Controlling Peltier with Variable resistor?
JetFuel
I have recently become more and more interested in peltier cooling
solution construction, I know the fundamentals ie... hotplates/coldplates,
insulation against condensation etc..... the question I have is would it be
feasible to employ the use of a variable resistor to dial up the amount of
cooling the TEC generates (Heat too), If by controlling the voltage to the
TEC, could one not tame the condensation problem (crudely I might add) by
cooling the CPU... not freezing it, to say a realistic level..... 15 degrees
C for example.
Cheers
Al
Warren Hawk
good Q - i was thinking along those lines for two reasons
one - some what I've read the peltier runs on voltage from 3V to 12 V
- and so the performance changes too.
two - the power drawn is about the same quantity as the heat moved
from one side to the other.
from this it would appear that your idea is possible, I'm looking for
more information at the moment, but I'm looking forward to other posts
about this.
Warren
D Zero VIBE
the same thing. I haven't done it yet, but I'll probably try it within
the next week. I think what I need is a potentiometer, which I think is
a variable resistor. The only thing I'm not sure of is what
specification to get for the potentiometer. Is there a way to calculate
what range of Ohms I need? Has anybody else tried something like this?
0
JetFuel wrote:
>
> Hey Guys-
>
> I have recently become more and more interested in peltier cooling
> solution construction, I know the fundamentals ie... hotplates/coldplates,
> insulation against condensation etc..... the question I have is would it be
> feasible to employ the use of a variable resistor to dial up the amount of
> cooling the TEC generates (Heat too), If by controlling the voltage to the
> TEC, could one not tame the condensation problem (crudely I might add) by
> cooling the CPU... not freezing it, to say a realistic level..... 15 degrees
> C for example.
>
> Cheers
> Al
po...@carolina.rr.calm
>Hey Guys-
>
> I have recently become more and more interested in peltier cooling
>solution construction, I know the fundamentals ie... hotplates/coldplates,
>insulation against condensation etc..... the question I have is would it be
>feasible to employ the use of a variable resistor to dial up the amount of
>cooling the TEC generates (Heat too), If by controlling the voltage to the
>TEC, could one not tame the condensation problem (crudely I might add) by
>cooling the CPU... not freezing it, to say a realistic level..... 15 degrees
>C for example.
WOW...I was just thinking about something similar about an hour ago.
I was thinking about it in reference to cascaded peltiers (with 2 good
power supplies, you could run 4-12volt peltiers and a pair at 5volts.
Any ideas (I'm trying to think of it on my own, but a little help
never hurts) on how to regulate BOTH so the consumption wouldn't be
too bad...but the final temperature would stay low?
Check out my site for information on overclocking, installation of fans,
heatsinks/thermal compound, etc...and of course, pics of my water cooler.
http://home.carolina.rr.com/poit
tiggre
of heat building up...
ill post if my 450@600 still runs good at 5v... also the pot
we're talking bout if you need it... should take 15v and 3amp
in my case... gonna do some calculas on it, promise... :)
// tig
<po...@carolina.rr.calm> wrote in message
news:37ce373b...@24.93.94.77...
Andrew Chin
you'd better make sure that the resistor can dissipate the same amount of
power as the Peltier.
To use a variable resistor better, you'd need it to control the base current
of a transistor. The transistor itself handles the large current.
The best method would be to use a PWM control. As long as the frequency is
greater than 2kHz (mentioned on spec sheets), you can vary the duty cycle
and thus control the amount of voltage reaching the Peltier.
JetFuel <jet...@sprint.ca> wrote in message
news:zCoz3.2011$qY3....@newscontent-01.sprint.ca...
> Hey Guys-
>
> I have recently become more and more interested in peltier cooling
> solution construction, I know the fundamentals ie... hotplates/coldplates,
> insulation against condensation etc..... the question I have is would it
be
> feasible to employ the use of a variable resistor to dial up the amount of
> cooling the TEC generates (Heat too), If by controlling the voltage to
the
> TEC, could one not tame the condensation problem (crudely I might add) by
> cooling the CPU... not freezing it, to say a realistic level..... 15
degrees
> C for example.
>
> Cheers
> Al
>
>
>
po...@carolina.rr.calm
wrote:
>iam just otw to go for 5v instead of 12... since i get loads
>of heat building up...
>
>ill post if my 450@600 still runs good at 5v... also the pot
>we're talking bout if you need it... should take 15v and 3amp
>in my case... gonna do some calculas on it, promise... :)
I meant using peltiers rated at 7-8volts for the 5volt peltiers and
ones rated at 15v for the 12volt peltiers.
Two power supplies can't provide enough power for 5-6 peltiers at
12volts, but the 5volt power output of many ATX power supplies is
rated at (hold on to your butt) 30AMPS! So you could easily power a
couple of the 7-8v peltiers with two power supplies.
To get rid of all that heat I was talking about using a water cooler
with a 15 X 15 X 7 cm. heat exchanger and a 15cm fan :)
It's actually getting into the range of vapor phase cooling...it's
just a little easier to work with in some ways.
