Magnetic Sensor

Thread Starter

ejs22

Joined Aug 19, 2021
7
Hello all,

First post for me but this forum seems to have a lot of great info! Thanks for any help in advance.

I am putting a 2016 Subaru Outback (referencing as "2016") motor into a 2013 Subaru Forester (referencing as "2013"). There are few differences between the motors, one of them being the Variable Valve Timing Sprocket which is sensed by magnetic sensors to create an analog waveform to send back to the car computer (using 2013). This helps the computer compare camshaft position to the crankshaft. The difference between the two is one sprocket has a protruding metal tab (2016) and the other is "valley" instead.

My question is what will the difference in waveform be?

2013 sprocket has a "valley".
1629429496866.png
I/O signal from maintenance manual. Appears the 2013 waveform reads 5V until it hits one of the valleys and then drops to 0V.
1629429636478.png

2016 sprocket has a tab.
1629429508363.png

2016 waveform is the same, shows 5V until it hits a tab and drops to 0V.

1629429883608.png

The part numbers are the same for the magnetic camshaft position sensor for both model cars, so it is interesting they both create the same waveform.

Any insight would be great
 

LowQCab

Joined Nov 6, 2012
4,023
Are YOU measuring the Waveform produced,
or is this a picture of what the Waveform is "supposed" to look like
on each of the various year-models ?

I'm guessing that these are not your measurements.

I would just hook it up and see if the Motor runs.
If it doesn't,
then start figuring out how you can put the old Reluctor-Wheel on the new Engine,
since You say that You are keeping the old-Computer, or just use the newer-Computer instead.

Keep in mind that the newer Motor may not play-nice with
the older Engine's Computer calibrations.

Are the Compression-Ratio, Cam-Specifications, and Fuel-Injectors EXACTLY THE SAME
between the 2-different year-model Motors ?
If not, You will have to tweak the Tune in the Computer.
.
.
.
 

Thread Starter

ejs22

Joined Aug 19, 2021
7
Are YOU measuring the Waveform produced,
or is this a picture of what the Waveform is "supposed" to look like
on each of the various year-models ?

I'm guessing that these are not your measurements.

I would just hook it up and see if the Motor runs.
If it doesn't,
then start figuring out how you can put the old Reluctor-Wheel on the new Engine,
since You say that You are keeping the old-Computer, or just use the newer-Computer instead.

Keep in mind that the newer Motor may not play-nice with
the older Engine's Computer calibrations.

Are the Compression-Ratio, Cam-Specifications, and Fuel-Injectors EXACTLY THE SAME
between the 2-different year-model Motors ?
If not, You will have to tweak the Tune in the Computer.
.
.
.
Thanks LowQCab. Sorry I should have specified those were not my measurements. They are from the service manual for each car.

The new motor is currently in the car and running great with older computer. Subaru used the FB25 motor for a lot of different vehicles and the "short block" is the same between a lot of model years.

There are two differences though between the motors that are throwing some codes... with the camshaft position sensors being one of them. Before looking into putting the old Reluctor-Wheel on the new engine, I wanted understand how the magnetic sensor sees the tab vs a "valley".
 

LowQCab

Joined Nov 6, 2012
4,023
Since You say that both configurations use the same Sensor,
maybe what You need to look at is the
exact positioning of the Sensor in relation to the Tab or Slot.

The exact positioning can be absolutely critical,
and may be slightly different between the 2-Engines.

Even a slightly mis-positioned Sensor could cause loss of a reliable pulse at higher RPMs.
Worse yet, it could also make any intermittent problems temperature dependent,
as the sensitivity of the Sensor may change slightly with Temperature.
.
.
.
 

bwilliams60

Joined Nov 18, 2012
1,442
Typically on an engine swap, you would bring the ECM along with the engine. That would alleviate a lot of issues. As far as waveforms, I believe they will be the same but you should be able to go to the PICOSCOPE library and look up your waveform for those engines under cam/crank signals. The cam signals are offset to identify left from right but signals should be similar between hills and valley identifier. If anything, you should find the valley to be the inverse of the tooth which would make sense.
 
Last edited:

Thread Starter

ejs22

Joined Aug 19, 2021
7
Since You say that both configurations use the same Sensor,
maybe what You need to look at is the
exact positioning of the Sensor in relation to the Tab or Slot.

The exact positioning can be absolutely critical,
and may be slightly different between the 2-Engines.

