SM4 2003 Subaru WRX engine sync signal missing.

[13larry]

Hello, having problem getting cyl sync from the cam reluctor sensor working. Using the monitor-sensor screen, the cyl sync is max at 15888 msec. If I reverse the pins on the sensor, it is min at 000. I have checked the signal with a oscilloscope and the signal is there. I have measured the resistance at pins 41 and 26 of the ecu connector and have 2K ohms (same as the sensor). Neither pin is short to gnd, has 224K ohms. I have include my cal file. I am also using the stock individual coils. I have use 4 O/P 4Cycl-MC and 4 O/P SUB0104 neither seems to work. Do I have a defective input?

[Ian]

Are you sure that the cam and crank sensor inputs aren't swapped? Unplug the cam sensor, crank the engine, check if the error reported is Sync or Cyl sensor missing. If error is Cyl sensor then the inputs are swapped.

If the error is Sync sensor then inputs are correct. In that case what does the rpm show when cranking? It should be around 180 rpm.

[13larry]

Ian, Disconnected the Cam sensor, got err Cam I/P missing, cranking rpm varies 159-213 while cranking. Larry

[Ian]

Ok. Is there any possibility that the cam gear has jumped back one tooth? If this happens the Autronic will not start as the Sync pulses are in the wrong position (It will go advanced one tooth though).

Is it possible to get a scope trace showing the Cam pulse waveform relative to the crank pulse. Expand / contract the trace so that the crank missing teeth can be seen together with the 2 cam pulses.

[MRMRacing]

STI variable camshaft Subaru to SM4
I have this information if it would be to any help!

[13larry]

I have not verified the cam timing. I will try to trigger on the crank signal and look at the cam, but this could be difficult with these pickups.

We are not running the variable cam timing, our engine does not have that feature. Please explain in the table the the reference to the 10 k ohm resistor, should I parallel each sensor with 10 k ohm?

What I think is happening is the cam sensor is counting but not resetting, do you think I am correct??

Thanks to all Larry

[MRMRacing]

I have checked this further and the resistors where necessary only with a special mix of older sensors and newer trigger wheels. Ian is very familiar with the Subaru's so try to test as he advise.

[Wroom]

Are you sure the cam wheels and the timing disc is both of either the old or new Subaru timing pattern? And that you have selected that pattern in the setup?

I ask because a friend of mine had precisely that problem on a Hybrid new/old Subaru engine.

[Madevelopments]

Does it have non factory pullies on it as I have seen these cause problems with sync because of a bad design of aftermarket pully.

Mark

[Ian]

13Larry, if your scope has an External trigger input option connect it to an Injector output. If the ECU is reading RPM it will output injection pulses so you will get 1 trigger pulse every 2 engine revolutions. That will give you a full revolution of the camshaft.

As Wroom said, you cannot mix crank and cam wheels from the different series. 36-(3x2) crank must go with fixed 1+1 cam and 6 tooth crank must go with fixed 3+1+2+1 cam. The E-throttle models have 36-(3x2) crank and 4-1 variable cams, this pattern is only available with development CPU's.

Autronic reluctor interfaces have two independant floating inputs. 97-00 Subarus have a common wire for the crank and cam sensors. This common wire and the floating inputs with the older design interface necessitated an additional load resistor across each inputs for correct operation on these models. The WRX99-00 PnP boards have the resistors on the boards.

The WRX01 onwards cars have gone back to completely independent wiring for both sensors and the while the the resistors are not required on WRX01-05 ECU's (with a new design reluctor interface) they are fitted as a carryover.

[13larry]

Hi Ian, This engine is from a 2003 usdm WRX, I do not believe that I have changed any crank sprocket or the camshaft pully. I am running a flying lead set and none of the factory wiring since, the engine is now in a GC8 Subaru. If this is the case when I trigger on the injector 1, how many pulses will I see with the 2003 cam sprocket? You mentioned that I could maybe be off a tooth, how will that show up? Larry:)

[TheBoostCreep]

Thank you all for helping Larry on this one. I have been over there today and I am positive that the facts are straight from him. Larry is very good with electrical concepts, better than any other customer I have had.

