Originally Posted by
spfautsch
Are these thresholds specified by ADC count or the conversion value (which is the "engineering" unit I mentioned previously)? I'd love to know their exact values and where the compare branches are in the program.
I dug into the LT-1 disassebly I've been annotating and have learned a lot, but it's unclear to me where these switch points reside.
I've been able to draw several connections between the 0xEED1 and 0xEFD1 tables and some other data points I've collected, so I feel like I'm very close to having this part nailed down. The biggest difference I've come across is that the AVR has a 10 bit ADC processor (0-1023) whereas the PCM seems to have an 8 bit ADC processor (0-255). So I'll need to divide all the ADC values by 4. Otherwise it looks like it will be a walk in the park.
Thanks! That's very informative that they published the actual resistance values and circuit design. Cool find!
I've been busting my butt to get all the information I have incorporated into the next firmware as well as working towards a build that handles temperature compensation. Best case, this is going to eat up at least 10-12 days to get to a "rough draft". But I'll do my best to make time to get some spark durations with different gaps and electrode materials.
Ultimately I'd like to build out a zener pile and test spark energy across the board. But at this point in the project I'd rather have a working (if incomplete) firmware for the one and only test subject, because it's a lot more fun to own if I can drive it.
Side note - since discovering that the coils I have require < 4ms of dwell, I'd been nudging the dwell target down little by little. But after a rudimentary analysis of the test data I decided to try boosting cranking dwell by quite a bit, and cold starts have thus far responded with great enthusiasm. I feel like it's about to get real (or blow up in a magnificent ball of fire).