Help with cam configuration Protege 1.8L FP engine

[CRF_MS]

Developing an MS3/MS3X v3.57 for an 1999 Protege 1.8L FP, which is similar to the 1999 Miata 1.8L BP engine, but with an entirely different trigger system. It has an 36-1 VR trigger wheel for the crank sensor, and a three tooth VR trigger wheel for the cam sensor.

Like the Miata, the engine has two coil packs, so the ignition runs in wasted-spark. We hooked up each injector to it own driver, and we initial got the engine running in batch mode, and have now moved to semi-sequential, but would like to be able to support sequential.

From what I can tell, the cam setup on the Protege FP (and FS) engines is a one of kind, with three big teeth on the cam wheel. There is one tooth after cylinder #1 TDC, and two teeth around cylinder #4 TDC (at and after). The service manual states that the missing tooth on the crank wheel is 90° before cylinder #1 TDC, so tooth #8 on the crank wheel should be cylinder #1 TDC (Tooth #1 Angle is 80). A picture of the cam wheels can be found on this web page (http://forums.haltech.com/viewtopic.php?f=23&t=3013 [1], first picture).

In not clear to me from the Mazda service manuals where the first cam tooth is on the cam wheel (cylinder #1), but it seems from the MS trigger logs, that it's about 20° after cylinder #1 TDC. The service manual does state that cam tooth #2 is 340° after cam tooth #1, and cam tooth #3 is 40° after tooth #2, and another 340° back to cam tooth #1. That should make cam tooth #2 around cylinder #4 TDC.

From the composite trigger logs, I see the following,

cam tooth #1, 10-13 crank teeth after missing tooth
cam tooth #2, 8-11
cam tooth #3, 13-15

If I use the poll level to determine the cam position, I have two choices, I can look for when cam tooth #1 is active when cam teeth #2 and #3 aren't, or I can poll when cam tooth #1 is not active, but cam teeth #2&3 are.

It looks to me that because of the overlaps, there is only one tooth that cam tooth #1 is active (12), and cam teeth #2&3 are not, but there are two teeth (8&9) when cam tooth #2 is active, but cam tooth #1 is not.

I have read the MS3 setup manual, the MS3 and MS3X hardware manuals, but I still don't understand how "poll level" works, which tooth is tooth #1 on the crank wheel, and which value "Level For Phase 1" to use (0v requires HIGH and 5v requires LOW, but that seem backward to me). I thought that I read somewhere, that it doesn't poll on the tooth given, but on the next tooth, so how many unique crank teeth are need to have a reliable cam sensor?

Is there a way by looking at the datalogs or trigger logs to see where the system is checking the polling level, and which cylinder it thinks is cylinder #1? I tried several settings, but each time I switch to sequential, the engine starts and idles, but it not happy, like it out of phrase, so I have switch it back to semi-sequential.

Off_Off2.csv

ON_ON2.csv

Protege 1999 1.8L FS.msq

[CRF_MS]

OK, no one has posted a reply to ANY of my questions, so after some more thought, I think that I can answer some of them by using a dual trace oscilloscope. I put one lead on the Cam sensor, and the other one on injector #1.

I set "Second Trigger Active On" to Poll level, "Level For Phase 1" to Low and "Crank At Tooth#" to 8. Now when I switch from semi-sequential to sequential, I see injector #1 go from two injections to one injection and by changing "Level For Phase 1", I see the injection location switch from before cylinder #1 TDC (single CAM tooth) to before cylinder #4 TDC (double CAM tooth).

I found that I could change "Crank At Tooth#" from 8 to 9 and there seems to be no difference. Changing it to either 7 or 10 would cause an "Cam fault".

So this does seem to work, and after a quick drive around, no Cam faults.


Now to my other questions, I found in the MS3XV357_Hardware-1.3 manual the following:

the 'tooth' should be normally start at least 20 crank degrees before tooth#1 and continue for another 20 crank
degrees afterwards. (The level is actually polled at tooth#2.)

so since my Cam teeth are about 20 degrees long, is this long enough to have a reliable cam sensor?

Because I'm using an old non storage oscilloscope, and it will NOT sync on the double tooth, it jumps between syncing with the double tooth back to the single tooth, so it difficult to be 100% sure where the injector fires, so, my question

Is there a way by looking at the datalogs or trigger logs to see where the system is checking the polling level, and which cylinder it thinks is cylinder #1? I tried several settings, but each time I switch to sequential, the engine starts and idles, but it not happy, like it out of phrase, so I have switch it back to semi-sequential.

still needs to be answered.


One last issue, when I switch between semi-sequential (autotune) and sequential using the test mode, I see the AFR going lean by at least 3 to 4 points, and it doesn't matter which setting of the "Level For Phase 1" is, it does basically the same thing. While I expected to see a change, I expected it to go a little richer, not a LOT leaner. Have I done something else wrong?

[Matt Cramer]

It looks like you'll need to go after tooth 2 as it's the only one with no overlap.