Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › Help! Design a Peltier/TEC driver
New Posts  All Forums:Forum Nav:

Help! Design a Peltier/TEC driver

post #1 of 6
Thread Starter 
In my project, I need a TEC for cooling, the paramater of TEC module I used is as below: Vmax=17V, Imax=6A, Qmax=38W, and TEC module can get the best performance when Vtec=13V(while Itec=3.5A), I've proved that in lab.
The current problem is that my DC source is 24V(200W), and I want to adjust the temp by PWM, below is my current schematic,one is used N-Channel, the other is P-Channel,
:
For getting Vtec=13V, I add a cement resistor in series with TEC module, I don't know if is there any potential problem, so I need ur help, or maybe have a better solution in my project condition.
Thx a lot.
post #2 of 6
Use a microcontroller or signal generator to spit out a sine wave. Feed it into a full wave rectifier. Feed signal into a high current op-amp which will drive your current...it'll need a big ass heat sink on it. You can use a transformer of your choosing to step up your sine wave signal voltage prior to feeding it into the rectifier. Cool thing about this design is you can vary the op amp gain and be able to generate whatever voltage you want (provided your op amp voltage rails are above the output voltage)

Been there, done that.

With your design your pull up resistors on the mosfet are too large in resistance. At 2 Khz your mosfets won't be pulled up/down fast enough and you'll get weird ghosting issues (been there done that). I'd guess off top of head you'd want 1k resistors or less. You also probably don't need such a large capacitor to hold up the output voltage. Run it through an online RC calculator though to verify.

Personally when I needed to regulate the temperature of an object with a TEC, I just used a pmos with a regular old relay. Didn't need any sort of fast switching control of the TEC itself since thermal mass took care of it for me.
Edited by dr/owned - 3/26/13 at 8:10pm
post #3 of 6
Since your power supply is 24V fixed. You need to lower your voltage, cement resistor is not a good solution because it wastes power. Search for "DC-DC Buck step-down converter" in google you will get the schematic you need. I would search Ebay for DC-DC step down converter. They are quite common. Otherwise buying a variable power supply is not that expensive, considering your current requirement is not high (high current ones are very expensive).

BTW your circuit seems OK, in principle it should work. The first mosfet is not necessary if you PWM signal can source enough current. The resistors on the second mosfet (R2, R4 in first figure) seem too big at 100k, several K ohm will be good.
Edited by foxrena - 3/26/13 at 8:38pm
Obsidian
(22 items)
 
CPUMotherboardGraphicsRAM
i7-3930K ASUS P9X79 Deluxe XFX HD 7970 Black Edition Double Dissipation Patriot Viper Xtreme Series 4x 4GB DDR3 2000 9-... 
Hard DriveHard DriveHard DriveCooling
OCZ RevoDrive3 X2 OCZ Vertex 4 WD 640GB  Swiftech 360 Radiator 
CoolingCoolingCoolingCooling
MCP 35X pump Swiftech Apogee XT waterblock Foxrena's Water Rack KOC TEC and Fan controller 
CoolingOSMonitorKeyboard
KOC 62mm TEC block Windows 7 Ultimate Dell UltraSharp IPS 27'' 2560x1440 Razor Black Widow Ultimate 
PowerCaseMouseAudio
Corsair AX1200i Corsair 800D Logitech G9X Logitech Z4 
  hide details  
Reply
Obsidian
(22 items)
 
CPUMotherboardGraphicsRAM
i7-3930K ASUS P9X79 Deluxe XFX HD 7970 Black Edition Double Dissipation Patriot Viper Xtreme Series 4x 4GB DDR3 2000 9-... 
Hard DriveHard DriveHard DriveCooling
OCZ RevoDrive3 X2 OCZ Vertex 4 WD 640GB  Swiftech 360 Radiator 
CoolingCoolingCoolingCooling
MCP 35X pump Swiftech Apogee XT waterblock Foxrena's Water Rack KOC TEC and Fan controller 
CoolingOSMonitorKeyboard
KOC 62mm TEC block Windows 7 Ultimate Dell UltraSharp IPS 27'' 2560x1440 Razor Black Widow Ultimate 
PowerCaseMouseAudio
Corsair AX1200i Corsair 800D Logitech G9X Logitech Z4 
  hide details  
Reply
post #4 of 6
Quote:
Originally Posted by dr/owned View Post

Use a microcontroller or signal generator to spit out a sine wave. Feed it into a full wave rectifier. Feed signal into a high current op-amp which will drive your current...it'll need a big ass heat sink on it. You can use a transformer of your choosing to step up your sine wave signal voltage prior to feeding it into the rectifier. Cool thing about this design is you can vary the op amp gain and be able to generate whatever voltage you want (provided your op amp voltage rails are above the output voltage)

Been there, done that.

With your design your pull up resistors on the mosfet are too large in resistance. At 2 Khz your mosfets won't be pulled up/down fast enough and you'll get weird ghosting issues (been there done that). I'd guess off top of head you'd want 1k resistors or less. You also probably don't need such a large capacitor to hold up the output voltage. Run it through an online RC calculator though to verify.

Personally when I needed to regulate the temperature of an object with a TEC, I just used a pmos with a regular old relay. Didn't need any sort of fast switching control of the TEC itself since thermal mass took care of it for me.

