Hi, I'm trying to power my boost converter from a microcontroller and I've 
encountered stability problems. I tried to make pseudo-PI controller, a PID 
controller and few other tricks, but the unloaded output just doesn't want 
to be even close to stable. 
Of course, I'm using a buffer between my MCU and the power mosfet 
(MCP1407). I run it from 72MHz clock on 100kHz, so PWM has limited 
resolution (only 720 steps). Of course, having the output voltage match one 
value of PWM is pretty much impossible, so I accept oscillation of PWM 
value by 1 as stable output. However, when the load is very light and the 
output capacitor is smaller than 10uF, the output is hard to control - it 
tends to oscillate by more than 100 steps, generating over 50Vpp AC wave on 
the output! This is unacceptable.
The only semi-working method is to reduce my PID controller to only I part 
with very small gain, which results with settling times of about 0,5s - 
which for a nixie clock isn't really a big deal. This is a compromise that 
I was able to achieve. I fear that a converter with so slow reaction will 
be unable to properly power a multiplexed clock.
I'd like to hear from you - have you tried to power your clocks with MCU 
driven boost converters? Did you encounter similar problems? Was there any 
other solution than placing a big cap (increasing "inertia" of the circuit) 
or pre-loading the output? Or is my PWM resolution just too low and 
lowering frequency to standard 20-30kHz would do the job?

I really like the idea of using MCU to drive the boost converter, in theory 
it is perfectly capable of doing the job, I just can't seem to get it to 
work properly. 


 

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send an email to [email protected].
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/8fdd3917-c7ca-4bd5-9ceb-3d62663a50d8%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to