[Success Story] Subaru SVX with custom RusEFI

[MHTSOS]

8.4/1 compression ratio calls for big boost. Even 8.8/1 can do 250kpa on a stroker EJ20.

Στάλθηκε από το VOG-L29 μου χρησιμοποιώντας Tapatalk

[Dron_Gus]

8.4/1 compression ratio calls for big boost. Even 8.8/1 can do 250kpa on a stroker EJ20.

Suba maniacs say 1bar is safe for pistons and rods used in this motor.
Any suggestions on turbo?

[MHTSOS]

I guess the pistons are from a EJ22 NA since only 22B had a turbo version of this engine. I don't know much about these particular engines since they were never sold around here. Generally EJ16 and EJ20 NA pistons don't tolerate much boost. People usually use 0.7 bar max but at much higher compression ratio than your engine. With 8.4 CR you'll be fairly safe from detonation but you'll have lag at low rpm. Having 3.3 liters will help a lot but before full boost the car will feel slower than before. The rods I guess are the same as the EJ20 NA and they are good up to 6500 rpm. They can tolerate a good amount of boost but no high rpm.
As for the turbo I have 0 experience with 6 cylinder Subarus. What's your empty space in the engine bay like? Do you plan to mount it in the same position as EJs or are you going for something extreme? What about exhaust manifold?

Στάλθηκε από το VOG-L29 μου χρησιμοποιώντας Tapatalk

[Dron_Gus]

Yes. Pistons from EJ22NA. Seems EJ22 has another combustion chamber shape. Thats why its pistons has bigger pool.
Rods are the same as on EJ20G, EJ20K and so on. AFAIR Subaru were using same rods for NA and turbo engines until late 1990x.
No space at engine bay at all.

Screenshot from 2022-02-26 01-13-57.png (3.14 MiB) Viewed 20396 times

There is almost no space between firewall and engine. So stock subaru's turbo place is not a solution.
Move battery to trunk? Narrow front-mount intercooler? Or some water intercooler?
Each head had has it's own iron cast 3 to 1 manifold. Then two pipes merged behind gearbox.
So this manifolds can be rotated 180 and routed to front mount turbo. But this will eat clearance.

xhAAAgE97-A-1920.jpg (217.5 KiB) Viewed 20396 times

[MHTSOS]

Yep, that's tight in space. If you place it where the battery is now how are you going to route the downpipe and the rest of the exhaust? Unless you go full mad max and just put a straight pipe through a hole in the fender. In that case you can also mount it low on either side of the radiator and use a scavenger pump for the oil return. This also makes it very easy for the intercooler piping.

Στάλθηκε από το CLT-L09 μου χρησιμοποιώντας Tapatalk

[Dron_Gus]

Yep, that's tight in space. If you place it where the battery is now how are you going to route the downpipe and the rest of the exhaust? Unless you go full mad max and just put a straight pipe through a hole in the fender. In that case you can also mount it low on either side of the radiator and use a scavenger pump for the oil return. This also makes it very easy for the intercooler piping.

Definitely no screamer pipes.
A bit of googling shows that these is possibility to mount turbo at stock subaru's place. But uppipe will have to much 90 rotations. And loooong pipes to front mount intercooler.
Going to pick up car from machine shop next week. This issue costs me $2K total. So no turbo in near future. Budget is over.

[Dron_Gus]

Engine was rebuild.
Then I had a problem with PCV, that costs me engine out and in again and set of gaskets.
Yesterday I done my first ride on RE this year.

Today it is +30C in Saint-Petersburg and I need to figure out why AC is not operational with RE.

Compressor clutch is not get activated.

Wiring is quite simple and straightforward. But I'm stuck.

AC and fans wiring

AC.gif (156.89 KiB) Viewed 19249 times

+12 to AC clutch relay coil comes from IGN switch through "LowHi pressure switch". I assume I should be closed. At least AC works with stock ECU.
Ground to AC clutch relay comes from ECU pin b8.
I use "General Purpose PWM 2" to drive this pin. I see low level on this pin during tests on table.

Settings

Screenshot from 2022-07-23 12-30-14.png (49.81 KiB) Viewed 19249 times

Input from "Intermediate switch" is currently ignored.
There are also two signals between ECU and AC unit: "A/C switch" and "Auxiliary air control valve". From documentation it is not clear if ECU is sourcing these signals or receiving.
Tracing stock PCB shows that "A/C switch" signals goes to some unknown soic IC through 0.5W 2.7K resistor. Looks like ECU outputs this signal?
"Auxiliary air control valve" I was not able to trace.