Audiophile grade TEC Control!

Wow, that's... complicated. Also inefficient. Why use a sine wave and a power op amp when you could do the same thing with a couple of transistors and a square wave? Cheaper, easier and more efficient. But then you've basically made a switchmode power supply and you might as well do that. I mean, you could even feed a square into the opamp and it'd still be more efficient...

Your idea has no advantages over a buck converter.
Tractorbox II
(6 items)
 
al dente
(11 items)
 
 
CPUMotherboardGraphicsRAM
Intel 4770k Gigabyte Z87-DS3H Nvidia Geforce 970 G.Skill DDR3-10700 
Hard DriveHard Drive
Samsung 840 Samsung 840 
CPUMotherboardRAMHard Drive
AMD Athlon 2 X2 240 ASRock 760GM-GS3 Kingston HyperX Blu Red Series 250Gig Seagate 
CoolingOSMonitor
Stock FX-6100 cooler Windows XP Wells Gardner 25K7194 
  hide details  
Reply
Tractorbox II
(6 items)
 
al dente
(11 items)
 
 
CPUMotherboardGraphicsRAM
Intel 4770k Gigabyte Z87-DS3H Nvidia Geforce 970 G.Skill DDR3-10700 
Hard DriveHard Drive
Samsung 840 Samsung 840 
CPUMotherboardRAMHard Drive
AMD Athlon 2 X2 240 ASRock 760GM-GS3 Kingston HyperX Blu Red Series 250Gig Seagate 
CoolingOSMonitor
Stock FX-6100 cooler Windows XP Wells Gardner 25K7194 
  hide details  
Reply
post #5 of 6
Quote:
Originally Posted by kevmatic View Post

Quote:
Originally Posted by dr/owned View Post

Use a microcontroller or signal generator to spit out a sine wave. Feed it into a full wave rectifier. Feed signal into a high current op-amp which will drive your current...it'll need a big ass heat sink on it. You can use a transformer of your choosing to step up your sine wave signal voltage prior to feeding it into the rectifier. Cool thing about this design is you can vary the op amp gain and be able to generate whatever voltage you want (provided your op amp voltage rails are above the output voltage)

Been there, done that.

With your design your pull up resistors on the mosfet are too large in resistance. At 2 Khz your mosfets won't be pulled up/down fast enough and you'll get weird ghosting issues (been there done that). I'd guess off top of head you'd want 1k resistors or less. You also probably don't need such a large capacitor to hold up the output voltage. Run it through an online RC calculator though to verify.

Personally when I needed to regulate the temperature of an object with a TEC, I just used a pmos with a regular old relay. Didn't need any sort of fast switching control of the TEC itself since thermal mass took care of it for me.

Audiophile grade TEC Control!

Wow, that's... complicated. Also inefficient. Why use a sine wave and a power op amp when you could do the same thing with a couple of transistors and a square wave? Cheaper, easier and more efficient. But then you've basically made a switchmode power supply and you might as well do that. I mean, you could even feed a square into the opamp and it'd still be more efficient...

Your idea has no advantages over a buck converter.

I bolded the advantage part over the OP's design. I also said "use a microcontroller OR a function generator with a sine wave", obviously the microcontroller would be a square wave althought it's possible to approximate a sine wave with more difficulty.

Personally, I don't like dealing with low level components. Some people get off on using 555 timers and capacitor oscillators...I say use a microcontroller. It's a preference thing. In the industry you actually want to use as much off the shelf high level stuff as possible, because you'll leave the company and then someone else is stuck trying to understand a convoluted analog design.
Edited by dr/owned - 3/27/13 at 1:51pm
post #6 of 6
A buck Convertor can very voltage just as well as your design. In fact, it can do it even better than your design, which will have series heat dissipation issues at lower voltage levels. An op-amp in such a configuration won't be much more efficient than a linear regulator when varying below 100%.
Tractorbox II
(6 items)
 
al dente
(11 items)
 
 
CPUMotherboardGraphicsRAM
Intel 4770k Gigabyte Z87-DS3H Nvidia Geforce 970 G.Skill DDR3-10700 
Hard DriveHard Drive
Samsung 840 Samsung 840 
CPUMotherboardRAMHard Drive
AMD Athlon 2 X2 240 ASRock 760GM-GS3 Kingston HyperX Blu Red Series 250Gig Seagate 
CoolingOSMonitor
Stock FX-6100 cooler Windows XP Wells Gardner 25K7194 
  hide details  
Reply
Tractorbox II
(6 items)
 
al dente
(11 items)
 
 
CPUMotherboardGraphicsRAM
Intel 4770k Gigabyte Z87-DS3H Nvidia Geforce 970 G.Skill DDR3-10700 
Hard DriveHard Drive
Samsung 840 Samsung 840 
CPUMotherboardRAMHard Drive
AMD Athlon 2 X2 240 ASRock 760GM-GS3 Kingston HyperX Blu Red Series 250Gig Seagate 
CoolingOSMonitor
Stock FX-6100 cooler Windows XP Wells Gardner 25K7194 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Peltiers / TEC
Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › Help! Design a Peltier/TEC driver