Is this a common problem..?

NZHall · Post by **NZHall** » Thu Jun 19, 2003 11:00 pm

Hi all. I'm hoping someone can help me. I live in New Zealand and I have just bought an 1990 Legacy 5 speed wagon with normally aspirated EJ20 4 cam motor that's done 143,000 kms or 90,000 miles. Very happy with this car, other than a moderate but noticable engine hesitation/miss under heavy load, most obvious between 2000 and 3200 RPM as a feeling of engine "surging" and not making full, smooth power. Car idles fine and runs beautifully on light load such as highway cruising speeds or puttering around town. I have been told by the local Subaru dealership this can be caused by high mileage coil packs developing pinholes and then misfiring so I replaced all four with genuine coil packs ($$.. ..ouch), to no avail. I also have replaced spark plugs with both standard and then platinum plugs, as well as replaced both the fuel and air filter (genuine again). Again, no improvement. I drove the car out into the country last night and checked under the bonnet with all the lights out and couldn't see any of the wires leading to the coil packs "arcing out". The Check Engine Light has NOT come on, though it is working (comes on just before cranking the motor). I am truly stumped!! Does anyone out there have any experience with this kind of problem? Electrical fault in the wiring loom or alternator? Faulty O2 meter or some other sensor? Glitch in the engine management computer? restrictive exhaust? (muffler has been replaced with cheaper aftermarket unit prior to me purchasing car). Am reluctant to keep throwing money at the symptoms without a better understanding of the problem. I appreciate any thoughts or advice you may have. Thanks and best wishes, Iain

SubaruCO · Post by **SubaruCO** » Thu Jun 19, 2003 11:11 pm

Even though most of us here don't have the EJ20 NA motor, I would suspect that your issues might be similar to the ones detailed through in the sticky topics at the top of the engine forum. I would predict that most of the other people here would tell you to try most of the ideas posted in those sticky topics. I would suggest doing an ECU error read to diagnose any glaring issues first. If there are any particular issues relating to US vs JDM-like models (which I think you have) there are a couple of Australian members that might be able to help on that front.

NZHall · Post by **NZHall** » Fri Jun 20, 2003 12:40 am

Thanks for the advice. I should have mentioned it is a 1990 Japanese Domestic Market Legacy, and it was imported into New Zealand in 1996 so it may have a different ignition system to US models. Sorry about the newbie questions, but where is the ECU located and how do you perform a "ECU error read"?

Post by **Legacy777** » Fri Jun 20, 2003 1:51 am

check out my site for code pulling instructions. There should be similar green & black connectors.

The hesitation thing is an issue on the us cars too.......wish we knew the cause.

NZHall · Post by **NZHall** » Fri Jun 20, 2003 2:55 am

Whoa...I just read back through some of the older forum discussions, (probably should have done that first, I guess) and it sounds like this hesitation thing was a Subaru factory option!!! This weekend I will check if I have the original knock sensor and check the error codes too. Is the best way to re-set the ECU to take the battery earth off or to do you prefer the under dash plug method? Thanks for the great advice, this gives me some hope of sorting the problem.

Post by **Legacy777** » Fri Jun 20, 2003 1:26 pm

The two plug method I'm not sure it actually "resets" the ECU. On our cars you can pull fuse #16 which is the EGI/TCU fuse for 10 minutes and that will clear both the ECU & TCU.

entirelyturbo · Post by **entirelyturbo** » Fri Jun 20, 2003 4:50 pm

Yeah, your EJ20 has the individual coils on each plug, that's why it was so expensive

. Us Americans have a coilpack that mounts on top of our intake manifold.

As for the hesitation thing, yes mine has come back a fourth time. However, boys and girls, I have another idea. We've discussed the O2 sensors around here a bunch, and we all know they're not immensely useful to the car. But I have noticed that when my car is first started in the morning (open-loop) that it has much more power, runs much smoother, and has NO HESITATION. (I think I've actually been stupid enough to call out "Hesitation Fixed!" those other times when I was in open-loop

). Of course that means when the loop closes after it warms up, it runs a little rougher, less power and hesitates. So I think I'm going to try replacing my O2 sensor. Since my gas mileage has been rather appalling these days, I theorize that when the ECU closes the loop, it takes incorrect readings from a faulty O2 sensor and sets the a/f ratio very rich.