Does anyone has experience with old Japanese cars AC systems? What is purpose of these signals? Who is source of these signals?

[Dron_Gus]

Ok. AC problem is HW problem (leaking system).

Now I have crank-run-die problem...

Screenshot from 2022-08-21 18-34-02.png (61.53 KiB) Viewed 18970 times

Tune https://rusefi.com/online/view.php?msq=1172
Log https://rusefi.com/online/view.php?log=987

Should I try latest greatest SW? Or this one look like HW issue?

[AndreyB]

There is a chance of trigger sync loss. I am confused why do we not see latest gap ratios in your log?

image.png (42.82 KiB) Viewed 18967 times

[Dron_Gus]

Ok, I was able to start engine, helping with foot. Then I slowly released gas pedal - engine idles at 1200. And then I start closing IAC valve and now engine works fine on 670 RPM. I see no gaps in Engine phase graph.
Log https://rusefi.com/online/view.php?log=988

[AndreyB]

Not causing you any problems but https://github.com/rusefi/rusefi/commit/25778f089065bc68a9d58462d9631066ec1df4c7#diff-cb45a588128ccaeda59b2b544ffc517eb73942c256477bc1f50387ba8b45a243

[Dron_Gus]

New exhaust.
Ready for RusEFI AFR vs reference testing

[Dron_Gus]

Ok.
I'm still getting trigger errors due to unreliable CAM sensor. Same issue on stock ECU.
So lets try to use CAM sensor only for initial sync. This is possible?

Screenshot from 2022-10-16 21-48-00.png (75.22 KiB) Viewed 18461 times

Crank has 12 tooth. Should I set 12 to "Total tooth counter (number)" or 24?
"Skipped wheel location" is it option to set where "custom toothed wheel" is located? Skipped = Custom?
How can I set CAM shape? It has single tooth.

[Dron_Gus]

Trigger wheels once again.

[AndreyB]

No-no-no-no, let me call you.

Also https://github.com/rusefi/rusefi_documentation/issues/241

[AndreyB]

What you really need is https://github.com/rusefi/rusefi_documentation/issues/227

[Dron_Gus]

Original composite SVX trigger/reference good signal.
coils times are almost the same. ignition time also very close.

[Dron_Gus]

Switching to "12 tooth crank" on crank and "single tooth second half" for cam:

Screenshot from 2022-10-19 16-19-44.png (155.86 KiB) Viewed 18391 times

[Dron_Gus]

Results:

fuuu2.png (30.43 KiB) Viewed 18390 times

[Dron_Gus]

Adjusting trigger offset in console does not actually moves green line.
Ignition time are quite different per cylinder...
What am I doing wrong?

[AndreyB]

Is your crank Hall or VR?

image.png (52.49 KiB) Viewed 18388 times

Not sure if that's 1% or 90% of your issue.

[AndreyB]

https://github.com/rusefi/rusefi/issues/4681

[Dron_Gus]

Is your crank Hall or VR?

Not sure if that's 1% or 90% of your issue.

Good question. It is VR.
Now it looks 50% better. But enjine is still unstable and I hear cylinder firing skips.

[Dron_Gus]

CAM use also only Rising edges.

[Dron_Gus]

Tune with composite original SVX trigger:
https://rusefi.com/online/view.php?msq=1270
Log of idling and reving.
https://rusefi.com/online/view.php?log=1082

[AndreyB]

image.png (2.36 KiB) Viewed 18375 times

image.png (2.46 KiB) Viewed 18375 times

[Dron_Gus]

image.png
image.png

Absolutely agree. This is going to be fixed.
As we decided on phone call I did logging with "tune that is known to work"...

[Dron_Gus]

One issue with Subaru documentation is that they don't care about actual VR+ and VR-. This is another area where issue can exist.

[AndreyB]

https://github.com/rusefi/rusefi/wiki/Trigger-Configuration-Guide#vr-variable-reluctance-sensors has relevant wisdom

[Dron_Gus]

https://github.com/rusefi/rusefi/wiki/Trigger-Configuration-Guide#vr-variable-reluctance-sensors has relevant wisdom

Problem is that crank wheel has no skipped tooth. while correct crank polarity is the question.
Seems I need to attach oscilloscope.