presslab

Sound like a problem with your charcoal canister. Try disconnecting it from your intake (and capping the intake side only) to see if that fixes it.

I've taken a peek at the software inside the TCU on my '92. The TCU will not shift to 4th gear until the trans temp reaches 64 F. It will lock up the torque converter at 162 F. The temperature lamp comes on at 230 F. I added a cooler on mine, and it would take longer to lock up the torque converter. So I modified the software and lowered the lockup temp just a bit.

I assume your ECU is also from a 97 EJ22? My 92 ECU is designed to work with the red top injectors, same as the EJ25D. What injectors came on your 97? If you build/buy a SSM cable you will have so much more information. The second thing would be to get a wideband o2 sensor. I have the MTX-L and with EvoScan I can log SSM and WBo2 simultaneously. With the cams you will run lean at idle. Same with the EJ25D shortblock. Combined you're likely exceeding the range of the ECU adjustment and running too lean at idle/low RPM. You said your ECU runs closed loop; is the o2 toggling while idling?

Well I'm not entirely sure. It could theoretically get the timing information from either one, as it takes the integral of velocity to get the position. With the wasted spark I'm not sure it even needs the crank - in the past I had a bad crank sensor (that triggered a CEL) but I never noticed any running problems other than weird starting. For that matter it could probably run without one or the other if the ECU was smart enough. What I was saying originally before we went down this road, is that I don't think this is the problem as the timing belt tensioner was in the same spot, so the cam (and possibly ignition) timing should be unaffected due to that aspect. I wonder if the way the crank is offset ground to get more stroke could cause a change in timing.

Cam timing can certainly change the dynamic compression ratio; I would say in some cases this could be significant, but I don't think that's the case here. In reality what I mean is that changing the cam timing will therefore change the ignition timing because the cam position sensor is giving the ECU the wrong info. Just a small change here will have a big impact. When I say total advance I mean maximum load at ~3500 RPM+. Under light throttle at higher RPMs, the timing is advanced more, up to 45 degrees BTDC, just like you see. You may or may not have seen my other thread where I have posted the stock maps for the "F9" ECU: http://www.ultimatesubaru.org/forum/showthread.php?t=134517 To get the effective timing, add from both the base timing map and the max advance map. This is what the timing will be if the ECU does not detect knock.

I've left both my Subarus on for the time it takes to download the ROM. It's not a problem. If your injectors leaked, they'd leak down every time you turned off your car as the fuel pump should hold pressure with a check valve. Put a charger on the battery and dump the ROM. Or start it up and let it idle for 45 minutes.

That's like my dog in the morning. He'll climb on the bed and go right for my mouth. :-p

Yes, the TCU certainly needs the VSS signal from the speedometer cluster. The second speed sensor is located at the rear of the trans. It needs both of these working to determine if the wheels are slipping. There is an open-loop part of the AWD control, and that's what you are seeing without the VSS. Also, the TCU expects a certain R&P ratio and speedometer drive gear. If these are too far off the TCU might throw a code and refuse to work the AWD clutches. I assume you swapped the rear R&P with a 4.44 unit as well?

Yes, I have about 26 degrees total advance. The stock EJ22E map has about 34 degrees. The thing is though, my buddy's engine with 10.0:1 CR could only be advanced a couple more degrees at most than mine at 10.6:1 CR. So say for every half point of CR, two degrees. But that would put the EJ22E 9.5:1 CR at around 30, with 91 octane. That's a ways off from the EJ22E stock map at 34 degrees, running 87 octane. There's something more going on here.

As long as the cam timing is the same between the EJ22E and the frankenmotor, I would think that the cam would have a similar effect. In any case it's interesting here to note that both dynos were with frankenmotors, one with stock cam and one with the 220. On the stock EJ22E cam frankenmotor, it's smooth at idle and feels like it has similar power to the EJ22E longblock; it just has a bit more all over, as would be expected. Now the cam timing is something I've been thinking about. So even with the stock cam, a frankenmotor needs the timing pulled back pretty much all over or else there is knock - even with 91 R+M/2 octane. Is this necessary because of the increased compression ratio? Longer stroke? Or is there something more insidious going on, like the deck height being different causing different cam timing? This doesn't seem to be the case as the timing belt tensioner seems to be in the same spot. Maybe the keyway in the crank is at a different position? For the frankenmotor, I think ideally there would be a cam that the exhaust valve opened sooner because of the increased piston velocity. But the stroke isn't all that different on the 2.5 (79mm) vs 2.2 (75mm), so I'm not sure how much torque would benefit from a cam designed specifically for the frankenmotor.

Yes it does look like it's the opposite of a "torque" grind; the other pulls look similar, so I don't think it's an artifact of measurement. The lope at idle also is a tattletale that there is less performance at low RPMs.

Different brand dynos are different enough that it's hard to compare between them. I haven't found a Dynojet plot for a stock EJ25. The Delta 220 cams are "torque" cams. Really though it looks like these mostly give a better top end. There is a lope at idle too, but I've tuned around most of that. Still haven't received my core charge from Delta, after numerous emails. Thinking of telling the credit card company.