So gimme some time to save up $100 for a OEM sensor

, and I will report back.

boostjunkie · Post by **boostjunkie** » Fri Jun 20, 2003 7:32 pm

But if you still get hesitation on anything from 1/4-WOT (depending on the gear and how hard you're stressing the engine) the ecu will automatically ignore the o2 sensor by going into open loop.

Did you have anyone test your charging system? Once I replaced my alternator I cured all of my hesitation issues. Some of them went away when I changed my sparkplugs, but the alternator swap completely got rid of it!!

vrg3 · Post by **vrg3** » Fri Jun 20, 2003 9:26 pm

It's not quite true to say that the ECU ignores the oxygen sensor signal when it's in open loop.

The difference between "open loop" and "closed loop" is that in closed loop it tries to maintain a stoichiometric air/fuel ratio. The ECU is always watching what the oxygen sensor is reading and using it to learn. For example, if it's in open loop but it sees the sensor giving a lean reading, it will start to richen the mixture, and also remember to richen it in the future.

A faulty oxygen sensor can affect driveability in many conditions other than cruise or idle.

I was actually really amazed by the flexibility of the learning code in my girlfriend's 1989 Mazda 626's ECU. When the engine backfired through the intake and blew apart the intake resonator, I plugged the hole (with an old Legacy shifter knob of all things). That really screwed up the airflow dynamics. The flapper door airflow meter requires the ECU to do a lot of guessing with regards to airflow, and it started getting it wrong a lot. The engine was badly in need of a tuneup, so the wrong mixture caused misfires and general bad running. It was really hard for her to learn to drive around the problems (it was particularly bad when you rapidly changed the throttle angle or when you had the throttle far open, and there were also just some bad areas, like 2000-3000 rpm in 3rd gear), but she had to get to work and back so she tried her best. Two weeks later, when I was ready to do the tuneup, I test drove her car beforehand and found no noticable driveability problems at all! Even at wide open throttle, rapidly changing throttle, lugging, etc etc etc, it ran smoothly. I reset the ECU and it went back to the terrible running. I couldn't believe how much that ECU learned in two weeks just from a one-wire oxygen sensor.

boostjunkie · Post by **boostjunkie** » Fri Jun 20, 2003 9:57 pm

Is this fact universal for all cars? That makes me wonder how a s-afc can be used on ANY car? If the ecu is constantly monitoring the o2 sensor signal to keep the a/f/ ratio within spec, then wouldn't the ecu learn around the s-afc in all cars? Are you required to reset the ecu when using a s-afc?

vrg3 · Post by **vrg3** » Sat Jun 21, 2003 12:35 am

I can't say for sure that it's universal, but it's one aspect of good engine management design.

As for the S-AFC, remember that the oxygen sensor only gives precise readings in a very narrow range of ratios around the stoichiometric ratio. It basically reads "lean," "a little leaner than stoch," "stoich," "a little richer than stoich," and "rich." Usually the AFC is used to move the mixture around to different values in the "rich" zone anyway, so the lambda sensor feedback wouldn't be very different.

If you try to use an S-AFC to lean mixtures down to near stoichiometric under high load conditions, I'm pretty sure the ECU will learn around the changes. The designers of the ECU know that running stoichiometric at wide open throttle isn't really good.

entirelyturbo · Post by **entirelyturbo** » Sat Jun 21, 2003 12:51 am

boostjunkie wrote:Did you have anyone test your charging system? Once I replaced my alternator I cured all of my hesitation issues. Some of them went away when I changed my sparkplugs, but the alternator swap completely got rid of it!!

