The goal of this project is to take power from the 12v rail of a common power supply and regulate the voltage to the peltier through the use of a thermistor and a potentiometer to adjust voltage to the peltier to maintain a set temperature. I have a small collection of peltiers. 40w 80w 220w 340w all 12v. although for this project I am only interested in the 40w peltier... maybe the 80w. Although, if there is a set of know components that can handle say 60w of power at a certain voltage, im sure a peltier configuration and be easily acquired to match voltage and wattage of the power regulator. There are many great applications for this, ie. cooling for soda, hard drive cooling, chipset cooling. some of these may even be used without the risk of condensation. I am not an electronics guru, but I believe this project would be a simple one to construct particularly if the wattage of the load is kept down to lower levels. So I am asking the electronics guru community for help on building this and help prepare a mod guide for others to follow. another great use of this would be to regulate voltage for fans (but fans do not require near the power a peltier will) Goals: 1. accept 12v dc from pc power supply 2. automatically adjust voltage based on temperature 3. adjust set temperature through a potentiometer 4. be able to be used for fan and peltier powering. 5. relatively easy to acquire components 6. Novice friendly with step by step instructions. I hope some one with the know how pics up on this and is willing to help me, and help the moding community by helping write a guide. I have basic soldering skills and know a little about electronics and what they do, but no idea on WHAT to get and how to make it work.
Some of you guru's browsing the electronics forum may not hit the mod guides section, so if your interested in this topic please click over to this topic http://forums.bit-tech.net/showthread.php?t=98336 Thanks EVRE
Sounds interesting. To help you, I will move this thread to Electronics where those people hang out who have the skillz you seek. When you write the guide, please post it in this forum.
The best method of power control to a Peltier is PWM (Pulse Width Modulation). Wasted energy is kept to a minimum, so you don't need big heatsinks, and cheap MOSFET switching transistors will easily handle the currents. Plenty of simple circuits around for fans up to an amp or two. For higher current you really need to split the load supply from the electronics supply to avoid false switching, but the same power source can be used. Also for peltier applications there may be a "best" frequency range to design to. Maxim go way overboard with this design, and no doubt you could use a PIC.
Thanks for the quick reply Looks like Bit-Tech's own Dave Williams has designs for a PWM Fan controller that can handle 60W! http://bit-tech.net/article/51 I am going to purchase the parts necessary to build Dave's PWM FC and try to use it with my 40w peltier. Im not good at reading a wiring diagram any suggestions on how to proceed? This is part of why I want to make this guide, I want to build this but dont know how to start. any suggestions on what to change to up its current handling ability? looks like this is off to a good start, basic design is already taken care of! :edit: I noticed the use of "strip board" is it a good idea to use it with these power demands? any other suggestions/ideas? THANKS!
I'll subscribe to this thread because I am actually looking for the same thing since last week (when I got my pelts). would that work? : http://www.oatleyelectronics.com/kits/k140a.html or this?: http://www.madshrimps.be/gotofaqlink.php?linkid=1950 Any clue how to make it from scratch? Thanks a lot!
I like that unit on oatlyelectronics. the only real problem is that the thermister is on the cpb, not on a wire that you could easily position. but maybe thats not a problem, we may be able to use a different thermister of the same value on a cable to position it.
However, I would prefer to build a unit from scratch, I believe that is the spirit of things. we could "barrow" their design and components used to make our own. that unit said it could handle 10a = 120w!
After a bit of digging, recommended frequecies for PWM are >15kHz, sometimes even >50kHz, smoothed back to a DC voltage with an LC filter. High frequency means low-value (and small size) inductors and capacitors can be used. From EDN The Oakley kit uses an LM324 oscillator/comparator, which is OK for fans with PWM frequency 30-120Hz but far too slow for high-frequency use, that and macroman's MIC502 circuit should work OK but may be down on TEC efficiency. I've built a MIC502 controller (which operates around only 30Hz IIRC ) with a 14A IRF530 MOSFET switcher (less wasted volts than with the 5A TIP122 darlington ), no change needed in the basic circuit apart from adjusting R1/R2 to control below room temperature. The thermistor can have its leads extended no problem, it's shown on the pcb just for the photo.
I've actually been working on this project off and on (mostly off) for some time, so I'll follow this with interest. We used it for the purpose of avioding the dreaded condensation around a chip when the pelt got too cold. The end result of our first unit was displaying to a small LCD screen and controlling by buttons what temp you wanted the pelt to hold to. Though useful, it still allowed for condensation when the ambient air would shift higher than the pelt was set to. Before we got into testing it on a chip (just the bottom of a glass of hot water), we got a new idea. The second model allowed you to set it to X degrees above ambient. This was by far more useful, because if ambient changed, then the peltier wouldn't accidentally drip below and cause condensation. The problem was, though the chip stayed cool, it seemed to crash a lot more. Third-party testing yielded the same results..chip stayed ~5C above ambient (which was my target setting). Our conclusion was that the PWM creates EMFs like crazy due to the pulse, helped by the high frequency that works for pelts. That 'on/off' of such a large amount of current (we used a 240w pelt) seemed to upset the chip under it. Chip ran fine again when we switched back to HS/F. We had no scientific way of validating our conclusion. I'm distilling this to the extent of my knowledge, as I was not the Elec Engineer on the team...so if we skipped a step that could have fixed it, I wouldn't know. Maybe our ideas and conclusions will help you yield better results? I'd love to see this work right from a 'homebrew', as it is very expensive to buy a pelt regulator.
Couldn't you use inductors and capacitors to filter out the noise? BTW, for the circuit pictured above, only an inductor, capacitor, and diode (for shunting the inductive kickback) are necesary.
Do I gather you weren't using an LC filter system to effectively convert the HF PWM supply back to DC? That would reduce, maybe even eliminate, spiking close to the chip electronics. To control to a difference from ambient, a pair of linear sensors (eg, LM35) could be used. And for the controller, maybe a TL494 circuit modified from this one by Oatley Electronics I've been meaning to try since 11 August 2002, 21:36:46.
Actually, you gather correctly. Beyond my knowledge, but my EE friend's reply was "DAMN!". Then there was some mumbling about the entire circuit being re-drawn and yelling at me for taking his patent-able idea and making it public. Now whoever builds one of these things properly, send me one.
It probably wasnt patentable as it has been done before. Infact there was a comersially available device last year. It is pretty easy to sort out you just need an LC filter on the output which isnt re-drawing the entire circuit.
Edit- decided to put up full beer cooler artical instead. full circuit info will now be able to be found there. http://forums.bit-tech.net/showthread.php?t=98828
