Carb'd (Now EFI) EJ22 Into '69 Bug

[Steve Arndt]

You are kidding me if anybody with a suby conversion is running bug CVs.
Either way, just copy their design. done.

[Buggin_74]

Are you serious??

I can show you plenty of bugs with stock CVs running 200+hp EJ20turbos that have never had a failure.

Jaks bug has gone over 6 years and 174,000kms on stock CVs with god only knows how many track days.

http://www.youtube.com/watch?v=PnEJiY3yEWk

[Mooosman]

You are kidding me if anybody with a suby conversion is running bug CVs.

Unless you're abusing the car (side-stepping the clutch constantly, drag racing it every week, etc.), I don't think you'd have a problem with the stock CVs.

[Mooosman]

Nick give it a shot anyway it may work fine for you, I used a door magnet switch which i got the design for from somebody on the subaru-vanagon website who it worked fine for, maybe I was just unlucky.

I'm using the most powerful magnets I could find at the hardware store, and the Subaru reed switch epoxied to a simple L bracket. When hooked to my multimeter, it picks up the magnets easily. I don't know how it's going to act at speed, though. We'll see.

Now, I have an automatic ECU that is currently in the Park mode (Park wire grounded). Now, once I add the VSS, do I have to ground the Neutral wire at the ECU, or will the ECU pick up the VSS while in Park?

Nick

[Big Dave]

My Ej25 conversion was similar, except its ECU is from an 03 automatic. Without the VSS hooked up, the car would run OK, but hit the rev limiter at 4K RPM and it would tend to die when coming to a stop.
My harness builder sent me one of these

and some rare earth magnets.

I used JB weld to "glue" them to the drive flange and made an aluminum bracket to hold the sensor.
Have full RPMs and no dying at stop lights ever since.

[Mooosman]

*UPDATE* VSS Time!


Okay, for the VSS I salvaged the reed switch from the Subaru donor's cluster, and decided to see if I could get it to work with some magnets. Testing on the bench showed that the reed switch closes with no problem when a magnet is passed over it, but that the distance at which it would switch changed based on the orientation of the reed switch to the magnet (I.E. what direction the magnet passed over the switch). I found that the switch worked the best when the magnet passed over it lengthwise. So, I had to mount the switch on the transmission vertically, so the magnets on the CV joint would pass by it lengthwise.

The rest was just a matter of making a simple L bracket for the switch, epoxying the reed switch to the L bracket, and then putting a thin coating of silicone sealer on the switch and connections to waterproof them. We then epoxied 3 small stacks of magnets (from the hardware store) to the inner driver's side CV joint. When it was all mounted, the wiring is a very simple 2 wire hookup: 1 wire to the 'VSS' pin on the ECU, and the other wire to ground.


Results?

Well, with no VSS hooked up and the ECU in 'Park' mode, the car started and ran perfectly, but would die when coasting to a stop and pushing in the clutch. The car would immediately fire right back up, but it was annoying having to restart the car every time I came to a stoplight. We wired in the makeshift VSS and started driving, and haven't had any stalling since. The ECU is still in Park mode, but it doesn't seem to matter. The idle drops very low when you first push the clutch in, but the ECU immediately brings it back up to normal, and the car idles steadily after that. So, it seems the VSS is working as intended. I haven't checked to see if the ECU is storing any codes, but the check engine light has not come on so far. Now, I don't know if this homebrew VSS loses its reading at higher speeds, but in my mind it doesn't matter; as long as the VSS reads when you're coming up to a light (say, under 30mph), it will serve its purpose. As long as it reads the signal well enough to keep the engine from stalling at the lights, I'm happy, and it seems to be doing just that. We'll see how reliable this solution is long-term, but for now it seems to be working.

So remember: save EVERYTHING you can from the donor car!

Excuse the greasy transmission; it's next in line for an upgrade (to a close-ratio box)






Lastly, a couple pics of the wiring harness installed in the car. The main harness will be hidden by a combination speaker box/ECU box that I'm going to build in the rear parcel shelf area, which will also house a Vanagon heater box at some point. The ECU, main relay, and fuel pump relay are all easy to get to, as are the fuses (3 total) for the EFI system.

[itawolf]

Hey Joel

Hope u dont mind--but here's ur setup that I followed

Cheers

[Buggin_74]

It's all good.

Nick,, that should work fine, the stock one read 5 signals each revolution so only 3 it should handle fine at higher speeds cos it will think the car is moving slower than it is.

The number of pulses doesnt matter with a gen1 car, it just needs to know that the car is moving or stopped.
The later stuff uses the speed sensor to dictate what it does with the radiator fans.

If it hasnt thrown 24 when you slow down you're good to go.

[Mooosman]

If it hasnt thrown 24 when you slow down you're good to go.

I got that code when I drove around before installing the VSS. I haven't cleared the code yet to see if it pops up again with the VSS installed. With 3 magnets instead of 5, it may actually be reading at higher speeds. When testing on the bench, though, it seemed that the reed switch didn't react to the magnet very quickly, though that may have just been my multimeter (maybe the meter took longer to show a reading on the screen).

It seems to be working fine right now, though. Now, I just need to delete my separate control unit for the radiator fan, and go ahead and wire in a relay to let the ECU control it. That should finish up the wiring.

Nick

[Buggin_74]

Have you got a procedure for clearing codes?

The ECU from my donor car was great, I just had to disconnect the battery for 30mins and the code was gone
Obviously the CMOS battery in it was almost dead.

This new ECU I'm running, that code I made it throw in my youtube video is still on there.
I've left it disconnected more than 24hours and it's still logged

Tried the get it to temp and drive at over 10kmh one and that didnt clear it.

[kyle_pc_75]

Have you got a procedure for clearing codes?

Find an independent mechanic (not a Subie dealer) with the proper equipment. I have a guy that does that sort of thing for free as long as I send business his way. Most will do it for a few bucks.

Kyle

[Buggin_74]

I have got a really good subi garage that have helped me out heaps giving me loads of parts and advise, I just wanna try fixing it myself before I go hassling them with trivial crap like this.

It's no big deal I know what the codes are from, just a pain when diagnosing and its throwing 3 million codes and only one is real.

Sorry Nick for the hi-jack

[kyle_pc_75]

Yeah, that makes sense, Buggin. Hope you find a good method. I'll need that info one day!

Kyle

[Buggin_74]

Well 3 days with the battery disconnected did the trick.

Now I just have to not get it wet

My knock sensor housing has lots of cracks as they tend to do and sometimes if i wash it the knock sensor code comes on for the first few minutes

[surfbeetle]

I don't know if this can help you both, but I saw this in another post here on Shoptalk, http://www.romraider.com/ I ended up with a Link ECU for mine that I hope to get running soon.