I have a 2002 WRX with the same trigger discs from the factory and we are currently investigating this and getting the oscilloscope trace from my car as well. I have also set the non plug in SM4 in 3 other Subaru engines without any problem using Cosworth's cam sensor holder (the Subaru variable cam sprocket has proper teeth and pattern).

Today we tried a new ECU. We checked the wiring for short to ground or open faults. Switched polarity on both plugs. Looked at the oscilloscope pattern from the camshaft sensor.

The interesting thing is that we do get a reading by connecting the cam sensor to the crank sensor input signal wire. Larry will try connecting the crank sensor to the cam input for more information.

[Ian]

Here is a scope trace of just over one crank revolution (from single gap to single gap) with the cam pulses. The 2nd revolution repeats the crank but there are no cam pulses in the 2nd revolution.

If the cam gear is retarded by a tooth then the left side cam pulse does not occur in the single gap and the middle cam pulse occurs in the 2nd gap instead of the first. If the cam sprocket is advanced a tooth then the cam pulses still occur within the same gaps as in the standard position.

[13larry]

HI Ian, I will hook up the scope today. Thanks for the great picture. I will give you an update after that. Larry

[13larry]

Hi Ian, Here is the scope traces, Please comment. Thanks Larry

[Ian]

Actually, after looking at it again, it looks good.

[13larry]

Ian, You answered me so fast that is really service. Thanks

Sorry I can not tell that it is but I will trust you. Thursday I will pull it apart, details to follow. Larry

[Ian]

The pulse just fits in the window but the amplitude looks a bit small if that scale is 500mv / div. I think someone said that is a non-standard pulse ring. What is the clearance to the sensor? You might have to reduce the spacing between sensor and teeth.

[Chris]

Hi Ian, Here is the scope traces, Please comment. Thanks Larry


Please ensure the scope or probe is not set to divide by 10 on that channel.

Chris

[13larry]

Hi Ian, I also thought the timing of the pulses was good. Yes I thought the pulse was a little weak. However I tried another sensor and got the same results. This is stock WRX cam sprocket and pickup, distance is not adjustable. Any ideas? Larry

[13larry]

Ian, Also the only thing connected to the cam sensor was the scope probe, so it was not loaded. Larry

[13larry]

Hello, Here is the latest photo from the scope. I had a bad probe on the cam signal last night, the amplitude is actually larger than the crank signal. Please help. Thanks Larry

[Wroom]

I read somewhere in the huge amounth of diagnostics info in the workshop manual that the signal amplitude must be bigger than 400 milliVolts, and it seems to be okay. But also read in another chapter that it should be +/- 7 Volts nominal on idle. Confusing.

Also read that it should be 30 events on the crank signal on a full crank cycle, and it is.
And the short events shall be 10 degrees and the long events 30 degrees. That is 27*10+30*3=360 degrees, which seem correct. ;)

Also, the ending event in the two long gaps on the crank signal should occur at 10 crank degrees BTDC.
And the single long gap ends with an event at 40 crank degrees BTDC.

And the cam events should occur at 50 crank degrees BTDC.
That is about where they are on your scope, but that they maybe should be a tiny bit to the left.

It's late here in Sweden, so i could be missing something here, but have you shaved the heads lately? ;)

[Ian]

Hello, Here is the latest photo from the scope. I had a bad probe on the cam signal last night, the amplitude is actually larger than the crank signal. Please help. Thanks Larry
The -ve zero crossing of the cam pulse shold be before the -ve zero crossing of the tooth after the gap. In this scope trace (compared to the earlier one) I think the cam pulse -ve zero crossing is very close to or just after the crank -ve zero crossing.

In the absence of anything else obvious you are going to have to try advancing that cam a tooth. Or has this engine got adjustable cam gears that you can advance without much problem.

[Wroom]

In the absence of anything else obvious you are going to have to try advancing that cam a tooth. Or has this engine got adjustable cam gears that you can advance without much problem.
You mean setting the gear one tooth earlier on the belt, and readjusting cam timing in the cam gear?

