.com/...W_spec_sht.pdf
3.5a @ 12v = 42 watts (55watt Qmax @ 20*dT)
42w + 55w Qmax = 97watts heat load x 8 = 776w heat load to be dissipated
2 x RX480 radiators with 8 x GT AP-15's @ 1850rpm can dissipate 800watts at >5*dT
Ambient = 20* + 5*dT = 25* Th
25* - 20*dT = 5* Tc
42w x 8 = 336w power consumption for 440w Qmax which should hold a cold side temp of 5*
And it only uses 28amps so should be fine with your controller.
Sound better?
Glad to see you back into the TEC realm again Ultrasonic. It seems we've got some good ideas floating around here.
If your going for ambient temps you want a 2:1 ratio,recommended 2.3:1
so 336w tec should move close to 800w.The Qmax should be more than twice the load for good temp,I would use a lot of tecs and under volt like crazy try 48 15v@ 5.Quality 1500w psu just for chiller would be needed or just use separate loop for gpu and adjust temp goal.There is some water chillers on ebay you can get a 12/1500w and plug up hoses much easier not as fun
Last edited by Hell Hound; 05-19-2012 at 12:31 PM.
Hell Hound, your posts never really make a lot of sense to me. 48 TEC chiller, seriously? that could be about $1000 worth of TEC's let alone the size/cost of the chiller blocks to cool it. I am also trying to fit all this inside my custom case, ive got 600x575x260mm for the whole PC, so far with 6-10 tecs that is fine.
My GPU's are not being chilled, they are in the hot side loop. I also think that 2.3:1 is a bit off. I'll stick to the numbers on the TEC performance graph and revise from there, i haven't been taking thermal resistance of the chiller blocks into account yet so need to get back closer to my target 30*dt.
EDIT: Double Post
http://www.ebay.com/itm/12709-136-8W...item484544df61
136.8W TEC Thermoelectric CoolerPeltier 12V 50PCS
300 usd free shipping first search on ebay,Probably can find some better ones for around 10$ per tec sounds reasonable.
If you can find performance numbers for that then great. i don't really trust ebay for TEC's, they are most likely sh#t! Even then a 48 TEC chiller block is huge.... I prefer to get my performance data/cost from http://www.customthermoelectric.com/tecs_vmax.htmlhttp://www.ebay.com/itm/12709-136-8W...item484544df61
136.8W TEC Thermoelectric CoolerPeltier 12V 50PCS
300 usd free shipping first search on ebay,Probably can find some better ones for around 10$ per tec sounds reasonable.
What delta T are you able to cool to with 336watts of power and 800watt Qmax? 0-10*? whats the point. This is the closest TEC i can find to that one on ebay http://www.customthermoelectric.com/...Q_spec_sht.pdf what i'm seeing is 720watts Qmax at 20*dT using 600watts of power, thats at 5v pulling 2.5amps/TEC
Last edited by Liam_G; 05-19-2012 at 07:33 PM.
I'm saying serial tec total wattage and qmax not one 300w+ tec.So undervolting multi tecs to achieve 2:1..Cpu + 2/3 gpu is @ least 750w add 750w of tecs so if qmax is not high enough temps will rise past desired goal.I think 5c above ambient loaded will require high ratio.My block dissipates 200w and rad 400w cpu output is 250w after 8 hr prime temps climb to 65c,I call that a gaming session.If I want lower temps I have to raise cooling ratio so adding 250w of tecs with a 800/1000qmax of cooling power and switching to a 480 rad would likely drop temps 20/30c.
yeah, and thats what i was saying, with 48 tecs with that TEC you linked would be about 600watts power usage at 5v which gave me a 720watt Qmax at a 20*dT... I don't think you are going to get 2:1 ratio of power to qmax if you want to achieve any sort of good delta T.
And i'll say it again, i'm not chilling GPU's. I only need about 400watts Qmax.
