2JZ 3/0 cam wheel problem

[WOOL]

Andrey: Told me that you need to move the Crankshaft offset into the right phase. Who is the right phase? I thought the main reference sensor was the crankshaft sensor. But here's the log.

lastest logs is https://rusefi.com/online/view.php?log=1279

[AndreyB]

2JZ 3/0 cam wheel is a unique one and it looks like while we have all the needed code, we never got it to properly run and never documented the magic way it's configured while we are figuring it out.

At the moment your log shows that engine phase was not detected - note the magic field names I am looking at.

image.png (233.21 KiB) Viewed 17346 times

First improvement: I've already improved the names for two of these fields https://github.com/rusefi/rusefi/commit/768eefbb36d4ffc195f9f24b82a2ca8fd2c74534

Now the real problem: you need to configure some magic settings. In the binary which you have at the moment it's hidden at Lua Calibrations -> Lua Script Settings, please set it like here (we care for two bottom values)

image.png (37.03 KiB) Viewed 17346 times

The method above is insane so I am now adding explicit special settings especially for 2JZ https://github.com/rusefi/rusefi/commit/8391bc4dee794071544babb796e6c0e783f6017b

https://github.com/rusefi/rusefi/issues/5557

[WOOL]

So I set the settings that you suggested above, it seems I have camshaft synchronization? Do these magic numbers correct the positions of the cams? I want to know more about it))

Last logs is https://rusefi.com/online/view.php?log=1281

[AndreyB]

Yes, these magic numbers are the range of randomly designated sync tooth on the cam.

Now please update to fresh snapshot and let's use proper configuration specifically for 2JZ

image.png (78.07 KiB) Viewed 17300 times

[WOOL]

I downloaded the last Snapshot and set the above values.and now I don't see the position of vvt1 vvt2
Here is the log
https://rusefi.com/online/view.php?log=1282

[WOOL]

So, we officially have a success on synchronization. I'll leave a link to the log here

https://rusefi.com/online/view.php?log=1285
Thank you Andrey , this is magic !!!

[WOOL]

There is a problem, VVT syncnorization disappears after 1800 RPM.

It seems to come from here https://rusefi.com/forum/viewtopic.php?f=4&t=2521&p=47394&hilit=lazy#p47394

javaw_xP6dXFZzcI.png (59.01 KiB) Viewed 16961 times

Do not change the camshaft sensor to a simple hall sensor without this joke?! I don't fully understand why this is happening.

Log is https://rusefi.com/online/view.php?log=1290

And last tune https://rusefi.com/online/view.php?msq=1544

[AndreyB]

We have "Maximum cam/VVT sync RPM(rpm)" setting since technically you do not need sync all the time

What is the problem you are facing? Looks like you have no VVT so nothing negative is happening? https://rusefi.com/online/view.php?msq=1544&dialog=vvtPidDialog

[AndreyB]

Looks like the log was taken at 100 reads per second not max reads per second?

[WOOL]

We have "Maximum cam/VVT sync RPM(rpm)" setting since technically you do not need sync all the time

What is the problem you are facing? Looks like you have no VVT so nothing negative is happening? https://rusefi.com/online/view.php?msq=1544&dialog=vvtPidDialog

Hi Andrey, in this setting I have the value set to 0 - disabled, always synchronize.
Now not a single vvt solenoid is connected until I understand that everything is ready for vvt with feedback. The problem now is that when my revolutions rise above 1800, the values ​​of the vvt angles freeze and do not change until the engine returns to speeds below 1800 rpm (I have a chain timing belt and when the revolutions rise, the vvt angles move slightly away from zero, this is a landmark for me) In the future I will hard-limit the vvt to 6200 rpm, but for now I'm losing the feedback from the vvt.

[WOOL]

Looks like the log was taken at 100 reads per second not max reads per second?

Yes, usually 100 readings per second is enough for me. Need to make a log with the maximum value of readings per second?

[AndreyB]

At 1800 RPM 100 per second is no longer useful.

[WOOL]

Okay , today I’ll make a log with the maximum value of readings per second and post it

[WOOL]

Is that it? will look the same as the previous log ?!

log is https://rusefi.com/online/view.php?log=1292

[WOOL]

So, since I have a camshaft sensor from mitsubishi, I began to look for similar ones in size and connectors. And I found.

WhatsApp Image 2023-10-02 at 15.58.13.jpeg (62.43 KiB) Viewed 16651 times

On the right is my native sensor, to the left of 4g18, 4g63 and it fits the pinout. Now about the my problem. This sensor does not have Invert signal after 1800 and one of the camshafts now always has a constant VVT angle. So my problem is precisely that the sensor inverts the signal (Lazy hall)

javaw_XFAZ3TMc3S.png (62.75 KiB) Viewed 16651 times

Is it possible to do something about it?

Log file in high speed here https://rusefi.com/online/view.php?log=1294

[WOOL]

Today I bought a second such sensor and decided to try to control the VVT closed loop and I have a crash again. If I try to turn the camshaft at least 3 degrees, then the engine starts to shake and the VVT readings of this camshaft freeze (This can be seen in the graph, a straight line)

javaw_ZXgxOQpGcl.png (70.42 KiB) Viewed 16593 times

So long I will not return the camshaft position to 0 degrees on the table. Is this again a loss of synchronization and a closed loop going out of the parameters? I also limited the maximum PWM to 50 to eliminate swinging using the PID regulator.


As always,Fresh log https://rusefi.com/online/view.php?log=1296

and tune https://rusefi.com/online/view.php?msq=1548

Only VVT1 output is assigned

[NormanAlphaspeed]

3/0?

2JZ cam's are either single tooth on non-VVT engines, or 4-1 on VVT engines. What is this 3/0?

[WOOL]

Good time everyone! Norman, unfortunately I don’t know about the origin of this camshaft, but I have exactly this on my Subaru ez30.

Now about the current state. At the moment, thanks to Dron_gus and Pavel, I managed to find out the duty cycle range and I worked a little with the PID regulator and finally the vvt with feedback works and holds the speed, but this continues until +10 degrees, after which the connection about the camshaft position is lost and the vvt angle readings freeze.

I made a log and uploaded a new tune with the current problem.
Log https://rusefi.com/online/view.php?log=1298
Tune https://rusefi.com/online/view.php?msq=1554


Please help with this, this is the last serious problem on the way to my happiness and the happiness of the engine

[AndreyB]

Looks like you are lucky with the relative angle of your two cam shafts for the code to synchronize both cam positions at all using same settings for both shafts. I might would need to change things later for two 2JZ cam shafts to work more reliably.

Anyway. You have the magic position = 95 and precision = 25 settings meaning it's looking for a tooth in the from 95-25=70 to 95+25=120 range. Well, it also looks for the offset by 360 range of from 360+70=430 to 120+360=480.

Note the top yellow values on my screenshot - the red bank 1 intake looses sync when the yellow line gets below 430.

Try changing precision = 25 to precision = 40 for now?

[WOOL]

Yay, those magic numbers helped me. Now I can set an angle greater than 10 and it is held. Now I have connected the wiring of the second solenoid VVT Right cylinder head and I do not see any changes, not seeing the PWM on this channel. If you swap the Exits on the solenoids, the PWM appears on the channel on the right head and disappears on the left.
I also don't see any readings in vvtTargets 2

NEEEE.png (45.22 KiB) Viewed 16402 times

I updated the firmware from the last snapshot

Log here https://www.rusefi.com/online/view.php?log=1299

Tune here https://www.rusefi.com/online/view.php?msq=1556

[AndreyB]

rusEFI usability issue: bank 2 intake is target 3

rusEFI functional issue: only bank 1 cam events were logged just fixed https://github.com/rusefi/rusefi/commit/0921a303d09696d8cbae37e4abab5fd783beae20

To be continued.

[AndreyB]

I am sorry I am using you for manual testing while proper way would have been to add some auto-testing of multi-channel VVT

Anyway I've just improved VVT output PID logging https://github.com/rusefi/rusefi/commit/0df3dbb06b75036f310e4172ad190dbb79befef1 it would not solve your issue but at least give us more data. Please make a new log once firmware snapshot builds.

[AndreyB]

Found the defect https://github.com/rusefi/rusefi/issues/5599 already fixed it!

[WOOL]

Hello everyone, today I tested what Andrew fixed, for which I am very grateful to him. When setting up the Precision VVt after setting the values to 40, everything works up to 28 degrees, anything above again a loss of synchronization. Now I've set it to 60 and have a successful sync up to 38 degrees, I haven't raised it higher in the vvt goal table yet

lkklkl.png (85.16 KiB) Viewed 16160 times

Success Log https://www.rusefi.com/online/view.php?log=1306

The Tune of Success https://www.rusefi.com/online/view.php?msq=1562

Andrew has magic numbers again. Thank you very much Andrew!