93 Legacy Misfiring??????

[Guest 93leg]

My 93 Legacy has a problem shortly after start up and a few miles down the road. I have read other very similar problems, and am wondering about the Knock sensor. The car starts missing around 2 miles from home. We travel up a hill to a stop sign on the other side when it really becomes noticable. The three cylinders it runs on is barely enough to keep it running. Sometimes it stops, but even if I would turn it off, and restat it, it would run fine. I know that the O2 sensor does not function before it's hot. The ECM has default information it uses before it relys on the O2 relayed information. What else, besides the knock sensor can I look for?

 

I have replace plugs, wires, fuel and air filters, and the ECM coolant temp sensor.

 

Thanks

[Guest intrigueing]

Did you check to see if you have any codes?

 

See: www.surrealmirage.com/subaru/subaru.html

 

for instructions on pulling the codes.

 

Could be several different sensors, Coolant temperature, cam, crank, knock. See if a code can help you pinpoint the problem.

[Guest 93leg]

I have not seen any indicated by the Check Engine light. But I have not tried to pull any.

[Guest intrigueing]

Some times codes are stored but no CEL is shown on the dash. See what you find, and let us know what happens.

[Guest AusLegacy]

Looks similar to my (and some other peoples) problem, see (search) stalling on incline thread.

[Guest meep424]

I had a similar problem. Replaced wires, plugs, to no avail. ECM reported misfire. Turned out to be the t-belt, tho it was only 40k miles old.

 

Just for kicks, tho, I'd run some fuel cleaner through and install a new filter first.

 

mike

[Guest scoobtech]

Is this isolated to one particular cylinder ?? You may want to inspect the ignition coil contacts . You know where the wire plugs up . These sometimes get corrossion . The power transistor may also be wigging out .

[Guest 93leg]

Using the Haynes manual I pulled the codes from the OBD. 14, 15, 16, 17, and 32. The Teen codes refer to the each of the 4 injectors, and 32 was an O2 sensor. I have checked the operation of the O2. The backprobing results fell right in line with the stated manual values. The knock sensor was also checked for continuity. It passed this test, but that doesn't tell me that it is working. The injectors I have not checked yet, but the codes had to do with some type of abnormal, or low return voltage. (Can't remember wording) Since all four have shown up, I am hoping that there is something else that has cause all of them to show this error. One interesting thing, the Haynes manual says to unplug the OBD connector to access the codes, but I had to plug it together to access the codes. I have now cleared them and will wait to see if others show. They could have been stored for a long time.

I have checked the coil contacts and their resistance, which doesn't tell me if they are truly functioning. Coils can breakdown under load.

What next????

[Guest Legacy777]

Next step is to just drive the car, wait until the problem happens again, and then check the codes. See what pops up.

 

Injectors are simply electro-mechanical devices, so if they show up again, and they check out ok, I'd check the wiring harness and/or the ECU.

[Guest 93leg]

I had a dealer look at it. They test drove the car and from that experience said it was a sticking valve, and would be happy to repair it for a small fee of $1,200. I have a hard time excepting that, because it goes away if you restart the car. They didn't even check for codes.

 

How reliable are the heads and valves on a 93 with 130,000 miles??

[Guest Legacy777]

I've got 170,000 miles on my 90

 

You may want to run a motor flush or something similar to eliminate the sticking valve theory.....most people have had good results with autoRX.

 

Me personally.....I don't think it's valves. Sticking valves like you mentioned aren't going go away when you turn car off and on again.

[Guest 93leg]

Does anyone have access to TSB 114695? I think it has to do with Knock Sensors. I have seen where Knock Sensors are listed as a "Problem Area" for 93's. I think it was on an Advance Auto Parts, or Autozone web site.

[Guest Legacy777]

There is an updated knock sensor for the first gen legacies.

 

They're like 35 bucks

 

Part number is 22060AA031

 

If you get it from liberty subaru www.newsubaru.com and mention you saw their ad in road & track, you'll get an additional discount.

 

I don't know if this will solve any of your problems though.....but it probably wouldn't be a bad idea to replace.

[Guest northguy]

I'm by no means much of a mechanic, but we change the timing belts religiously at 60 k as we've been warned about the potential for problems such as this. How many miles do you have on it?

