Erratic crank signal

[401amx1969]

Hi
I need help to resolving this problem.
Here is my system:
1- V8 engine
2- dual wheel ignition, 4 teeth on crank and 1 tooth on cam.
3- both sensors are VR, pot R52 and R56 are fully counterclockwise adjusted.
4- software noise filter is ON: time= 1.5ms and percentage is 50%.
5- I have installed a 10K resistor in line on the crank sensor.
6- Firmware version 1-3-2

2015-05-19_17.32.01.csv

You can see on the data log page 9/20, the signal from the crank sensor start to come crazy at approx. 4500 rpm and over.
On the last test, the composite logger show 9 pulses instead of 8 per cycle.
Before I have installed the 10K resistor, this problem was started at 3000 rpm.

thanks in advance.

[Matt Cramer]

Try adjusting the trim pots and/or adding a larger resistor.

[401amx1969]

Try adjusting the trim pots and/or adding a larger resistor.

Thanks Matt for your time.
But maybe I was not clear enough in my request. I agree to admit that the 2 pots and resistance in series will have a direct effect. But why start with one or the other? I research a technical explanation of the direct effect of each of these actions. From the data log, why change the zero crossing or introduce hysteresis, or increase for higher inline resistance .....
Where do I have to start?

[Matt Cramer]

This link may help. http://www.diyautotune.com/tech_article ... _index.htm

[401amx1969]

2015-05-30_14.52.47.csv

2015-05-30_12.36.34.csv

OK, I tried to adjust the level with R52 first. The result was not bad but requested 6 turns!.Now with a larger resistor (20K) R52 & R56 are fully counterclockwise and composite logger looks much better.
BUT, a other problem. At 5000 rpm exactly, the computer always lost the sync......? I have tried 4-5 times and always 5000. Above that, test and tune with Auto Tune works well.
I have checked the rev limiter and pass all the parameters and unfortunately find nothing to explain why.
Here is the composite logger and tooth logger.

Any idea?
Thanks

[Matt Cramer]

Try turning off noise filtering.

[401amx1969]

Try turning off noise filtering.

Mat, I did 30-50% on both sensor without success.
The pecision of the shutdown make me feel more on programing mistake.
Data logger always show tach shutdown @ 4990 ~ rpm simultaneous with the lost sync.

[prof315]

Try turning off noise filtering.

Mat, I did 30-50% on both sensor without success.
The pecision of the shutdown make me feel more on programing mistake.
Data logger always show tach shutdown @ 4990 ~ rpm simultaneous with the lost sync.

No Matt is suggesting that you try now WITHOUT any noise filtering. Also you are using the Tach Rejection based on your description of the settings. Try leaving that off and using just the Primary Tach Noise filter curve. I have had pretty good luck with a 200 to 400 uSec curve starting at 500 rpm (200 uSec) and ending at ~ 7500 (400 uSec) on low tooth count VR signals

[401amx1969]

My apologies on my misunderstood (thank prof 315)
Your support is appreciated.

This week-end, I have tuned with an oscilloscope the crank and cam sensor. On crank sensor, I added a resistor of 10K inline + a shunt of 10K. On cam sensor, I added a shunt of 1K. Without these resistors, composite logger was showing intermittent pulses from both sensors at any rpm. From this done, signal was clean to 5000 rpm before the shutdown.....

Following your ideas Guys, I tried today as you said:
1- I turn off Tach rejection and did a test as is. Same result as before.
2- Removed resistors and without any noise filtering (all off). Problem was started at low rpm.
3- Re-installed the resistors and programming the noise filter as prof 315 suggested. Finally I was able to break the fence and grow to 5500.
4- I tried to tune by increasing and-or decreasing the time (from 400 uSec to 200-600) at 5000 rpm target, but still with a poor result confirmed by the composite logger.
5- With noise filtering back to off, resistors still in place, I tried pot R52 turned 2 first, and 1 more turns later clockwise, the engine was growing to 6000 and composite was very nice at 3 turns total. Please, see data logger page 42. You maybe able to see something than I don't.

Now, I have to tune the VE table before to try 7500 rpm.

I don't understand the effect of the noise filtering. I'm agree to turn on, but how I can see the effect?

2015-06-01_17.38.38.csv