~Slideways~

The clock wasn't working, I got another from Pick a Part. It also wasn't working. Took it apart and found cracked solders on these: Added a bit of new solder and all working again. Also fitted a few bulbs to the gauge cluster, it's still a bit dim so I might see if there are any bright LED options.

Another thing I've noticed is the front grill surround is quite different and looks much better than the samples below where it is more flat. While on mine is extends out much more:

I'm not sure what they were going for with the round lights, maybe copying Toyota? Same car: Re rpm's the US V7 WRX's are usually 3.9:1 and 2400rpm @ 60mph so I reckon it'd be quite different to drive.

I've been driving this despite the cam sensor error, the engine check light for the 2 cam sensor codes are: P0340 and P0365 Interestingly the check engine light only stays on at low rpm, and there is an exact rpm where it turns on or off, but it changes that point each time (may be heat related), while always being under 2000rpm. So it seems fine under most driving conditions. For example as soon as you start the engine, it has the check engine light and it turns off at around 1800rpm then back on as soon as it drops below that. So while driving it is almost always off until you are idling or even between gearshifts if you let the rpm's drop that low. There are 2 cam sensors which are located on the back of each head at the top and another sensor on the passenger side behind the top cam gear. I have tried: - cleaning all three sensors. - testing ohms readings: the 2 main ones are around 1950ohms (identical) and I can't remember what the third one was (the cam gear one). But all within spec. - replaced split wiring and connector on the drivers side one (I thought this solved it originally but it came back). - passenger side wiring looked fine. - cleaned all loom plugs and pins. - the alternator is noisey but still good voltage. Tried removing the belt to just run off battery but no change to the check engine light. - one internet theory was bad grounding on this gen wrx. So I added three thick ground cables, but made no difference. Some possibly reasons according to google: - Alternator amp's could explain it going away with slightly higher rpms (but removing drive belt kind of defeats this theory). - one guy with a oscilloscope found noise being generated so directly soldered extra grounds to the ecu pcb itself. I could just buy 3 new sensors, at around $100 each. But heaps of reports that this didn't solve the problem for other people and it seems weirdly unsolvable? I decided to take it for a reasonably spirited drive through Ohariu Valley > Takarau Gorge > Karori, it goes very well for basically stock. I don't know what the stock heavy flywheel would have been like but this light one makes it really fun to rev match and quick down shifts etc. Absolutely no problem in heavy traffic like some internet people suggested. Some things I'd like to change soon: - Remove old alarm system (20-25 years old, randomly going off and only one remote) and probably install something else. - Seems pretty understeery, so want to try change the rear sway bar (only 17mm on wagon, but 20mm on some sedans of same gen) - Most speakers are buggered. - Tyres that came on the BBS wheels are almost full tread but are definitely not performance tyres, ok in the dry but first wet roundabout felt like it just wanted to understeer. Just managed to catch the Sun going down: I put in it's first full tank of 98, will be interesting to see what the fuel consumption is like. Maybe I'm kidding myself but with the brand new o2 sensor, maybe it'll be decent?

Success, I found a couple of 2nd hand genuine idle control valves but they didn't really make any difference. So since I had a spare I completely disassembled it and found that the shaft was stuck even after being soaked in various things. I got it moving nicely and reassembled, it is adjustable with oval holes on the motor/coil part, like some TPS's so I got it as close as I could. Now it idles and has no code for the IACV. Win! So now the oxygen sensor and iacv codes are solved. I've been driving it a bit and there is still a cam sensor code coming flashing the check engine. It runs totally fine and keeps getting better, from what I under stand the ECU gradually increases the ign timing and boost control based on knock and o2 sensor (after the ecu is reset). It's now up to 14.5psi exactly. I swapped in an old Gizzmo boost controller with a digital display just to use as an accurate boost reading. Threw the old boost gauge in the bin. One of these: It's definitely well geared for fun driving, probably not so much economy. Pretty certain it's the 4.4 final drive version based on the code on the gearbox and chassis plate. It also does just under 3k rpm at 100kph. So this cam sensor check engine flash, I'll have a look at the other sensors but for now it seems to be running really well.

