Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Peltier Project: 50% PWM -> alternate circuit switching for increased performance?

  1. Apr 30, 2012 #1
    Peltier Project: 50% PWM --> alternate circuit switching for increased performance?

    Hi all,

    I have a project where I am making a drink cooler device using four peltier devices which can draw up to 30amps or 24 volts each.

    we are considering battery power (pretty ridiculous but anyway..) and were considering Lithium polymer batteries.
    http://www.maxpoly.co.uk/shop/index.php?main_page=product_info&cPath=3&products_id=85 [Broken]
    my thinking is that 2 of these could provide 15volts at up to 110 amps (we should be fine with 20-25amps per module) constant current with a capacity of 216000 joules which should last 4-5 cycles.

    we are planning to use PWM with a control circuit to monitor temperatures and output the correct voltage which would provide the current giving us the best COP to extend the battery life.
    as the peltier devices are very high power and dissipate a lot of heat we would probably only run them at under 0.5 PWM so the high voltage and current is mainly so that we can select the optimum PWM for whatever temperature the device is at.

    I was wondering if there was a way to split the circuit in two so that 2 peltiers were separate from the other two and they could be alternately PWM controlled so that while one was off, the other could be on. high frequency 1000Hz + is required as the peltiers do not like low frequency cycling.
    this would let use more of the battery performance allowing us to have up to 30V through the devices at the same current by putting the batteries in series as only half the current would be needed.

    any thoughts on all of this please? and the main question:what would I use to switch alternately between 2 circuits at high frequency? im thinking this would be added in series after PWM is applied which would let us control the PWM separately to stop the device overheating.

    side question: is there a simple way of powering this device off mains power as high current power supplies are expensive!! at the moment we are having to restrict ourselves to low currents

    Thanks for your time!
    Last edited by a moderator: May 5, 2017
  2. jcsd
  3. May 1, 2012 #2
    Re: Peltier Project: 50% PWM --> alternate circuit switching for increased performanc

    i have suggestions which may not help you but give you some ideas.
    Actuvally, i am working on a project to control a 1000Hz welding operation where we got bad results due to continous weld current as output. Harware consistes of of transformer and later to that section with 4 IGBT combinations for switching operation. for example, Two IGBT's are ON at a constant time and a cycle(positive and negative) of current will be drawn for each combination continously. Because of this continous current production, we have had some problems with realtime weld that had bad spots on welded material after operation. There, we used burst current to stabilize wled electrodes heating from continous current. for example, we turn on IGBT's for 10 cycles of current and then blank period of one or two cycles to make electrodes cool. This cool period is also depend on previous ten cycle weld current. If the current drwan is more then cool period increases and vice versa and so on.

    Does it make any sense? apolozies for non related answer.
  4. May 1, 2012 #3
    Re: Peltier Project: 50% PWM --> alternate circuit switching for increased performanc

    im note realy sure how that would apply..

    I have been looking around though and while looking up what an IGBT was (not sure i fully understand..) I realized I wanted a high frequency/high power SPDT relay switch or something that would do similar.

    Alternativly as I am using a 18f4331 PIC chip, I might be able to use a 2 complimentary PWM outputs set to 50% which I am guessing would work alternately. Im thinking I could then combine this with 2 and gates and buy a new MOSFET so that each circuit has one.

    or even easier I could use a single PWM at 50% and use and and gate on one circuit and an or gate on the other.

    now is that a better solution to the original problem? and would it be better 2 get 2 high power logic gates or 2 normal logic gates and a new MOSFET? or would there be a way to get the PIC chip to output my signals without buying logic gates? will the logic gates operate fast enough?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook