2JZ VVTi

[gross polluter]

Ok, here's a trigger log of the cam progressively advancing from the full retard position at cranking. When viewing the trigger log in MLV, on page two you see where the code goes from sync to no sync roughly 40 degrees into the sweep, if the VVT angle gauge was reading correctly.

Don't have an MSQ to accompany. Trigger mode was set to "2JZ VVTI"

[porelmundo]

Are you crancking with the sparkplugs? I have some jumps on the cam when crancking different than your log. I will post my log this afternoon I was out programing other cars.

[gross polluter]

No, that's cranking with the spark plugs out to get the cranking speed as quick as possible.

[jsmcortina]

Ok, here's a trigger log of the cam progressively advancing from the full retard position at cranking. When viewing the trigger log in MLV, on page two you see where the code goes from sync to no sync roughly 40 degrees into the sweep, if the VVT angle gauge was reading correctly.

Good information.

The way my code is written, it looks for the cam between certain crank tooth numbers. My "window" of teeth wasn't large enough. I'll email you a test code.

James

[porelmundo]

well its looks like i have some noice at the cam sensor, provably is my vr conditioner, but i was looking if i keep cranking it's sync but way off from crank timing i have 155 not 160 on rising edge. here are the logs with some noice i will keep working on that this afternoon.

[porelmundo]

It sync on the third turn of the engine, if at last cranck timing is fixed I can turn on the engine.
Thank's for your support.

[gross polluter]

James, I'll test the new code when I return to the shop on Tuesday.

[jsmcortina]

Ok,
I need to have a look at the timing offset issue too. The code I sent you has no changes for that.

James

[porelmundo]

Any time to check the offset timing yet?

[gross polluter]

Alright, tested the new code today and can't get constant sync at all. During cranking, regardless of cam angle, RPM will momentarily read ~200rpm then drop to zero and no sync. Seemed to do this once every engine cycle. I've attatched two trigger logs and the MSQ that contains the correct trigger settings for the crank sensor. The first in the order of the trigger logs is just cranking at full cam retard (no power to oil control valve.) The second log in the order was done using the same method as I had used from my original post, cranking with power to oil control valve, sweeping through the cam timing range from retarded to advanced.

Hope the info I provided helps. I can do other tests if need be.

[jsmcortina]

Updated code on its way shortly.

I've moved the "tooth windows" to look for the cam pulses based on those latest composite logs. I have also built in that 160 deg tooth#1 angle and the timing looks the same on the scope.

James

[gross polluter]

Fantastic. Will try the new code this evening.

[porelmundo]

I will do the same tomorrow morning.
Thank you

[gross polluter]

Alright, Here's where I'm at.

With the latest code the car stars right up and runs well. Timing is rock solid at the commanded value. I ran the cam through the sweep with the engine idling using the VVT test output. Engine seems to maintain sync the entire time, albeit running really poor to the point where it will stall by the time it gets to well into cam advance. But, that's a lot of overlap so that's expected.

One of the issues I ran into, however, is that the cam angle gauge can't seem to keep a relatively constant reading. With the cam at full retard (zero duty) the reading bounces between 136 degrees and 158 degrees absolute. I managed to make the cam advance slowly by reducing the PWM frequency and setting the duty to 80%. During advancing, about 40 degrees into the sweep, the the cam angle would bounce between 194 and 13. I could not keep the engine running very well beyond 40 degrees, but I think the data I collected should conicide with what I am describing.

I've included the MSQ, datalog, and trigger log of this event. If you need any other data, let me know and I'll do my best to provide what I can.

[jsmcortina]

Looks like your crank VR sensor is wired backwards - take a tooth log and you'll see that your missing tooth is unclear.

Looking at the composite logs by eye, it is also clear to see that one of the cam teeth is uneven ?!? I'm sure I thought I was seeing this in earlier logs too. Perhaps changing the capture polarity on crank and cam might help this too.

I'm going to send you some debug code as well.

James

[gross polluter]

I've thought about that too. Back when I provided the original composite log I had tried changing the polarity of both the cam and crank sensors to see if I could yeild a different result. I could never get sync from any other polarities than what I am using now. Now, if my cam sensor was wired in reverse polarity, wouldn't I see discrepancies in ignition timing through the RPM range? As it is now, if I lock the timing at 10 degrees, it will hold timing rock solid all the way through the rev range.

I'll try to verify polarity through the wiring manual this evening, in addition to trying the test code you sent me.

[gross polluter]

Alright, it looks like both sensors are wired backwards. I don't have an english wiring diagram for this engine so a lot of stuff I had to figure out on my own. I'm taking a look over the english GS300 manual right now and it shows that the export market cars use the same wire colors as my engine has on the crank and cam sensors. If the colors are the same, both of my sensors are reversed. I suppose I'll correct that issue and report back.

[gross polluter]

Quick update: polarity on both cam and crank sensors were indeed reversed. Swapped them to the correct polarity and everything seems to be working fine now. Cam angles are stable, vvt follows target table, etc.

But, since my polarity was reversed I'm now seeing the same 155 degree trigger angle that porelmundo is seeing. For the time being I've added 5 degrees to my entire ignition table to compensate.

[jsmcortina]

For a quick 5 deg offset you can use the "trigger angle/offset" setting. I'll build that revised angle into a new code.

James

[porelmundo]

I say it before 155 not 160