subaru FB25 engine interchange

jdf60 said:

In reply to mildensteve :

Hardest part was stabbing the engine back onto the tranny by myself without a load leveler. Took me an hour to get it lined up just right. Would be much easier with two people or a load leveler of some kind. 

 https://www.harborfreight.com/3-4-quarter-ton-capacity-heavy-duty-load-leveler-67441.html?_br_psugg_q=load+leveler

Not the same one I got from HF years ago but worth every penny.  I used heavy duty links directly to the leveler instead of the dangling chains to shorten the whole thing.  https://www.harborfreight.com/38-in-quick-links-3-piece-68899.html

It is cool here today and I was just able to put in a good hour before running out of breath. I still need to finish pulling the radiator and double check everything else, then I guess take off the intake manifold in prep for separating the engine/transmission.

Having (old) TGV hardware in hand to play with should help with questions about how to proceed, even if it takes me a while to get the engine all the way out.

I have a hoist, engine stand, (NAPA) load leveler and pallet to put the old engine on. I have never worked on a boxer engine before. What is the best way to mount it for bench work? For each project I try to find something new to have a "learning experience" about :-) Just about to the fun part.

In reply to Skenton :

I was going to mount mine on my engine stand but found out I needed to remove the flex plate and crank trigger wheel before hand (wouldn't be able to get the plate off).  So I just placed the engine on a Harbor Freight furniture dolly.  I can roll it around the garage and disassembled it on the dolly.  Guess I can still mount it on the stand.  I did have an interesting sight greet me when I took off the right valve cover...

Broken valve spring, rocker and spring cap fell out.  Interestingly enough it was running fine when I pulled the engine.  The cam lobe was a little brown indicating it had been that way for a while.

How big of a mess are the transmission cooler lines? I got the radiator ready to pull yesterday but did not have my waste oil container with me so stopped - working in my daughters garage. I was thinking about looping a hose back on each end to minimize leaking.

Skenton said:

How big of a mess are the transmission cooler lines?

Nothing that a Campbell's soup can can't handle.  I disconnected mine at the hose clamp vs the radiator b-nut.  drained straight down into the oil catch pan. 

OK very anticlimatic, just a few tablespoons of oil. Radiator is out and most of the intake stuff is disconnected. Should get it removed next time and will take a look at the old TGVs. I never heard back about my offer on the damaged new one so I guess keep hunting.

In reply to mildensteve :

That's odd that you have that. Not sure if you've read all my posts but I also had a rocker just laying in the side of the head when I pulled the valve cover. But my spring and valve and everything else was fine. The "lifter" or whatever it's called was laying there too Undamaged. It's not really a lifter but a pivot I guess. Engine was also running fine. Had the knock that I still have not figured out. 
 

In reply to Skenton :

that's some progress at least. It's sure not a trivial project!

intetested to see what you figure out on the tgv still and If I can help with anything. I'll try to look around for a newer style tgv also. 
my engine seems to be starting snd running good. Has a little rough idle when cold. Maybe tgv related, not sure. But otherwise is running good. Still planning to check engine codes as I'm sure I have the tgv codes going, among others maybe. 

Upper and lower manifolds removed, just got home with old TGVs in hand. Next on the list is to clean them up a bit and then open them up for diagnostics. I will check magnet poles, left/right symmetry, construction and as soon as I rummage around and locate a 1K Ohm resistor and 5/12V sources quantify the output signal voltage vs position. Plus all those other questions I asked :-)

So we all have 3 cylinder old engines, huh?

I did not take a picture, but I found most of a piston rod in pieces inside the oil pan. The piston is still in the bore with the remainder of the rod dangling down. When I tapped it with a finger it swung back and forth so the wrist pin is not seized. It looks like the rod cap cut loose. It still cranked over OK once I got past the tripped security lockout but I did not try to actually start it due to lack of oil.

I wonder if the knock you heard could have been rod slap due to worn bearings or wrist pin bushings?

Did you get the cooling system dealt with and stuff pretty much buttoned up for the moment? Good to hear the 2015 swap is running well enough for now. I will post what I find about the TGV situation.

Edit

Having trouble with the TGV actuator screws too, I hope my impact screwdriver is over with the tools at the car 'cuz I cannot find it here.

In reply to Skenton :

Yes I believe we all have the older forester engine. I'm going to tear mine down and possibly rebuild if all it needs are bearings because apparently they are worth $2500+!!

and yes cooling system flushed and new radiator. Working great so far.  Took a chance and reused old water pump because I didn't want to wait for the new one. And it's not bad to change later on if it leaks. 
 

I do have a cyl 3 misfire code. Is that likely a bad coil pack and if so, anyone know if I can use the 2015 coils?  The plugs seem different if I remember right so I'm guessing not. 
 