Even a slightly mis-positioned Sensor could cause loss of a reliable pulse at higher RPMs.
Worse yet, it could also make any intermittent problems temperature dependent,
as the sensitivity of the Sensor may change slightly with Temperature.
.
.
.
Good point, ill have to check to see the spacing.
 

Thread Starter

ejs22

Joined Aug 19, 2021
7
Typically on an engine swap, you would bring the ECM along with the engine. That would alleviate a lot of issues. As far as waveforms, I believe they will be the same but you should be able to go to the PICOSCOPE library and look up your waveform for those engines under cam/crank signals. The cam signals are offset to identify left from right but signals should be similar between hills and valley identifier. If anything, you should find the valley to be the inverse of the tooth which would make sense.
Thanks, that was my initial thought, that it would inverse the signal. If that were the case and changing the sprocket is not easily done, is there a way to simply inverse the signal through an Op-Amp before it goes back to the ECM?
 

LowQCab

Joined Nov 6, 2012
4,023
Probably not.
The Phasing between the Crank and the Cam Sensors may be critical,
and reversing the signal phase isn't as simple as it may appear to be.
.
.
.
 

bwilliams60

Joined Nov 18, 2012
1,442
I would recommend against it. Cam phasing is very critical on these engines. I stick with my initial recommendation and would put the original engine ECM in this setup. You will thank me later. I have been involved in a lot of engine swaps over the years and it never goes well when you start customizing things unless you have the tools, knowledge and patience to do it right the first time.
 

Thread Starter

ejs22

Joined Aug 19, 2021
7
I would recommend against it. Cam phasing is very critical on these engines. I stick with my initial recommendation and would put the original engine ECM in this setup. You will thank me later. I have been involved in a lot of engine swaps over the years and it never goes well when you start customizing things unless you have the tools, knowledge and patience to do it right the first time.
Thanks for the input and appreciate the advice. I think in this application, I would have to look into how the ECM interacts with, if at all, the transmission (4AT in the 2013 vs CVT in the 2016). Also if it would pass smog in California due to new ECM checks.

Also cam phasing being critical has been mentioned a few times. The cam is mechanically linked to the crankshaft thus phasing cant change, so I assume this in reference to the computer reading where the cam is at and trying to correct the waveform could cause issues here.
 

geekoftheweek

Joined Oct 6, 2013
1,201
Chances are the transmission difference will be another headache unless the transmission has it's own controller... Even then it may take some reprogramming to get it right. A CVT and an automatic aren't even remotely similar other than the reverse gear.

If I understand it right you are trying to use a variable valve ECU on an engine that doesn't have variable valve timing. I would think the cam sensor codes have more to do with the cam not advancing or retarding as it should rather than the pickup.
https://en.wikipedia.org/wiki/Active_valve_control_system

Edit...

Of course if the valleys and ribs of the two are off a few degrees that would cause problems too, but you're still going to have issues with the lack of valve timing.
 

Thread Starter

ejs22

Joined Aug 19, 2021
7
Sorry for the delay on following up on this...finally got an oscilloscope. Image below is of the 2016 motor with a 2016 cam in my 2013 forester body with a 2013 sensor and 2013 ECU (need to use 2013 ECU due to transmission)1635921791357.png

I also a found another reference waveform for the 2013 (I think the one I originally posted above is not correct)
1635922042894.png
I suspect the reason I am getting crankshaft / camshaft correlation codes, is the ECU is looking for a falling edge (at least per my understanding) which is now off by a few degrees due to difference in sprockets between the years, valleys vs tab.

Is there a simple way to take the digital camshaft position sensor waveform above and effectively invert it? 0V becomes 5V and 5V becomes 0V
 
Last edited:

Thread Starter

ejs22

Joined Aug 19, 2021
7
Chances are the transmission difference will be another headache unless the transmission has it's own controller... Even then it may take some reprogramming to get it right. A CVT and an automatic aren't even remotely similar other than the reverse gear.

If I understand it right you are trying to use a variable valve ECU on an engine that doesn't have variable valve timing. I would think the cam sensor codes have more to do with the cam not advancing or retarding as it should rather than the pickup.
https://en.wikipedia.org/wiki/Active_valve_control_system

Edit...

Of course if the valleys and ribs of the two are off a few degrees that would cause problems too, but you're still going to have issues with the lack of valve timing.
To clarify both are VVT...my post should of read crankshaft is mechanically linked to the VVT sprocket which can then alter the timing of the cam
 
Top