[Guest 93leg]

The timing belt was replaced at 100,000 miles. It now has 130,000 miles. I ran the car in the OBD "D" check mode last night and found the Check Engine flashing code 32, O2 sensor. I did however run a voltage check on the O2 sensor and found it to be within the specified operating range, although the heater resistance of the O2 sensor was 8 ohms, the manual calls for 30 ohms. I may replace that for $100. Not sure yet!

[Guest PaganQWA]

From my experience with similar issues it was either spark plugs, plug wires or fuel injectors. You replaced plugs so lets rule out injectors. Take your digital multimeter and do an OHMs resistance test on them all. Try and get OEM specs to check against. If you cant get them just test them all and compare the numbers to eachother. They should all be similar numbers and if one or more is significantly higher its going out. Mine didnt even throw codes when they were going out

 

Hope this helps and good luck.

[Guest camosuba]

Subaru here in Oz, have a pressure pack spray can for the removal of carbon build ups, you spray about half into the manifold leave for 15min then start and spray the rest into the manifold while its running, some guys call it a tune up in a can, not sure if it cleans valves as well it usually creates a cloud of smoke

 

Regards Camo

[Guest 93leg]

I plan on testing the injectors and anything else I can before replacing anything. Do autoparts stores carry OBDI meters that measure sensor signals in live time while the engine is running? Or do they just read stored codes? It would certainly be useful, and maybe help pay for itself by stopping me from buying unnecessary parts.

[Guest PaganQWA]

you need no special tool to pull codes from first gen legacys. please find a post by Legacy777 and go to his homepage to find proper instructions on how to pull and read your codes. Its basicly a Subaru morse code

 

As for testing all you need is a cheap $10 digital multimeter found at Wal-Mart to do an OHMs resistance test.

[Guest 93leg]

Over the weekend I checked the resistance of all four injectors. Their values were all close to each other and within range. I also looked for the proper operation of the charcoal canister purge solenoid. Battery voltage was present with the ignition on, the valve would physically operate using an external voltage source, and I could feel a vacuum at higher RPMs by disconnecting the line at the canister. This valve is supposed to stay open at long as the key is turned on. Correct?

[Guest Legacy777]

The valve is only supposed to be open if rpms are above a certain point.

 

At idle and below this set point rpm, the valve should be closed.

[Guest 93leg]

Thanks Legacy777, that makes sense. You wouldn't want the valve to be open at idle. It would most likely affect the idle quality if it were left open. Is there anything that can be cleaned with the MAF sensor? The connectors I have cleaned, but what about internals?

[Guest Legacy777]

honestly you're better off leaving the MAF alone....I've heard some people said they cleaned theirs, by spraying brake cleaner or something in there.

[Guest 93leg]

The knock sensor is to retard the timing if a knock is detected. Is there a default timing advance limit stored in the ECU? I am thinking of unplugging the knock sensor, but I wouldn't want the timing to advance through the roof and cause damage.

 

I do have an annoying vibration coming from what I would call some type of bracket arm, like a radio bracket, or AC duct holder bracket. (I don't know if there is one, but this is what it sounds like) It is coming from behind the radio, right around the AC vents. I'm wondering if the knock senses actual vibrations and this is what it is mistaking for a knock? Just thinking out loud. I may try to fix it.

[Guest meep424]

this does not sound like knock sensor to me. It would not retard the timing to the point where the engine would begin to miss. It would just seem to have not as much power.

 

Is the misfire violent, or merrily the absence of ignition? Seeming as it is temperature related, a violent misfire could be the result of preignition or detonation, both of which can result from too lean a mixture -- pointing me towards injector problem. Absence of ignition, especially if it was only during acceleration or power at low RPMs might suggest lack of spark, caused by faulty plugs, wires, coil pack or igniter (amplifier that powers the coil pack).

 

Are you certain it is one cylinder, or could be alternating between two? The rhythm will be more regular if it is just one, which suggests to me fuel delivery. If between two, spark, since the plugs fire in pairs.

 

So, we have mixture, which could be a number of things, most simple of which is O2 sensor. Unhappy fuel injector, possibly clogged, but unlikely since this would show up more when cold. Maybe wiring harness for that injector, but there will be a code for that. Or, ignition. Depends on the symptoms?

 

Meep