Hello,
I have to design a circuit for "proportional" temperature regulation: a PTC resistor is used to heat a fluid and to measure its temperature (existing design that can't be modified). The PTC heats the fluid up to a target temperature then a 5s cycle is started in order to regulate the temperature:
-- 1. What does the PTC's R/T curve look like? 2. Is the 12 to 19V spread used to control the amount of heat generated (12V = low heat, 19V = high heat) or is it the drop across the varying R with 7A through it, or is it something else? 3. Why can't you use the output of the 24VDC supply directly by switching in appropriate resistances?
The temperature is controlled by switching between a few voltage steps from
12V to 19V. For example, if the temperature is greater than the target after the measurement (PTC was initially powered by 19V), the voltage is switched down to the lower voltage step (for example 17V) in order to reduce power in the PTC. And so on.If you mean a simple voltage divider, it would be less efficient than a buck converter and would require large power resistors. The temperature regulation is a part of a whole system with other features.
--------------------------------------- Posted through
TC
how about two supplies?
12-19V buck (or just pwm) for the heating,disable when stabilizing/ measuring
You can't PWM?
Isn't that getting kind of personal? ;-)
-- Politicians should only get paid if the budget is balanced, and there is enough left over to pay them.
This doesn't seem to be a prohibitively difficult loop to stabilize, particularly considering that you don't need a huge degree of precision except during the measurement.
Have you identified where and why it wants to oscillate? Is it because it has to be in discontinuous mode some of the time? Does it need to go into "skip cycle" mode, and if so does it meet its ripple requirements for the measurement?
Using a half-bridge driver (i.e. synchronous rectification) instead of a pass transistor and catch diode would make the conduction continuous, for a vast simplification of the loop dynamics. You may not even lose that much efficiency overall if the better efficiency at high currents pays back the worse efficiency at low.
Even if you can't do that, designing a circuit that is stable for the end points of the load envelope (which is probably low voltage/high PTC resistance vs. high voltage/low PTC resistance) would probably be sufficient to make it stable throughout.
-- Tim Wescott Wescott Design Services http://www.wescottdesign.com Do you need to implement control loops in software? "Applied Control Theory for Embedded Systems" was written for you. See details at http://www.wescottdesign.com/actfes/actfes.html
TC
So, meter the applied voltage and the resulting current, both. The ratio (of the RMS averages, if it's done with AC) gives you the resistance, thus the temperature, within the uncertainty resulting from self-heating. The change in resistance, phase-locked to the low-current part of the cycle, gives you some kind of measurement of the self-heating, too.
I though it was getting rather practical.
Sigh. No sense of humor? :)
-- Politicians should only get paid if the budget is balanced, and there is enough left over to pay them.
@ Lasse: Yes, I was thinking this way. Separating the power supply from the measurement supply appears to be a good solution.
@ Tim: For the measurement, it would be always in discontinuous mode: L package size should remain reasonable. At higher temperature than the target, it would be in discontinuous mode too. Theoretically, the circuit is stable in continuous mode.
The circuit being designed for continuous mode, I still have to predict its behavior and stability in discontinuous mode (maybe output ringing?). What is the best way to ensure good behavior even in discontinuous mode? I'm not (yet) an expert of power design... :P
I will take a look at and maybe consider a synchronous buck controller.
@ Whit3rd: Of course, the self-heating of the PTC must be taken into consideration (if relevant).
Thank you for your answers, Johann
--------------------------------------- Posted through
Just chronically short of sleep....
Be careful. That can destroy your health. :(
-- Politicians should only get paid if the budget is balanced, and there is enough left over to pay them.
To late - I'm disabled.
Me too. Hard work tried to kill me, and almost succeeded. :(
-- Politicians should only get paid if the budget is balanced, and there is enough left over to pay them.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.