blitz

Phil, a couple things. First, since the return (ground) connection is made by physical contact with the crankcase, you'll want to make sure that the contact patch is clean and free of corrosion, and that the mounting bolt is tight. Second it's possible that one of the several connectors along the electrical path between the sensor going back to the ECU may have an intermittant open contact. Last you may have an abrasion situation where the lead has rubbed through it's insulation, and is intermittantly grounding out at some point. I believe a short would code as "sensor low" and an open would code as "sensor high". The sensors may have all been fine.

Keith, is it the spin-on filter (mounted on the side)?

A couple months ago, I went completely around my Sube and changed all the lamps. I used standard parts in all the positions except for the headlights. Instead of the Sylvania lamps that were installed OE, I opted for a slightly higher output Sylvania lamp (same part #, but with an extra 2-letter suffix at the end). They're not the obnoxious blue-white blinders, but a regular halogen with an extra 5 watts or something like that. I wish I could remember the details. I really didn't notice an increase of light with the new lamps, but I'd probably notice a decrease if I put the old ones back in.

Keith, if I'm not mistaken, didn't the EJ-22 make a change from DOHC to SOHC at some point? The earlier version being set up more as a high-rpm turbo screamer (Turbo legacy, Impreza 22B, etc.) but also available in N/A form, then in the later version converted into a dedicated N/A motor with a punchier midrange. I'm just guessing.

Well, it could be this, that, and the other. It's impossible to diagnose an unfamiliar vehicle remotely, but if filling up with a fresh tank of gas fixes the problem, then assume it was bad gas. The big secret to problem solving is actually allowing for verifiable problem resolution as, opposed to jumping to premature conclusions or any involment re: wishful thinking. Always make one change at a time, then measure the result.

The answer you're probably fishing for is the head gasket issue, which is actually two problems: The first part involves combustion gasses getting into the coolant, eventually causing an overheating and warpage issue if not tended to promptly. The second part involves total dissavowment by Subaru. "Problem, what problem?"

Well... the ones with mass airflow type do. My '02 OBS employes an alternate MAP-type system which uses manifold pressure, atmospheric pressure, air temperature, throttle angle, and vehicle speed to guess at how how much air is being injested without actually measuring it. The cleaner-running '05's have gone back to the mass airflow type, which is clearly the superior system whenever tight control over fuel delivery is a requirement.

Myles, every Subaru OEM filter I've looked at HAS an ADBV (there may be some that don't, but I haven't seen them). The reason for the ADBV is not to hold oil in the filter, but to keep the oil from siphoning down out of the engine galleries backwards through the filter, pump, and pickup tube back into the pan.

OK, one more well-seasoned quote: "The hesitation experienced in many Subarus during low speed acceleration is due to the engine management computer's overreaction to noise that is sensed in the knock detector, which is essentially a microphone on the engine. The knock sensor is a piezoelectric microphone that is situated on the top of the engine block, approximately seven inches below the bare throttle cable near the back of the intake manifold. It is black in color and shaped like a stack of eight, 3/8 inch size, flat washers. That is to say that it is a doughnut shaped piece, the diameter of a quarter. It is bolted to the engine with one, vertically situated, eight millimeter bolt that is about 40 millimeters long which accepts a 12 mm socket wrench. It has one wire that plugs into a wiring harness just below the throttle cable. To save the computer from sensing ambient engine sounds that may be mistaken for engine knocking, the sensitivity of the microphone needs to be reduced. The easiest way to accomplish this is to insert a layer of vibration dampening material between the knock sensor and the engine block. This may result in the need to replace the 40 mm bolt with one 50 or 60 mm in length. The prototype that has cured the hesitation problem is a teflon spacer beneath the sensor with a second layer of urethane beneath that. The final solution may be a less sensitive knock sensor or an adjustment to the engine management computer."

HESITATION ON ACCELERATION - 11-53-98 (2/99) Applicability: 97-98 Legacy, Impreza, and Forester Manual Transmission Vehicles with 2.5l and 2.2l engines. In the event you encounter a customer complaint of a slight engine hesitation between 1500-2500 rpm's when the engine is cold or hot, perform the following Eliminate all external influences, such as an incorrect or dirty air filter, loose or tracked intake duct, dirty fuel filter, low fuel pressure, PVC system, or low engine vacuum that would indicate an external leak or an internal engine component. If all external components are confirmed to be operating within vehicle specifications, the hesitation may be caused by the ignition control logic in the ECM. Under certain low rpm driving patterns, the ignition control system can pick up engine vibrations through the knock sensor and may retard the ignition timing. This ignition timing is learned by the ECM and placed in memory. NOTE: This area of memory can not be viewed by using the Select Monitor. When the vehicle is driven uinder these conditions, the timing may be retarded and could cause the engine to hesitate on acceleration. To confirm this condition, road test the vehicle while viewing the Knock Sensor Signal on the Select Monitor. If you duplicate the hesitation, and the reading on the monitor is around -10 degrees, you will need to change the ECM to correct the concern. [my note: this is an enhanced ECM, not a replacement with the same component]

Only a guess, but the symptom as you describe it, has the hallmark classic knock-sensor false-triggering thing going on. Subaru's flintstone-era knock sensor has a tendency to pick up mechanical resonances from the crankcase (to which it's mounted) and mistakenly interpret it as spark knock. The ECU then rapidy removes about 15 degrees of spark lead (which is allowed to return to normal in the absence of further triggering) causing the sudden "air pocket" in the acceleration. If you'd changed to a different oil recently, that could be the difference in resonances which the knock sensor is picking up.

