Since you are doing a gated oscillator switching converter, you could consider using the on-board comparator of this particular MCU instead of the ADC. Might save you some code/CPU time.
Alex On Monday, November 25, 2013 7:55:12 PM UTC, petehand wrote: > > > > On Monday, November 25, 2013 7:13:52 AM UTC-8, Chris Stalin wrote: >> >> LOL, Love it. Great work! >> >> Are you PWMing a FET for power from the atmega158? >> >> Yes I am. Given the current discussion here about AVR and Arduino, the > technical details are probably of some interest. > > Timer 0 is set up as fast PWM on the undivided 8MHz clock, with PWM output > to the OC0B pin (PORTD.5 in this case). The register settings are TCCR0A = > 0x23, TCCR0B = 0x09. I have the output compare register A (OCR0A) set at > 250 – this is the top limit of the counter and sets the PWM rate about > 40kHz. You can make it lower for higher frequency operation. Output compare > register B (OCR0B) sets the pulse width and is initially set low, around > 40. The OC0B pin goes high when the counter resets to 0 from 250, and goes > low again when the count matches OCR0B. The FET gate is connected to this > OC0B pin and turns on when the output is high. This counter is completely > free running and never stops, but the OC0B pin only drives the FET when its > port pin is set as an output, so the converter can be suspended without > disturbing the PWM by changing the port to an input. The port is > initialized as an input to avoid accidents, and is only set as an output by > the regulating routine once this is on the job. > > Regulation is performed with ADC channel 0. This is set to interrupt on > completion. The register settings are ADMUX = 0xE0, ADCSRA = 0xDE, ADCSRB = > 0. The high voltage is divided down with a couple of resistors – 200k to > the high voltage, 1k to ground, with the ADC input connected to the > junction. The ADC then reads 1/200 of the high voltage, comparable with the > ADC internal reference of 1.25V. The interrupt routine compares the read > value to a constant setpoint value, determined by experiment (200 for 175V > in my circuit). If the voltage is lower it increments the value of OCR0B, > and if higher, decrements it. If OCR0B is ever decremented to zero it sets > the OC0B port pin as an input to suspend the PWM, otherwise it sets it as > output. Finally it triggers a new conversion. Everything then happens > automatically, with no program intervention. Timer 0 OCR0B ramps up until > the set point is reached and thereafter hunts around that point as the load > changes, giving essentially constant voltage output. > > For switching, I use an IRL640 MOSFET. This is a logic level FET with a > 200V rating, about the highest voltage rating you can get in logic level > devices. Its gate is connected directly to the AVR OC0B pin, with a 2.2k > resistor to ground to make sure it stays turned off when the pin is set as > input. The inductor is a Coilcraft MSD-1278-224 – this is a 220uH SEPIC > choke with dual coils. In my application the coils are connected in series. > The FET drain is connected to one end of the pair, the other end to the > rectifier, and 12VDC comes in at the junction. With this arrangement the > FET only sees half the peak voltage, so it has a good overload margin. > > > > -- 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 neonixie-l+unsubscr...@googlegroups.com. To post to this group, send an email to neonixie-l@googlegroups.com. To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/9b90a928-68b6-4b0d-b4e7-541059792e89%40googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
