Decoding Nissan style Mitsi Electric 360 slot cam sensors

[SleepyKeys]

Not yet, but with FreeEMS now running an engine the priority just moved up a few notches. The decoding is simple enough(although there are still some things missing from the current 360-CAS code). Sounds like your idea should work just fine. Here is another guy who used a PIC for decoding the CAS. http://www.bailey-eng.com/LTCC.html


-sean

[HOODEY]

How much degrees BTDC does the start of the trigger window represent for six cylinder applications?

[SleepyKeys]

Not sure about the 6cyl but for the 8 it's 0deg.

[KW1252]

I know you like to do as much as possible with code, but this might prove some further insight on the issue:

http://www.allaboutcircuits.com/vol_4/chpt_11/3.html

It's a simple rotary encoder... using two opto gap sensors will let you know also direction, which might be a considerable issue with 1 degree resolution.

[SleepyKeys]

Find me a circuit that has say 10-bit logic then we may be in business as we could just poll an entire port register based on an interrupt generated by x bit.
But then wer'e almost back to using the accumulator set to interrupt on x count.

Maybe I missed something ???

[beav]

Stumpled across this thread on google a while ago while looking for a solution for a RB25DET wheel and thought I would let you guys know I got a solution working using a dedicated pic16f628 @ 20mhz to decode the wheel.

Without getting too in depth the outer holes trigger the interupt which increments a counter if the inner hole is low. On the first interupt where the inner goes high it switches on the timing trigger(the leading edges are an equal 60deg on this wheel) to the ecu and then decodes the cylinder and sets a respective output pin. On the next interupt it resets the trigger output and does nothing with further interupts until the inner goes low and it starts counting all over again. I started with an old pic16f84 @ 5mhz I already had just to see if I could make it work at low rpms to test the concept not expecting it to be capable of a working product, I actually got it working with code stable to over 6k rpm so I ported it too this 20mhz for plenty of headroom. At something like 29k it cannot process the cylinder in the 1 degree window so the triggers falling edge would be a degree late on the next interupt, nothing catastrophic happens till closer to 100k rpm.

I apologize for not really being up to speed on this forum and your guys' efforts so I hope I am not offending anyone by posting a link to the setup on the other forum, just looking to help others with some of my efforts. If I did something wrong let me know.
http://www.anonym.to/?http://www.msextr ... 04&t=35517

[Fred]

The only thing you did wrong was failing to proxy the link :

http://www.anonym.to/?http://www.msextr ... 04&t=35517

Links to ms sites are blocked by referrer, believe it or not. Sad but true.

Just add http://www.anonym.to/? in front of the link to make it work...

Well done on the pic decoder. It sounds like you do it exactly as I intended Sean to do it. Hopefully using the accumulators we can keep the CPU load to a minimum.

Are you planning to share your work under some FOSS license? I'd love to read the code if you are willing to share. Even if it's in private and not FOSS, provided that doesn't compromise my ability to work on this decoder for FreeEMS, that is.

Thanks for posting! :-)

Fred.

[SleepyKeys]

Awesome work !

[HOODEY]

I started work on this as well using a pic18 and got it to work
using a pic16f877 at 20mhz. I used a distributor for a four cylinder.

I am using the outer circle to drive a counter and the inner for cylinder information for port b interrupts thus counting the width ofthe windows.
I can use CCP peripheral to set events off but I need to know which edge reperesents TDC.

When I rotate the wheel I can see the effect on some LEDs I have on Portb.

My real issue is understanding the nissan 6 cylinder format.

So far I understand that the rising edge of the inner circles are 60 degrees apart. Does this translate to 120 at the cranks.

Are the rising edges on the rb26det TDC or is it the trailing edge?

What are the slot counts for the six inner windows and what is the firing order.

I really need this information for my Knock monitor which does windowing.

[HOODEY]

Based on the first page it seems that the start of the window is 120 dgrees BTDC and on the nissan rb distributor the smallest window represents cylinder 6.

Nissan is actually counting between the inner windows so by the start of a window their now which cylinder and how many degrees before top dead centre.

In my application I need to knwow where 10btdc is and 40 degrees ATDC.
I can accomplish this by using compare port and the timer connected to the outer ring for the time base.