uS Timing accuracy problems

[grippo]

What rpm were you running when you did these tests ?

[grippo]

I re-ran your msq at 120 rpm but changed the trigger rise to calculated since the 120 rpm put it in cranking mode, which means it sparks as soon as it gets a vr1 edge. By using calculated it sparks according to the table I then set up a situation like you had. To do this I had to change trigger offset to 175 deg and used 5 deg spark adv in the table. That gave a picture that matched your scope shot. I then advance/ retarded from about 10 deg to 0 deg in 1 deg increments and everything stepped as it should. I used spark as the scope trigger since this is what should change relative to the wheel teeth when you change either trigger offset or the spark table advance. If you use your original 135 deg it puts the TDC pt between teeth 7 and 8 as opposed to between teeth 6 and 7 on your scope shot. Tooth 1 is the falling edge of first tooth after the gap. I checked the motion on that also with the same result.

[bgeisick]

All of my previous testing and data was run at 1450 RPM. This was the reading in TS and I also verified it by looking at the dt between the t1 signals of a full revolution.

I wanted to make sure I was understanding all of this correctly, so I marked up the picture I took of the scope at 135 offset and 5 deg spark timing, 1450 RPM. Could you verify that I have this right? I still don't understand how the timing signal can occur before tooth 7 trigger. What is still more confusing to me is why I see a huge jump in timing for only a 1 degree change in requested timing.

135offset_5spark_FallingTrigger_markup.JPG (63.25 KiB) Viewed 920 times

Last night I collected more data at different rpms and noticed the large jump in timing occurs at different requested timings. I ran this test in two ways and got the same results. The first time I set timing to a specific value and swept rpm until I saw the jump in the timing signal on the scope.

Sweeping speed:
RPM Timing where jump occurs
1250 5
1470 4
1780 3
2100 2
2400 1
2950 -1

The second way I ran the test was to fix the speed and sweep timing at each speed until I saw the timing signal jump.

Sweeping timing:
RPM Timing where jump occurs
1200 5.05
1500 3.95
2000 2.15
2500 0.75
2975 -1.05

I realize that these are not realistic timing values, but if I choose a different offset for my trigger I can easily get the jumps to occur in the range of useful timing values. This seems like maybe it's another clue as to why my setup is doing this. If it's helpful, I could run the same sweeps with different trigger offsets and add it to the plots below.

Timing Jump vs. RPM.JPG (27.55 KiB) Viewed 920 times

Is there anything I can log that would help solve this? Is it possible to add different internal code variables to the log files by changing the *.ini file?

[grippo]

Here are the pictures I got at 1450 rpm. This is with 135 deg offset The reference is the falling edge of the first tooth after the gap, so going back from there about 139 deg for the 1st photo (135 + 4 deg adv) this is 3.88 teeth and as you can see the dwell stops (= spark) right about there. The 2nd photo is same but 3 deg more advanced. The dwell/ spark stays where it was because that is my trigger, but the bottom trace moves to the right relative to the top trace indicating a bit more advance. The 3rd photo shows a large 31 deg advance, again where it should be.