I also have the intake flap open codes. As expected. Doesn't seem to affect the engine operation so far. No limp mode or anything. 
 

I thought about rebuilding but parts are scarce. I saw a few scattered gaskets, rings and bearings but no rebuild kits or master kits. Places said out of stock due to supply chain disruption. If your engine is otherwise in good shape  that may be fine. Mine is worse off :-(

Try swapping two coils and see if the misfire follows the coil or stays with the wiring and injector etc. I believe you are right about the coil connectors changing.

Glad to hear no limp mode!

Edit

If cylinder 3 was the one with the failed rocker it is possible the computer was not using that injector to protect the cat, in which case it might have had a long time to get "stuck" or "crudded up". If that turns out to be the case, maybe pull that injector and try cleaning/flushing it manually? Have you done a compression test?

In reply to Skenton :

Good idea to swap injectors! 
 

 Cylinder 3 wasn't the failed rocker. 

In reply to Skenton :

No I didn't do a compression test on the newer engine. I don't have a tester and it didn't occur to me. 
 

I did swap coils and still got the 3 misfire. Might just swap an injector from the 2015 engine if they are the same. The connectors are the same.  I guess it could be something in the ecu or harness. I never checked the codes before I swapped engines because I had lost my reader.  Darn it, so many "hindsight's". I guess I'm out of practice working on engines. 

btw-I just fully read your post,  your failure was quite a bit more dramatic than mine!  Had that bearing seized up due to lack of oil? 

jdf60 said:

In reply to Skenton :

I'm going to tear mine down and possibly rebuild if all it needs are bearings because apparently they are worth $2500+!!

I think I will rebuild as well.  I am pretty sure I can get all the parts from ebay and rockauto.  I am concerned about my oil burning issue.  Gotta figure out how to prevent that via new pistons & rings or just rings.  If not, it will be a fun learning experience.

I bought my Forester cheap at an impound auction as a "project". It was flagged as no-crank/security, but the oil was low so I was not surprised by what I found in the pan.

From what I read the original oil problem was a design/QC issue with the rings, so I would guess that modern rings would deal with the oil consumption. Bearings while you are in there too.

There is a free Torque Lite app for android tablets and phones. The Pro version is just a couple of bucks. Uses a cheapo Bluetooth "Elm" OBDII dongle available from Ebay. Code reader and live data like RPM display. I gave my daughter one when she took off cross country in her rebuilt Prius project. She used it with Torque on her smart phone to track the battery pack state.

Edit

We had electricians working at the house today so I did not get to the car. Maybe tomorrow.

Got the TGV actuators off. Sprayed Pblaster on both ends of the screws/holes and let it sit for an hour then the impact screwdriver bit spun them right out.

Preliminary test with the magnetic pickup probe indicates the black magnet poles are reversed left/right. Probe seemed attracted to clockwise end of the left one and counter-clockwise end of the right one. I will check again once the are labeled and off the shafts and away from other metal/interference. It looks like the magnets are mounted by pushing a tab into a slot.

Edit

For the old TGVs the gear parts are identical left/right including the clocking angle of the indexing holes.

The black magnet parts *LOOK* identical including the clocking angle of the indexing holes, but I believe the magnetic poles are reversed on the long axis of the magnets. I need a smaller probe to double check after I get some sleep though. Tomorrow I will see what I can find or maybe magnetize a needle or something. I will also take pics.

If this holds up I think just swapping the two black magnet pieces left/right would invert the open/closed signal to the computer. The proof will be the measured signal voltages once I build the test circuit.

What the new one piece manifold with the white magnets does is still an open question.

Edit2

I added a small cotter pin to the end of the magnetic probe for a finer tip and confirm the magnets are different from side to side. See pics above.

Jdf60:

If the white magnets in the 2015 one piece manifold TGVs follow the pattern of the older black ones where they look identical but have the magnetic poles reversed left/right, which I think is likely given that the actuator sensors are the same, then swapping them left/right but otherwise unmodified will likely keep the TGV codes away after you initially clear them. This assumes you have also either swapped the harness left/right or swapped pins 4/5 on both TGV connectors to trick the Forester computer to reverse the open/close directions as required for the 2015 Legacy engine.

A bit of a bother to do but seems safe enough from the point of view of the sensors.

In reply to Skenton :

It wouldn't hurt to try for sure. 

I think the magnets are sloped instead of being symmetric like the older style so hope that doesn't complicate it. also it would be hard to work on in the car with the little pieces and tight access so I'll have to remove the intake,  likely  the pins would be easier to swap than the entire connector for sure.  Could you post a pic showing the pin layout?  I think you had before but this post is getting very long! 

