VVT with Flying Miata 36-2 wheel

[Peter Florance]

Has anyone used this setup?

Tried to dyno a car today but ran into a problem I could not solve.
Miata VVT motor (2004 ish) with Flying Miata 36-2 wheel. No sync issue.

when I made a pull, the VVT Target 1 would stay at whatever the highest value it encountered in the VVT1 Target table as it pulled from low to high.

It the log, the VVT 1 target just flat lines.

Just so they could drive it, I left the max target somewhat lower to keep the high rpm reasonable. They will have to come back for sure.

Tried with MS3 1.3.3 release and 1.3.4 release. Log and msq attached from 1.3.4.

Thanks

Thanks!

[Peter Florance]

I should note this is an MS3 on V3.0 board. Not MS3 Pro

[Peter Florance]

Hmm... not sure I like this composite log I got car owner to take.

Note CAM pattern as engine is revved up (page 7 or so).

miata_composite.png

But no loss of sync.

[DIY-Ben]

Couple of things I see.

1 -- Your VVT load axis is TPS. This seems like an odd choice to me. Why not use MAP for load?
2 -- Your VVT slew, the factor that resists target change, is set to 20. Default is 2. Have you tried reducing it?

Unfortunately I couldn't open your composite log. It seems we ran out of TS licenses. Hoping Phil gets that fixed for me soon.

[Peter Florance]

Couple of things I see.

1 -- Your VVT load axis is TPS. This seems like an odd choice to me. Why not use MAP for load?
2 -- Your VVT slew, the factor that resists target change, is set to 20. Default is 2. Have you tried reducing it?

Unfortunately I couldn't open your composite log. It seems we ran out of TS licenses. Hoping Phil gets that fixed for me soon.

1 I've always used TPS for load. As cam position can affect MAP, there is a threat of instability. Sorry of like boost control targets.
2 Discovered problem with Slew= 2. Tried 20 as a test.

[Peter Florance]

Had owner adjust R11 CW 2 turns and now it has double teeth throughout the log.
He'll do a street pull.
I'm surprised I get sync. I guess code doesn't check?

[jsmcortina]

Had owner adjust R11 CW 2 turns and now it has double teeth throughout the log.
He'll do a street pull.
I'm surprised I get sync. I guess code doesn't check?

Without correct cam detection you should get half-sync. The missing tooth crank wheel is enough to run the engine.

James

[Peter Florance]

Hmm... Shouldn't Status 1 go from 136 to 8?

[jsmcortina]

Once (if) the code has full sync by observing the 2, 1, 2, 1 pattern on the cam, it doesn't need to check again. Unless the crank wheel falls off (!), maintaining crank sync is enough to know true engine position.
EDIT: This is done to make the code more noise tolerant. Previously there were cases where a single noise spike on the cam would then cause the engine to lose sync. This was causing more harm than good.

James

[Peter Florance]

Adjusting R11 CW seemed to fix the problem. Really odd that it showed up more as VVT Target problem. I wish I had been smart enough to do a composite log on the dyno.

vvt_test_7_9.png

I think the issue is value of C13. When I build them, I change C13 to .001uf to keep phase shift low. I didn't build this one, but I'm surely going to update.

ms3x_cam_in_schem2.png

Thanks for the help and input.

[Peter Florance]

Update, changing C13 to .001uf was a huge improvement. VVT started working perfectly.

Still wish VVT had some error reporting ("I don't like this pattern"). I find it is not easy to take a composite log through a complete dyno run. At some point I always lose records. Maybe I need better hardware.

After fixing cam input, I did have some sync errors at probably at about 6800 rpm. Changed the crank sensor gap from about .9mm (one credit card) to about 1.8mm (two credit cards), and it worked perfectly after that.

I thought the Miata sensors were conditioned VR sensors; my results suggest that they are (undershoot at gap causing extra tooth).

[aidandj]

Update, changing C13 to .001uf was a huge improvement. VVT started working perfectly.

Still wish VVT had some error reporting ("I don't like this pattern"). I find it is not easy to take a composite log through a complete dyno run. At some point I always lose records. Maybe I need better hardware.

After fixing cam input, I did have some sync errors at probably at about 6800 rpm. Changed the crank sensor gap from about .9mm (one credit card) to about 1.8mm (two credit cards), and it worked perfectly after that.

I thought the Miata sensors were conditioned VR sensors; my results suggest that they are (undershoot at gap causing extra tooth).

Peter, I'm having what appears to be the same issue you were seeing.

What did you set R11 to after switching C13 to .001uf.

I have a bunch of logs of my issue here: http://www.msextra.com/forums/viewtopic ... 31&t=66104

[Peter Florance]

I don't think it will be critical with the new C13 value.
Maybe 2.5 volts to start with.

Sent from my XT1080 using Tapatalk

[aidandj]

Thanks. I'll test.

[aidandj]

I spaced when I first read this. But it's on the cam input right? So C-13 on the ms3x?

[aidandj]

I replaced C13 on MS3x with a .001uF cap and now VVT is working fine. Thanks Peter.

James: Is the not checking for cam sync only with the 36-2 wheel? Or the standard Miata 4 tooth wheel also.

[Peter Florance]

I replaced C13 on MS3x with a .001uF cap and now VVT is working fine. Thanks Peter.

James: Is the not checking for cam sync only with the 36-2 wheel? Or the standard Miata 4 tooth wheel also.

This is how I found the problem.
The 36-2 wheel has enough information to crank sync. The stock wheel does not
So I had sync but no VVT on the 36-2 wheel.
C13 is much too large in this circuit.


Sent from my XT1254 using Tapatalk