Danny Ayers
You'd need either a big variable resistor (look at the power ratings in the
catalogues), which would incidently give off a fair bit of heat itself, or,
prefererably something like those low power drill speed controller kits. This
would most likely reduce the power by switching fast, and thus not have the same
heat problem.
You could be really smart & link the power supply to a temperature sensor and
use a feedback loop to keep the cold plate at a constant -1 or whatever. Bet you
could do that with an old boiler thermostat & a bit of duct tape...
Cheers,
Danny.
Danny Ayers
another thought on this one - there are bog-standard voltage controller chips,
only need like a variable resistor and a heatsink, and can handle quite high
powers. They'll give nice continuous voltages and are good and stable. In the uk
this kind of thing can be obtained from Maplin etc, in the States - Radio Shack?
po...@carolina.rr.calm wrote:
> On Thu, 2 Sep 1999 02:16:57 -0400, "JetFuel" <jet...@sprint.ca> wrote:
>
> >Hey Guys-
> >
> > I have recently become more and more interested in peltier cooling
> >solution construction, I know the fundamentals ie... hotplates/coldplates,
> >insulation against condensation etc..... the question I have is would it be
> >feasible to employ the use of a variable resistor to dial up the amount of
> >cooling the TEC generates (Heat too), If by controlling the voltage to the
> >TEC, could one not tame the condensation problem (crudely I might add) by
> >cooling the CPU... not freezing it, to say a realistic level..... 15 degrees
> >C for example.
>
> WOW...I was just thinking about something similar about an hour ago.
>
> I was thinking about it in reference to cascaded peltiers (with 2 good
> power supplies, you could run 4-12volt peltiers and a pair at 5volts.
> Any ideas (I'm trying to think of it on my own, but a little help
> never hurts) on how to regulate BOTH so the consumption wouldn't be
> too bad...but the final temperature would stay low?
>
Jason Hubbard
in a very irregualr way (i.e. very temperature dependant) - remember TEC's
commonly dissipate in excess of 50 watts. A better answer here is to use a
voltage regulator, like say an LM337K - This simple device will supply upto
5 amps - parallel them via 0.22 ohm resistors (to share the current) if you
need more amps. The power supply should really be built outside your PC case
and will of course need a heatsink for the LM337K. The size of the heatsink
will depend on the difference in input/output voltage - i.e. 30 volts in, 15
volts out gives 5 volt difference, multiply by current of say 4 amps gives
60 watts of heat dissipation - get the idea?
Good luck
Jason Hubbard
JetFuel <jet...@sprint.ca> wrote in message
news:zCoz3.2011$qY3....@newscontent-01.sprint.ca...
> Hey Guys-
>
> I have recently become more and more interested in peltier cooling
> solution construction, I know the fundamentals ie... hotplates/coldplates,
> insulation against condensation etc..... the question I have is would it
be
> feasible to employ the use of a variable resistor to dial up the amount of
> cooling the TEC generates (Heat too), If by controlling the voltage to
the
> TEC, could one not tame the condensation problem (crudely I might add) by
> cooling the CPU... not freezing it, to say a realistic level..... 15
degrees
> C for example.
>
> Cheers
> Al
>
>
>
Warren Hawk
problem - I'll look into it :)
Warren
>I think the best way would be to use a thermostat to turn it on and
>off. The basic circuit is very simple, you probably built it in your
>science class in school, you make a voltage divider, with your
>thermistor on one side, and a variable resistor on the other. These go
>to the base of a transistor, and the variable resistor is adjusted so
>the circuit turns on and off at the desired temperature.
>
>You'd have to get the thermistor in a nice position, and you could use
>your computer's own temperature monitoring to calibrate the circuit,
>ie have it switching on and off just where you want it.
>
>I'd switch the peltier through a relay. That way, you can use a small
>transistor, and you don't have to worry about it being not fully
>saturated and dropping a load of voltage and heating up and exploding.
>The relay would give you a simple on-off action, and your computer
>would make an interesting clicking sound.
>
>I could draw this circuit if anyone wants to try it, it will be cheap,
>you'll have to test it yourself but you've nothing to lose.
>
>Oh yep, one issue is, how conductive of heat is a peltier when it's
>switched off? If the answer is "not very" then forget all this.
>
>------------------------------------------------------------------------
>
>Why pamper life's complexity,
>when the leather runs smooth on the passenger seee-eee-aaat?
>
> http://www.greenaum.demon.co.uk/
Andrew Chin
Peltiers don't like to be switched on and off slowly.
http://www.tellurex.com/resource/txfaq.htm#quest1
Has some info on how peltiers should be controlled.
Incidentally, I am constructing a PWM motor speed controller. This circuit
could easily be adapted for a Peltier.
I'll have a crack at it and e-mail my results to someone who has a half
decent homepage.
<sam@greenaumARSE!ARSE!ARSE!.demon.co.uk> wrote in message
news:37d15595...@news.demon.co.uk...