Id really prefer not to have the check engine light all the time since it causes owners to ignore it. (My son).  But doesn't seem to effect the running at all. As expected I think. 
 

still looking for the misfire issue. It's not the injector either. I swapped those and still cylinder #3. I suspect the connector might be bad. It's not a constant misfire but still annoying and not good for the engine. I did confirm that injectors won't swap from the engines. 
 

also noticing these used parts are expensive. It might be worth the effort to sell parts on ebay if the engines can't be cheaply rebuilt. 
 

im trying to figure out how to add an oil pressure gage. Any ideas there? 

The Legacy engine has an oil level sensor connection on the upper oil pan that the Forester engine/harness does not. Some people hooked that up to an LED indicator. As for oil pressure, #23 below the oil filter is the oil pressure switch. Maybe Tee off of there for an after market oil pressure gauge? #24 right above it is the engine oil temperature sensor.

I checked around with my friends to see who still had some discrete resistors, I should be able to get some tomorrow to build the test circuit so I can check sensor output voltages for the old TGV. If you open your new ones up again, side by side images of the white magnet pieces would really help nail this down. Do they look identical? Be sure to mark which is which :-) Checking the magnet poles with a random magnet and unbent paperclip on the end along the lines of my pictures should tell if the poles are reversed left/right.

I have a set of cheap Harbor Freight injector test lights that clip inline between the harness and injector. Also one of their inline sparkplug lights. Might help with wiring checks. FYI. At least one my 2011 the coils are controlled directly by the ECM with nothing in between.

Did either of you folks get the private message I sent to both of you via the forum? That feature took over a week to be enabled after joining. "Contact an administrator or just wait a few days". I think it uses the email address we register with.

Edit

Got the resistors will try to test later today.

Edit2

Stock configuration gave 5V when open 0V when close and in between voltage when partially open.

Will swap magnets left/right and repeat the test tomorrow. I expect the voltages to reverse.

Skenton said:

Did either of you folks get the private message I sent to both of you via the forum?

I did but thought it was a spam / virus.  Now that I know it is legit... cool.  Thanks. Very helpful.

OK, swapping the old black magnets left /right worked and inverted the open/closed signal on pin 1

To summarize:

The 2015 and later Legacy engines reverse the direction of the TGV butterflys relative to our older Forester engines.

We must change wiring a bit, such as swapping the motor 12Vdc polarity to actuator pins 4/5 to fake the old Forester computer into going the right way for the !egacy engine.

In addition we need keep the Forester computer from throwing codes by giving it the expected signal voltages for open/closed on actuator pin 1

In normal operation the old black magnet TGV signal voltage is close to 0V  when closed, close to 5V when open and in between when partially open.

The actuators contain bipolar Hall effect sensors that feed the signal on pin 1.

A common way to invert the output of those type of sensors is either flip the sensor over or swap the magnet poles that rotate past it. The sensor in the actuators is fixed and cannot be flipped over. The black magnets for the left and right TGV butterfly assemblies appear visually identical, including the clocking angle of the slot for mounting on the shaft flat spot. However, probing with an external magnet shows that their north and south poles are reversed.

Swapping the black magnets left/right does in fact invert the measured voltages when open/closed. See picture for test setup. What exactly the newer TGVs with the white magnets do in unknown but incompatible with the old Forester computer. Hopefully measurement of the white magnet TGVs will tell if swapping or replacing magnets will "fix" things.

Edit

I cut the end off an old USB cable for sensor power. Green and white are data and can be ignored, red is 5Vdc and black or bare wire is ground. I plugged it into a spare USB charger block to turn the tester "on". To change positions for each test I removed the actuator, moved the butterfly manually and then reinstalled the actuator. The resistor is 1K Ohm but anything from 0.5K to 2.0K should work. I slipped some snug heat shrink tubing around the actuator pins so I could just stuff wires in temporarily for testing without solder or clips. Measure the voltage between pins 1 and 2. Yes, that is white duct tape :-)

https://s3.us-east-2.amazonaws.com/prod.mm.com/uploads/2022/10/18/1666110119_20221018_104707_mmthumb.jpg

In reply to Skenton :

i will probably never need the info in this thread, but i am thankful that smart and motivated folks like y'all are part of this community.

Skenton said:

Yes, that is white duct tape :-)

It isn't a fix unless there is duct tape involved.

Wow.  If you don't figure it out, not sure who will.  Now to get you a single-piece intake manifold. 

I got back to the car today, loosening engine/transmission bolts. While in there I noticed a cable the engine removal instructions did not mention. Turns out my Forester has the all weather package block/oil heater with wiring running from the bottom of the engine to the lower grill for plugging in an extension cord. Missing that would have been fun with the hoist :-/

I still cannot locate a one piece manifold with new TGVs locally. Grumble.

In reply to Skenton :

One idea to get ahold of one would be to buy one on eBay for about $130. I saw a few in that range. And then resell yours if/when you buy your replacement engine to get your money back. Just an idea. The upside is that you would have a working intake ready to go when you get your new engine.