PWM Phase control

Also recommend searching forum for eFlexPWM library. One of the examples is 3 x PWM at 0, 120, 240 deg as you would have for AC motor control. The examples run right out of the box on T4.x
 
joepasquariello said:
You're welcome. I always recommend this when I have a chance. The author is an academic, and he did a really nice job.
Click to expand...
next is to create a PID and use the eFlexPWM as the throttle to set the temperature of five Peltier 1206s to almost freeze a can in my 500W soda cooler. (-10c in 4.5 minutes)
 
MarkT said:
How are you going to sense can temperature?
Click to expand...
I use a 0.3% 100ohm at 0C temperature sensorMAX31865 and digikey 1759-1020-nd
I 3D printed a rotation/compression ring to squeeze the sensors onto the side of the can for direct measurements
1745432941880.png


this is a 2.8mm sensor. I bring it down to -10c for five minutes and regulate it a +1.5c until the can is at 2c
1745432698599.png

1745433133681.png


this is the first driver and it was too shallow
1745433295626.png
1745432627321.png
1745432698599.png
1745432941880.png
1745433133681.png
1745433295626.png
 
So I think you have good thermal sensing, I wonder if a simple bang-bang approach will work as the contents of the can has much more thermal mass than all the metal work, being aqueous.

Once the peltiers are off the temperature drop will stop rapidly, you can probably just turn on full cooling till about 2 degrees, then turn off and it will probably slightly overshoot to 1 degree.

Convection inside the can should keep internal temperature gradients low.

But experiment and see, and beware integral-wind-up...
 
MarkT said:
So I think you have good thermal sensing, I wonder if a simple bang-bang approach will work as the contents of the can has much more thermal mass than all the metal work, being aqueous.

Once the peltiers are off the temperature drop will stop rapidly, you can probably just turn on full cooling till about 2 degrees, then turn off and it will probably slightly overshoot to 1 degree.

Convection inside the can should keep internal temperature gradients low.

But experiment and see, and beware integral-wind-up...
Click to expand...
the old way was to run the Peltier's at max cooling and then turn the power down so it was at 1c steady state were it would take time freeze the soda solid again 😂 soon I will be able to analyze the new cooling profile and have an analog PI(D?) feed back. I figured out the shape of the air pocked in the can and will put a senser under the can and over next to the air pocket and see when the water inverts in the can
SCTESC 1.jpg
1745453463236.jpeg
 
You must log in or register to reply here.
Back
Top