Read my post in your stickied thread. I put a new alternator on my car, and found a vast improvement. It did run smoother for a while, but then it started hesitating again. I did an ECU reset for the hell of it just for the sake of experimentation, but to no avail.

I really find it to be a surprising difference between open-loop and closed-loop. It feels like a new car when I first start it up, but once it reaches operating temperature, it still runs okay, but not near as good as it was. I don't know exactly what the ECU does with the loop situation depending on throttle position, but I can say that my car really doesn't feel more powerful at WOT then at 3/4 throttle...

What sucks is that I was hoping to get all caught up with the car this month, having no major driveability problems to worry about so I could concentrate on more important things in life, like girls

...

ciper · Post by **ciper** » Sat Jun 21, 2003 1:47 am

vrg3: You stated that the ECU still watches the Oxygen sensor when at wide open throttle. I have always known that not to be the case. My understanding was static (safe rich) maps using only throttle position engine rpm and air intake.

legacy92ej22t · Post by **legacy92ej22t** » Sat Jun 21, 2003 2:09 am

dump the platinum plugs,ngk v-power is the only way to go on the ej22 probably the same on the ej20.by the way what were the standard plugs?

Post by **Legacy777** » Sat Jun 21, 2003 3:34 am

subyluvr2212 wrote:
boostjunkie wrote:Did you have anyone test your charging system? Once I replaced my alternator I cured all of my hesitation issues. Some of them went away when I changed my sparkplugs, but the alternator swap completely got rid of it!!
Read my post in your stickied thread. I put a new alternator on my car, and found a vast improvement. It did run smoother for a while, but then it started hesitating again. I did an ECU reset for the hell of it just for the sake of experimentation, but to no avail.

I really find it to be a surprising difference between open-loop and closed-loop. It feels like a new car when I first start it up, but once it reaches operating temperature, it still runs okay, but not near as good as it was. I don't know exactly what the ECU does with the loop situation depending on throttle position, but I can say that my car really doesn't feel more powerful at WOT then at 3/4 throttle...

What sucks is that I was hoping to get all caught up with the car this month, having no major driveability problems to worry about so I could concentrate on more important things in life, like girls ...

My car pretty much same thing. Feels really good when cold, and then poop after it warms up.

vrg3 · Post by **vrg3** » Sat Jun 21, 2003 5:46 am

ciper wrote:vrg3: You stated that the ECU still watches the Oxygen sensor when at wide open throttle. I have always known that not to be the case. My understanding was static (safe rich) maps using only throttle position engine rpm and air intake.

I can't say I'm an expert, but I'm pretty sure what I said was true.

The open-loop maps are not exactly static; the oxygen sensor still provides feedback, although it is less helpful.

Neither of us has looked through the code of our ECUs (yet

) so we can't really resolve it, but I do know that my girlfriend's car's ECU did learn to use the proper mixtures at WOT.

entirelyturbo · Post by **entirelyturbo** » Sat Jun 21, 2003 4:57 pm

Okay, Josh, I remember you telling us that you had unplugged your O2 sensor as an experiment on your way to work. Your CEL didn't even light up until the loop closed...

However, did you notice a difference when the sensor was unplugged? Did it seem to run better, smoother, faster, etc. etc.?

I'm not suggesting we all unplug our O2 sensors, but I'm curious as to how much the ECU relies on it for the proper a/f ratio...

If it's a big difference, then I will be more expectant of noticeable results upon installing a new sensor...

Post by **Legacy777** » Sat Jun 21, 2003 9:49 pm

Umm....I don't remember exactly when it happen, but the results weren't exactly what I would call better.

vrg3 · Post by **vrg3** » Mon Jun 23, 2003 3:42 am

Our ECUs in particular have very well defined maps programmed into them. If you unplug the oxygen sensor the engine runs pretty well -- you might not even notice any difference. As a safety measure, it retards the ignition timing a little if the oxygen sensor is missing; that's pretty much it.

legacycentral bbs

A forum for 89-94 BC-BF(BJ) Legacy Owners and Fans

Is this a common problem..?

Is this a common problem..?