I always ask the questions that don't get answered (mostly because my brain works just outside of normalcy), but I digress... Perusing through the filters at Pepe Boyz today, intent on picking up a few P14460 (no ADBV ), I started opening boxes lookin for "visual" matches, similarities, etc... Anyway I found the P22821. All the mechanicals match. Same thread diameter and pitch, same number & size of inlet holes, same gasket, same convex mating surface (outward facing threaded portion), and same physical diameter of external filter body. Different however is: 1. Slightly greater overall filter length (more media surface area). 2. HAS AN ADBV!! (like the Subaru OEM part). 3. Costs only $2.99 US ($2.49 for the PROline house-branded equivalent). I'm thinking of giving his filter a try in spite of the fact that I have no idea what I'm doing. The only variable that comes to mind which was not already mentioned would be the filter's internal bypass valve (specific pressure). Anybody that has knowledge of oil filters have any sage advice for me? Is the bypass valve calibrated for (integrated into) the filter itself, or for the engine?

Does the acceleration delay happen at full throttle, part throttle, or both? Did the problem suddenly get worse, or did it develop in a gradual fashion? Are there any driving techniques you've found that minimize the flat spot? Also, is there anything else that changed about the performance of the vehicle? Noise? Vibration? Smells? Fuel mileage, etc? Any CEL codes?

Yeah, increased elevation and increased temperatures both result in "thinner" air and therefore reduced power. The ECU compensates in both instances by reducing fuel accordingly. One difference however is that increased elevation causes the ECU to advance timing, wheres increased temperature causes the ECU to retard timing. Driving style/highway speed and tire inflation will affect your fuel ecomony more than air tenperature will, except in very cold weather where the lubes get thick.

I also used Dawn on my Colt for 11 years and never had a problem with it, yet now all I hear is how dish detergent is really dangerous for the finish. It seems like the propaganda started right about the same time that all the specialized car wash products began showing up on the shelves. I don't know if all dish detergents are created equal.

Your comparison between the K&N and an open intake puts forth my point in a way I was reluctant to do, thanks.

I've tried both the Amsoil oiled-foam element and the aftermarket Purolator dry-paper element, and neither seemed to fit the airbox satisfactorily, so I've gone back to the OEM oiled-paper which honestly seems to be a high quality piece. Up till now I've avoided the oiled-gauze types on account of being able to see a frightening amount of daylight through the element. Not scientific I realize, but it gives me peace of mind.

Well, it looks like I won't be completely dumping IE. :^) Thanks for walking me through that. ------------------------------------------------- Hey Joe, thanks for the thread. I was on vacation and missed that topic.

I've heard of opera, is it free-ware?

Mmm..., doubly interesting. I first posted the url as: "http://axispowerracing.net/engine.htm", but when the link didn't work for me, I immediately edited it down to: "http:// axispowerracing.net", but the linkey still no workey. The edited url appears correct as viewed on my 'puter, but you're saying that it still shows up with the "/engine.htm" on your machine? ROFLMAOL!!! I rest my case about Firefox, it's messing with my ability to do an actual edit. Another thing is it's not allowing me to do is cut-and-paste quotes and url's from outside sources into a post. I was forced to hand-type the url. I wonder if it's an issue with the way certain Windows control settings may be configured on my machine. That may also explain the variable results between different users.

I switched to Mozilla Firefox v0.9.2 about three weeks back and I'm getting mixed results. I appreciate the increased security against viruii (the reason I switched from IE), also the pop-up blocker works flawlessly and transparently. The bad news is Mozilla seems to run really slow, almost like machine language emulation or something. Also a lot of little things don't work correctly, e.g. I can't get smilies to show up in message posts, the last url I posted (axis racing) doesn't seem to wanna work, etc. Anyone else using Mozilla-ware with similar or different results? I'm debating whether to go back to IE.

That was me. I just did a little math and discovered I was mistaken. EJ-25 Bore: 99.5mm Stroke: 79mm EJ-22 Bore: 97mm Stroke: 75mm The 97mm bore combined with the 79mm stroke yields 2.335 liters. Axis racing builds an awsome 2.4 that's based on a 2.2 turbo block using a 2.5 crank, but apparently there's some sleeving and proprietary whatnot going on to arrive at the 2.4 liter displacement. http://axispowerracing.net Apologies for the confusion.

THAWA, that's interesting, not unlike the "dead-zones" on the temp & oil pressure guages. It's loud commentary about human nature, but troubling. If one of the options available on the build-sheet for a new vehicle was "true gauges", would anyone opt for it? The public loves good news, albeit B.S.

It just seems wierd to me that a company who'd earned a reputation with engines like the EA-81 (so legendarily good that the experimental aircraft industry probably wouldn't exist today without it), would do something so foolish as to actually try to calculate what they could barely slip by with (and REALLY get bit in the rump roast by it), when traditionally their design philosophy was more like: "Calculate what's necessary, then add more". At one time GM was the envy of the world too. Maybe it's just a coincidence that uh... mmm...

The sad thing is that for all the trouble that the over-bore has caused Subaru (owners), it's a miniscule displacement gain. Most of the displacement increase over the 2.2 is from the stroke increase. If they would've left it at just the stroking, they would've ended up with a rock-solid & honest 2.4 liters. It must've been greed and wishful thinking that tempted Subaru to hog the bore to get an extra lousy 57cc (or therebouts) and call it a "2.5". It was a deal with the devil and they (you & I) lost. I'm assuming that the 2.5 turbo must a semi-closed deck to help stabilize the thin bores?