Since an adjustable cam gear will only reposition the cam, and not the notches on the rim of the gear that set the cam sensor timing.

[13larry]

:)I just retard the crank sprocket, one tooth, so all the cam timing marks line up and this has now given me 24msec sync pulses with 128msec crank pulses (19%). With this change will I be able to correct the crank timing? Should I leave it this way or modify the cam sprocket? Or should I only move the left intake cam forward. Please let me know the best solution. This is a rally, hill climb machine needing the max power, featured in the Jan 2008 issue of Sports Compact Car. (The real deal) thanks Larry

[Ian]

One crank tooth !!! That is 30 degrees which wiil move the cam pulse to the opposite side of the gap. But Cyl/Sync % is normally around 33%. Can you check all your marks again, particularly with the crank in the +90 position where all the marks should line up.

Wroom, you are right, moving the adjustable side of a cam gear does not move the pulse. Unless you file out the standard locating pin hole and that is not recommended for repeatability of location.

[13larry]

Ian, Okay that was a lot. Reinstall timing belt, new belt last night, still no sync. So I moved the intake sprocket 1 tooth forward. 19-20% Sync. I noticed that I am about 1/2 tooth off with this sprocket. Stock with marks retard 1/2, using next tooth advanced about 1/2 tooth. Larry

[Ian]

Ok, if it has a Sync/Cyl % then you should be able to get it going. As it is an SM4 and not a PnP board you may have to move the Cyl I/P lead parameter. I think you have 70 in your cal, so either 70, 250, 430 or 610 will get it spark in the correct place assuming you have wired the coils in firing order.

[13larry]

Hi Ian, here is the latest photo. I also check spark and it is present on all four, what will this do to power advancing only one cam? Larry

[Wroom]

That should work.

But it is some degrees earlier than it should be.

Either you have shaved the cylinder heads a bit, or installed thick head gaskets. That affects the cam timing.

You should set cam timing on both cylinder banks equally. Maybe you need adjustable gears?

And, as you probably know, early cams give the engine some extra torque with the penalty of being a bit rough in the tone and less power in high rpm's. And late cams will do the opposite.

[13larry]

Hi Wroom, That is weird, engine came from a wrecked 2003 WRX with 15k miles, other cam marks line up perfectly. I have ask my certified Subaru mechanic and he said he has seen this a few times before. I have been looking for adjustable cam sprockets but they are real costly. Which cams are earlier? Any in the USDM market? Thanks for helping Larry

[TheBoostCreep]

Congratulations Larry on getting closer to the solution. Perhaps we will be on the Dyno Dynamics sooner than later. Give me a call today.

I have not had to mess with oscilloscope traces on Subaru engines before and this is very good information.

[Wroom]

I have ask my certified Subaru mechanic and he said he has seen this a few times before. I have been looking for adjustable cam sprockets but they are real costly. Which cams are earlier? Any in the USDM market? Thanks for helping Larry
For the Mits EVO there are different lateness on the cams available for especially EVO8 and EVO8MR models to play with. But for the Subaru i dont know any cams with different "lateness".

I think you may have to invest in four adjustable cam gears and an evening in the garage with indicators and timing discs to set it perfect.

The problem with shaving the heads, or on the other hand using very thick head gaskets, is that one often find that the cam timing moved half a notch. This is because the distance from the left bank exhaust cam gear and the crank varies, and that the belt tensioner takes up the slack on the other side. So there is a shift in the cam timing. And it will be twice as wrong on the right cylinder bank as on the left one, because the distance between the two heads will vary with the deck height and gasket thickness on both heads.

So i guess one easily can end up with the left cams half a notch wrong, but the right cams can be set to the timing marks. But then the number of teeth on the belt between the heads will not match what the workshop manual says.

[Franz]

For what it is worth, the left side cylinder case is typically .002" taller than the right case half, when measured from the crank centerline.

[sudnrush]

what did you find was the fix for this problem?

[13larry]

It was all mechanical. I had to adjust the timing pulleys so more length was taken up on the left side of the engine. Once I did that everything workedl.