Have since discovered that the idle control valve is dead, resistance test shows OL between pin 2-3. 1-2 is too high as well. There are 'cheap' aftermarket IACV's available but most say they are crap so trying to find a good 2nd hand one. This style IACV was only used on the 2002-2005 turbo wrx's and a few early 2000's turbo models Foresters and Legacy's. Nothing at Pick a Part Wellington, all N/A engines there except one which is the wrong kind. I've checked all PaP listings around the country and nothing. Contacted some local wreckers and no replies, most likely they'd only keep good engines and won't want to separate. Where did all of these go?! Found a guy with one but have not heard back for a few days. Frustrating since it's so close to being ready to drive!

Cool thanks I think RHDJapan was one I checked out, there was a place in the US called Flatironstuning and they had them listed at a good price, until it came to shipping being $500ish USD. Insane.

Gah bugger it, Subaru has no stock on that IACV gasket and is a month wait from Japan. So I might as well order it from Amayama again for 1/5th the price. In other news the o2 sensor arrived yesterday so will get that installed as soon as I have time, then see how it runs.

Hey, bump for a cool project (I know its gone now but still). Just wondering where you sourced the above?

Its possible the idle control is just gunked up, so I'll get a o'ring gasket (funny shape that never goes back in) from Subaru then take it off and see what it looks like inside.

New plug arrived and matches: Properly crimped new pins onto some new wire to replace the crunchy section: Installed with conduit and heat shield tape since its near the turbo: Started it up and checked codes, no longer getting a Cam Sensor code, win! Still has the idle and o2 sensor, code though. I was hoping idle code could be linked to bad cam sensor signal. Test drove it and no longer rev cutting at 5k but still seems to be limiting boost and quite sluggish, so is it still limp mode without a rev cut? I have ordered a Denso o2 sensor from Rockauto, should arrive in a week. To make it worth it for shipping i bought a bunch of oil filters for a few cars. I also tried a few more things that the internet told me might make the Tactrix cable read the ecu. Including jumping 12v to the 'flash block' plug. But made no difference. So that means at least for now in order to drive it I need to have these 'dummy' TGV's wired in. So I used some conduit and extra plugs in order to run the cables neatly through the engine bay then through the fire wall. Now I'll have the dummy TGV's under the driver's seat. Its actually surprisingly neat for a temporary workaround. Here it is in the driver's footwell, and the wires are long enough to tuck under the seat. Weird I know, but it works for now. Engine bay, cant tell there is extra wiring anymore.

I don't keep track other than thinking, can I justify this vs the value of the car and/or how much I like the car and how long I want to keep it. But I also spread project car costs over a long period and don't pay labour by doing stuff myself like most people here, so I can justify (there goes that word again) the big ticket items by "saving" on labour. I have never tracked fuel other than the occasional calculation of roughly how many km's I do in a week, but only when fuel prices were almost $4 a litre. At that point I thought, maybe a cheap leaf makes sense for the work commute, but then what's the point in having interesting old cars if I don't use them?

Got the drivers side cam position sensor out, was pretty dirty but don't think that was an issue. Ohms reading is within spec: The wiring looks to be the likely culprit though. 1. The wires are both cracked at this plug: It still has continuity through to the cam sensor connector though, but from what I understand continuity doesn't necessarily mean it's ok. Next problem is the sensor connector itself is brittle and the retaining clip inside has snapped off and the wire easily slides out. Whether this wasn't touching the pin before or it happen when taking it out this time I don't know. Wire pulled out: You can see the broken inside clip that is meant to retain the pin on the top hole here: So I've cleaned up the sensor, reinstalled and searched for replacement connectors. Turns out it is a standard Nippon Denso plug mostly used for injectors, but it has to have the low keyway. MSEL had some so have ordered 2, the other side is probably ok since it isn't near the turbo heat. So hopefully this solves the limp mode.

Yeah thats pretty much the case it seems, they work for everything except early 2000's subaru 16bit ecu's (yay for me lol). Lots of Google results with people saying the same, even an HP Academy article saying 16bit subaru ecus are difficult. Unfortunately I cant just swap to a 32bit ecu.

Good question, but I reckon probably pretty simple no canbus or anything in these. Looks like there is a plug in Link ecu too. No money for that with this car though. Should be good on factory stuff though, just need to sort these last few issues.