After further discussion with other members i have revised again to using 6 x 245watt TEC again http://www.customthermoelectric.com/...W_spec_sht.pdf
If i get a SE 600w-24v Meanwell PSU i can run 2 parallel sets of 3 x 245w TEC's in series each receiving 8v and pulling 11amps across the 3.
This will give me a 400w Qmax at a delta T of 30* and use 528watts of power, so not as efficient as using the 8 x 255w TEC's but better delta, once I take into account 0.2 c/w thermal resistance across the chiller blocks and delta T of radiator i get about 20-22* real delta T from ambient. So if ambient is 20* i can hold 0-5* at full CPU load.
I guess if you did similar sets in parallel/series of the 48 tecs you could achieve similar but i would never go there!
Well there is a difference between theory and reality while the stats say the MOSFET N-CH 40V 327A can have a constant draw of 327amp's the reality is it would have caught on fire long before that. I mean it's legs are 1.5mm of aluminium there is no way it could ever pass 327 amps.
This is why you'd be better of with a mosfet for each TEC. this wont cost to much they are like 5$ each
No they dont make sense to me either
When he says a ratio of 2:1 he means raw Qmax of the TEC's ie 245 * 6 = 1470Qmax
YES DO THIS end of story ( assuming your maths is correct )
by the way you said your trying to move 400Watts of heat ? but your CPU wont ever put out that much will it ?
By your math I see 75w more usage but 320w Qmax gain.
The more expensive tecs would prob net better gain w/ less tecs,36 maybe.
@ Ultrasonic..
Well there is a difference between theory and reality while the stats say the MOSFET N-CH 40V 327A can have a constant draw of 327amp's the reality is it would have caught on fire long before that. I mean it's legs are 1.5mm of aluminium there is no way it could ever pass 327 amps.
This is why you'd be better of with a mosfet for each TEC. this wont cost to much they are like 5$ each
ah well, moved on from that now
No they dont make sense to me either
haha, glad i'm not alone on that
When he says a ratio of 2:1 he means raw Qmax of the TEC's ie 245 * 6 = 1470Qmax
I hope that's what he means, i can't really make sense of it.
YES DO THIS end of story ( assuming your maths is correct )
I'm glad you like this idea, means i can stop stuffing around and just roll with the idea for now. I'm pretty sure the maths is correct, worked it out with another forum member.
by the way you said your trying to move 400Watts of heat ? but your CPU wont ever put out that much will it ?
well, yes technically it won't but i want headroom just in case. a 3930k at 4.5ghz uses about 350watts of power, so maybe 300watts of heat or so, total guesstimation there :P, i want a little bit of room to move because this is a power hungry chip and with 5+ghz OC it will use about 400watts power. So with a bit of overkill built in i will be safe and can either get a better delta T or use less power in the real world, win win situation.
Last edited by Liam_G; 05-19-2012 at 10:53 PM.
75w more power at 320Qmax gain but less delta T...... Delta T, do you know what that is????? stop thinking about total watts dissipated and start thinking how much difference between your hot and cold side. once all the thermal resistance of your blocks and rad is taken into account your only getting about 10*dT, so your using more power than me but chilling the cold side less... for 600watts of power usage thats not enough gain for that to be worthwhile for me. granted you have GPU's in your cold side and you can take care of that load but i just don't see the point to it all. thats a sh%t load of cost and power just to keep your case a little cooler.By your math I see 75w more usage but 320w Qmax gain.
The more expensive tecs would prob net better gain w/ less tecs,36 maybe.
EDIT: actually 720Qmax + 600watts of power usage is a 1320watt heat load, so your lucky if your radiators will keep that at a 10*dT so your probably going to all that effort and expense for 5 degrees cooler temperature than what you would have if you just watercooled alone without any TEC's
Last edited by Liam_G; 05-19-2012 at 10:50 PM.
is this yours? whose setup are you calculating?
Last edited by Liam_G; 05-19-2012 at 11:05 PM.
Posting Permissions
Reply With Quote
Bookmarks