Subaru 36-2-2-2 VVT decoder wrong phase

Testing and development of Megasquirt 3

Moderators: jsmcortina, muythaibxr

Post Reply
jhoefer
MS/Extra Newbie
Posts: 5
Joined: Mon Jun 22, 2015 10:34 pm

Subaru 36-2-2-2 VVT decoder wrong phase

Post by jhoefer »

I've got a Subaru JDM EJ205 engine. This has a 36-2-2-2 crank wheel, a 4-2 (2 teeth spaced 90 degrees apart) main cam wheel, and two 4-tooth (with no missing teeth) cams for intake-only VVT. It uses VR sensors on all four locations. This is different from the US version of the Subaru intake-only VVT engine.

I have an MS3 v3.57 with MS3X running 1.5 alpha 4 firmware. When using the "Subaru 36-2-2-2 VVT" decoder with Full Sequential spark and fuel, the spark and injector phasing is off 360 crank degrees (180 cam degrees). For testing this, I have set the Cranking Degrees to 0 and the Injector Timing Table to all zeros in order to capture the baseline timing of events on the oscilloscope. Here is the chart of the engine timing events based on the raw oscilloscope signals with spark and injection signals for cylinder #1:

SiliconStew_subaru 36-2-2-2 VVT decoder wrong phase.jpg
This is not a wiring issue, all injectors and coils are matched to their correct cylinders in the firing order and verified with Output Test Mode.

The engine will run if changed from COP to Wasted COP to work around the incorrect sequential spark phasing.

Two things of note, I have only included the lefthand VVT cam signal, not both, and because I was turning the engine over slowly and did not have full oil pressure I cannot at this time guarantee that the VVT timing lines shown represent the fully retarded position. Also, this tune was only for testing this issue and is not from a running car so consider it dangerous to use on your own engine.

Suggestions, fixes, workarounds, or something I missed in the config? While not ideal, Wasted COP does work and I suppose I can try to mess with the Injector Timing Table to see if I can fudge things there, but I don't know if there's any other spots in the configuration I have to watch out for with the timing being out of phase like this.
Matt Cramer
Super MS/Extra'er
Posts: 17507
Joined: Thu Apr 16, 2009 8:08 pm

Re: Subaru 36-2-2-2 VVT decoder wrong phase

Post by Matt Cramer »

Try using the VVT signal on the other cam for your primary cam input.
Matt Cramer -1966 Dodge Dart slant six running on MS3X
jhoefer
MS/Extra Newbie
Posts: 5
Joined: Mon Jun 22, 2015 10:34 pm

Re: Subaru 36-2-2-2 VVT decoder wrong phase

Post by jhoefer »

Matt Cramer wrote:Try using the VVT signal on the other cam for your primary cam input.
I can't. Like I said, on this engine the VVT cam wheels are 4 teeth, 90 degrees apart, with no missing teeth. Cylinder #1 has to be determined off the 2-tooth "Main Cam" wheel.
Matt Cramer
Super MS/Extra'er
Posts: 17507
Joined: Thu Apr 16, 2009 8:08 pm

Re: Subaru 36-2-2-2 VVT decoder wrong phase

Post by Matt Cramer »

Sorry, I misread your post. Is the 2 tooth wheel the same as a non-VVT engine?
Matt Cramer -1966 Dodge Dart slant six running on MS3X
jhoefer
MS/Extra Newbie
Posts: 5
Joined: Mon Jun 22, 2015 10:34 pm

Re: Subaru 36-2-2-2 VVT decoder wrong phase

Post by jhoefer »

Matt Cramer wrote:Sorry, I misread your post. Is the 2 tooth wheel the same as a non-VVT engine?
Yes, as far as I know. Here's one reference to the crank and "Main Cam" signals (unknown subaru year and model):
SiliconStew_subaru unknown 36-2-2-2 crank and cam signal.jpg
Here's traces from a USDM 1999 Legacy (no VVT) that has those same crank and "Main Cam" signals:
SiliconStew_subaru 1999 Legacy 36-2-2-2 crank and cam signal.png
And the crank and "Main Cam" trace in my original post is for a JDM 2001-2005 STi (intake-only VVT), which matches these same crank+cam signals.


When the USDM engines finally got VVT for the STi model only, they switched to using two 4-1 intake cam wheels and eliminated the "Main Cam" sensor. Here's a reference for a USDM 2004 STi (intake-only VVT):
SiliconStew_subaru 2004 USDM WRX 36-2-2-2 crank and VVT.gif
However, the USDM engines without VVT (for the same 2004 year) still used the same crank and single "main cam" sensor as the 1999 Legacy and other examples I've shown.
Matt Cramer
Super MS/Extra'er
Posts: 17507
Joined: Thu Apr 16, 2009 8:08 pm

Re: Subaru 36-2-2-2 VVT decoder wrong phase

Post by Matt Cramer »

In that case, it sounds like you need to use regular 36-2-2-2 mode, not the VVT version. That should get you working.
Matt Cramer -1966 Dodge Dart slant six running on MS3X
tpsretard2
Master MS/Extra'er
Posts: 662
Joined: Thu Feb 14, 2008 4:59 am

Re: Subaru 36-2-2-2 VVT decoder wrong phase

Post by tpsretard2 »

i do not know where i posted, but i never got my car to work with the 36-2-2-2 and the 2 tooth cam wheel, but i never tried in wasted cop mode. :O
alpine227
MS/Extra Newbie
Posts: 5
Joined: Mon Oct 24, 2016 2:30 pm

Re: Subaru 36-2-2-2 VVT decoder wrong phase

Post by alpine227 »

Any luck? I have been wondering why I couldnt get sync. I am in the same boat with a JDM ej205 with intake avcs. Seems to work fine on wasted spark cop but loses sync when I try cop. I have two cam sensors at the rear of the heads and two at the front and the oil pressure solenoids in the middle. I was going to log the wheels today but ended up pulling the engine due to a faulty water pump. If there is something I can do to help let me know I would very much like to run the vvt and cop.

Respectfully,

John.

Sent from my XT1565 using Tapatalk
acab
Experienced MS/Extra'er
Posts: 243
Joined: Tue Aug 21, 2012 5:52 am

Re: Subaru 36-2-2-2 VVT decoder wrong phase

Post by acab »

any luck?
Post Reply