-
Notifications
You must be signed in to change notification settings - Fork 0
Home
Welcome to the ice-nucleation-cold-stage wiki!
Circuit works fine if just switched on or off; however, if we use a PID controller driven by the raspberry Pi it seems that power is lost when the temperature goes to around -20C or so.
My current hypothesis (no pun intended!) is that the 3A LM2596 switching regulators shutdown and this may be due to the DC power supply becoming negatively biased. I think the PID controller tends to switch the buck mosFET off and on rather quickly. When the mosFET PWM goes from 0, to 1 at the gate it may cause the inductor to generate back EMF of the same magnitude as the DC supply, which may mean that the 3A bucks do not have high enough Vin.
Test this by measuring the DC voltage as this happens.
I may be able to fix by placing a diode across the terminals of the supply, such that current flows from ground to Vcc if ground if higher voltage than Vcc (e.g. https://en.cosel.co.jp/technical/qanda/a0018.html). Another option might be to the play with the PID parameters in the Arduino firmware, so it doesn't flip so quickly.
It might also be worth setting the minimum value to be greater than zero - e.g. 10 - so that it is not switching off completely.
Turns out not of the above really worked, but I realised that the PID controller was set to update every 200ms. However, because the timer is divided by 62, it was really 200/62 ms (which caused problems with current - the DC supply was switching off). Setting the PID update time to 200*62 solved the problem.