Decoding Nissan style Mitsi Electric 360 slot cam sensors

[deviousKA]

Question, how is your "virtual CAS" being incremented?

I follow your case structure and count methods, but how are you going to break down virtual events in between?

Edit, nm. Id guess you could increment it at 1/5th the rate of the incoming 360 pulse.

[Fred]

The trick is, for consistent counting he need to reset it when het gets sync from pure counting. I like the idea and I hope it all works :-)

Fred.

[SleepyKeys]

Question, how is your "virtual CAS" being incremented?

I follow your case structure and count methods, but how are you going to break down virtual events in between?

Edit, nm. Id guess you could increment it at 1/5th the rate of the incoming 360 pulse.

Yes, exactly(set the pulse accumulator to over flow at 5), this code is still missing quite a bit, but I thought I would throw it up there for comments/suggestions as we go.

Thanks

[SleepyKeys]

The trick is, for consistent counting he need to reset it when het gets sync from pure counting. I like the idea and I hope it all works

Fred.

Thanks and yep I think it will, ECT FTW!

[deviousKA]

When first looking over the ECT I failed to notice the hardware counter/pulse accumulator, and was a little confused. Im certainly going to be playing around with it myself as I will be using the Nissan type mitsubishi optical CAS in many nissan/datsun applications. I dont plan to split it down any further than 24 events per 720deg engine rotation though (24 +1sync perhaps) for 4 and 6cyl.

I have done some reverse engineering of the nissan vh45 v8 ecu which uses a disc which I assume is similar to your Lt1, but unfortunately havent gotten into the CAS decoding on that. I do know that the CAS signals are fed into at least two of the MCU cores on the ecu (of which it has 4). I assume it uses a different approach to figuring out the engine angle (non timer based) where it reads engine angle directly from the CAS. Reading both rising/falling edges on the disc would provide 1 deg crank resolution per signal, so it can just use a compare match to perform fuel/ignition action.

[SleepyKeys]

The LT1 had special hc11 cores made for GM, I think it's called hc11-F6. The LT1 ECU reads rising and falling as you said to give it a 1deg update. But I think they went a bit further and added similar math used in decoding lower count wheels. Since it gets a real 1deg update it could "guess" with great accuracy much finer values(.1 deg or something). In the timing table there are some fraction values.

I figured divide by 5 because it would yield 72 teeth per cycle, scale the numbers up and you have 50/720 I think it will interface with the scheduler quite easily.

[deviousKA]

Cool, the vh45 uses a hitachi 6303 (motorola 6801 with a few extra functions and higher pin count) as the main controller and a few other in house (JECS) MCU and/or PLC for supplementary control. The ecu provides sequential injection and C.O.P. on that engine.

[SleepyKeys]

Hey guys I need your help finding info/examples on how to use the accumulator in 8-bit mode on PT0. EVERY piece of documentation I have come accross says how to the pulse accumulator in 16-bit mode on PT7. As Fred said 8-bit mode on PT0 would serve us better.

Thanks in advance,
Sean

[SleepyKeys]

I suppose a an ISR could be fired to check the 8-bit count register associated with a pin, then use the modulus operator to see if it needs to branch and run the full primaryISR loop, but that means an ISR is fired on ever 360x event. Or I totally missed something (I hope that's the case)

[Fred]

HUH!!!

When PACMX = 1, PBOVF bit can also be set if 8-bit pulse accumulator 1 (PAC1) reaches 0x00FF and an active
edge follows on PT1.

You phail at reading pdf files :-) Which is good because having jumperable inputs and outputs would have sucked...

Now get on with it :-)

Fred.