Is that a super high compression motor on race fuel with 8,000 RPM redline? Your numbers are the same power as a WRX EJ205, as shown below. Increasing compression ratio gets maybe 5%. The cams get another 8%. The EJ22E intake and small heads then brings that back down a bit, but keeps good torque down low.

Here's my dyno results from my EJ25D/EJ22E frankenmotor w/Delta 220 cams. And here's my buddy's EJ251/EJ22E frankenmotor, stock cams, and a half-point lower compression ratio. You can see I have just a bit more torque across the board (maybe because of the slightly higher compression) but the real gain is on the top end at 6k RPM where my torque is better.

Yes it's always fun to figure out the PO's wiring mess! I did document the change in my car's maintenance log, so if there ever is someone after me maybe they will have a chance of figuring out what I did.

I had a leak where the water would drip right through the fresh air vent under the windshield. The source of the problem was the plastic cowling below the windshield was not sealed anymore to the body at the end. It would especially be worse when parked slightly uphill.

There's a little bit. The strength is from the bolt clamping down. I've gone back and filled in the old holes with the welder, and then ground them flat. Can't even tell they were there. There were no signs of slipping before that though.

The headlight relay that I connected to is a load reduction relay. The relay coil is connected to the black wire going to the ignition switch. With the ignition switch off the relay opens up and the circuit goes dead. See below, the black/yellow wire at B15 (middle right). https://picasaweb.google.com/lh/photo/QFK0sj1MpAXMwQH6Oohf7-t4bszuSWzHLxcb_sc0RE0?feat=directlink

And... It was as easy at that. With a sharp side-cutters I was able to cut the bus bar between fuse #11 and #12. Then I soldered a short wire from the large black/yellow wire on fuse #8 to the now free fuse #12. No more dimming backlight with the blinker. Based on my highly scientific WAG (wild-rump roast-guess) I'd say moving the current to the other circuit will be fine. It's no more than 15A (the fuse #12 rating) and the I used the headlight relay with only one 15A fuse on it (the other one has two 15A fuses). The red fusible link is good for 50A total, and the headlights would draw only 11A, the other lights max 15A, so grand total is less than 41A.

Okay I dug down into the schematic. Fuse #10 goes to the blinker. Fuse #12 goes to the instruments. They share a common supply, the black wire from the ignition switch. Bypassing one or the other to the battery eliminates the dimming. That also means it's not a ground. So that means something has some resistance in the circuit back to the battery. There's no voltage drop at the fusible link (the 1.25mm link to the white wire). From there to the ign switch has about a 0.4V drop. The ign switch itself is 0.15V. The black wire from the ign switch to the fuse panel is about 0.15V. So in total the drop from the fusible link is around 0.7V when the blinker lights are on, and about 0.4V with the blinker off. Most of that drop is in the white wire. I don't think the ign switch is bad with a 0.15V drop, seems ok but probably would be better with a new one; even still that wouldn't fix all the voltage drop and the dimming. I suspect it's not just my car, but that's how they all are due to the drop in the wiring. Yeah I'm kinda picky, these little things bother me. So how to fix it? It seems to me the easiest way would be to tie the instrument cluster power into one of the relays going to a headlight. This would power the instrument cluster through the black/white wire and not the white wire, but still switch off/on with the ignition switch. I'm not sure how to break the connection inside the fuse box to isolate fuse #12. Basically I want to connect the supply from fuse #8 to fuse #12.

I'd bet if he left the outer casing in, the same bushing I used here would fit: http://www.ultimatesubaru.org/forum/showpost.php?p=1108054&postcount=24

Thanks for the comments so far! So I should mention that it's just the dash illumination (and not the cluster because I have digidash) that is dimming. The headlights don't dim at all. I have an EJ engine/alternator with ~50k miles so that should be fine. The battery is pretty new, no corrosion. I did check the fusable links a few months ago and they were all tight with no corrosion. Grounds are new as of a few months ago also. I think I need to dig into the schematics to see what circuits are shared between these things. So maybe if I ask a different way, does anyone have the trip computer that does NOT see backlight dimming (headlights on of course) with the blinker?

I have the trip computer in my GL-10, and it's always done something that's a little annoying. The clock/display backlight will dim in sync with the blinkers. There must be some shared circuit somewhere, and something is dropping some voltage. Maybe a bad connection, relay, or something. Anyone else notice this? Anyone try to fix this?

Yes, the one inside the tank fails, I had to add another valve as well.

I'll go with Ivan here, yeah did you check the oil pump? My EJ25D had a few screws that had backed out. Also maybe you have a cracked pickup tube. I assume you're not low on oil, maybe your dipstick is reading wrong. When I installed my Delta 220 cams I didn't have any HLA shims and after a few minutes when breaking in the cam everything was quiet and still is today.