Peltier Driver using Mosfet problem

Thread Starter

Luiz Henrique Marques

Joined Sep 26, 2017
6
hi people i wonder if anyone could help me out

I'm working on a cooler box. In order to heat ,i'm using peltier elements attached to an aluminum heatsink. its block is composed of 2 peltier in parallel.

In order to drive these cells i'm using a voltage follower circuit leading its input voltage( Microcontroller) to the cells . Although my circuit works ,it seems to be halted. whenever my input voltage surpasses 5V , its voltage drop though does not follow ,stalling at that point. Thus it only drains 5Amps(2.5amps each cell)

i wonder why this is happening ... could anyone solve?

best regards



upload_2018-3-15_14-49-35.png



Here's the operation point where i'm at...
loll.PNG
 
Last edited:

MrChips

Joined Oct 2, 2009
23,505
IRFZ44 is an N-channel MOSFET.

You want the MOSFET on the low side, i.e. interchange the TEC with the MOSFET.

You don't need the feedback to the opamp. In fact you don't need the opamp. You can drive the MOSFET directly from the MCU if you use a logic level compatible MOSFET, such as IRL530.
 

Thread Starter

Luiz Henrique Marques

Joined Sep 26, 2017
6
IRFZ44 is an N-channel MOSFET.

You want the MOSFET on the low side, i.e. interchange the TEC with the MOSFET.

You don't need the feedback to the opamp. In fact you don't need the opamp. You can drive the MOSFET directly from the MCU if you use a logic level compatible MOSFET, such as IRL530.
Sure, however that's the available Mosfet and i'd like to control the TEC power. If i interchange then wouln't it be a switch on/off structure?

i've searched and found a similar configuration

https://www.allaboutcircuits.com/projects/embedded-pid-temperature-control-part-1-the-circuit/
 

wayneh

Joined Sep 9, 2010
17,152
i've searched and found a similar configuration
That will work, as you're seeing, as long as there is not a lot of voltage drop across the load. At higher current, you're dropping enough voltage that your MOSFET's gate-source voltage isn't high enough to keep it fully on anymore. The MOSFET is probably getting warm - even hot - as you put 5A through it with its Rds-on higher than it could be.

Also, is your 12V power rail holding at 12V with a 5A load? It might be sagging as well, contributing to the problem.

Certainly part of the problem is that your op-amp cannot set an output voltage anywhere near the upper power rail. Check the data sheet - it may be topping out at 8-9V. Subtract the voltage drop across the load, and you're left with little Vgs.
 

MrChips

Joined Oct 2, 2009
23,505

Thread Starter

Luiz Henrique Marques

Joined Sep 26, 2017
6
That will work, as you're seeing, as long as there is not a lot of voltage drop across the load. At higher current, you're dropping enough voltage that your MOSFET's gate-source voltage isn't high enough to keep it fully on anymore. The MOSFET is probably getting warm - even hot - as you put 5A through it with its Rds-on higher than it could be.

Also, is your 12V power rail holding at 12V with a 5A load? It might be sagging as well, contributing to the problem.

Certainly part of the problem is that your op-amp cannot set an output voltage anywhere near the upper power rail. Check the data sheet - it may be topping out at 8-9V. Subtract the voltage drop across the load, and you're left with little Vgs.

i got it...my power supply is stable when i apply 5V or more .

you are right! Opamp output is about 9,5V . when i apply 5V drop to TEC element my VGS ends up being 4,5V which leads me direct to 5amps okay.PNG

so... how could i overcome this problem?
 

MrChips

Joined Oct 2, 2009
23,505
The article from AAC that you referenced uses analog control of the TEC current. When you do this the transistor will get very hot.

Using pulse-width modulation (PWM) means that the transistor is either fully ON or fully OFF. There will be minimal heating with PWM.
 
Top