emissions trouble w/f.i. EA82

[Dark Solar]

I have a '91 subaru Loyale that failed emissions testing today...for the second time. I love this car; it's been a workhorse that has never let me down until now.

 

Here's the low-down on the problem: first time thru testing idle rpm was 1137/HC level 143. Checked idle air control valve--operating within factory spec. Soooo... put in new plugs and manually turned back the idle speed until it was reading right on the tach. Went back for second test-run...idle speed still high (900+) and HC level 207 (epic FAIL)...

 

I've noticed that there's a little bit of fibrillation in the tach needle at idle which kind of points back to the idle air control valve but again, it's reading as good.

 

I'm stumped. Any suggestions?

[Suba9792]

bad cat? egr? new to subies but hey its worth a shot right?

PCV

VAccum lines

maf

[Dark Solar]

That's what I thought too at first...only trouble is that the emissions test sheet shows pcv is sound and egr is working within spec. Exhaust NOx levels are low so Ox sensor/catalytic team still performing. Currently tracking vac lines as I have time.

 

Somebody get me off this effin' stump!! I don't like heights!!!

[Dark Solar]

Doh!!! Mass air flow sensor...(slaps forehead)...hadn't thought to check there. (slaps forehead again) (and again)

[WoodsWagon]

If the engine coolant temp sensor isn't working right, or the wire leading to it is corroded, you will get a high idle, and rich mixture. It's like a carb leaving the choke on because it doesn't think it's warmed up yet.

 

So that would explain both the high idle and the high HC numbers.

 

Now, by screwing with the throttle stop screw (it's not an idle adjust) you've made a bunch more work for yourself. The throttle position sensor needs to be set in a particular spot for it to idle correctly. You need to get the throttle stop back to the right point. If you counted exacly how many turns you backed it out, try turning it back in. Also look for the corrosion line on the threads and thread it back in until they line up with the top of the bracket. If that doesn't work, then you will have to use a mulitmeter to get the TPS reading right for idle.

 

Is the check engine light on?

 

Here's a good quote with a link to the FSM from another thread:

Did you tamper with the idle stop screw on the throttle body? If you did that may be your problem. You will need to check the idle switch portion of the TPS, to verify/disprove your theory. A faulty coolant sensor or fault in its cuircuit, will cause high idle issues. Ohm it out and check the specs. A stcky IACV (idle air control valve) is another posibility, you might consider cleaning it, regaurdless. I have included GD's link to the factory service manual part you will need. Just incase you dont have it. http://home.comcast.net/~trilinear/SubaruEA82-ServiceManualPart2.zip

Edited April 1, 2010 by WoodsWagon

[ruparts]

I have a '91 subaru Loyale that failed emissions testing today...for the second time. I love this car; it's been a workhorse that has never let me down until now.

 

Here's the low-down on the problem: first time thru testing idle rpm was 1137/HC level 143. Checked idle air control valve--operating within factory spec. Soooo... put in new plugs and manually turned back the idle speed until it was reading right on the tach. Went back for second test-run...idle speed still high (900+) and HC level 207 (epic FAIL)...

 

I've noticed that there's a little bit of fibrillation in the tach needle at idle which kind of points back to the idle air control valve but again, it's reading as good. I'm stumped. Any suggestions?

 

 

You need to put the throttle stop back to its orig position if you moved it, that is factory set to avoid problems with binding and not sticking, then the throttle position switch needs to be set to acheve certain specs for that to work right. the coolent sensor is very likely to cause these symptoms, but i lean toward the cat being the bad, or the O2 sensor, before having made any adjustments. adjustments are a bad idea on fi systems , they usually need parts or sensor connections cleaned to restore original operation levels, seldom adjustments needed if it has not been "worked on". if you can get a digital meter and test the O2 sensor lead it should read between .2 and .9 volt dc (less than 1 volt dc) with motor and exhaust hot, the reading should vary between those readings as if it is hunting for the right place, if it does not the o2 sensor is bad and that will sure mess with the emissions readings. i have taken cars with the same problems you described and put a new cat or o2 sensor and went back and passed fine.

[GeneralDisorder]

50% denatured alcohol in the tank - retard the timing till it barely runs. It will pass.....

 

A new cat would do it also - you can get them for $50 to $75 on ebay and just weld them into the mid-pipe.

 

Both options cheaper than finding the cause - if it runs alright and you are ok with the mileage I say don't screw with it. If it were me - I would fix it - but for you that depends on how poorly it's running now and if the mileage is affected. You can dump a lot of money at a problem like that and get nowhere or only improve things slightly - it's also sometimes beneficial to get it to pass and THEN worry about fixing all that stuff. If it's due for tags and you don't have time to do it right - just make it pass and worry about it later.

 

GD

[WoodsWagon]

It's not just a high HC reading, it's a 1,200 rpm idle.

 

If the SPFI is never idling down, then it thinks it's still cold. So, it will be dumping more fuel to richen up the mix while it's cold. So you get high HC readings.

 

This has got to have taken the gas millage down a fair bit too, if it's allways in warm up mode.

 

Replace the engine coolant temp sensor, and check the connector and the wire leading to it for green corrosion. You may have to replace a couple inches of wire and the connector to get it working right again.

 

Get the throttle stop back to the right position.

 

The SPFI system is really simple, and it's better and easier to fix it right rather than rigging it with retarded timing, denatured alcohol, and a backed out throttle stop screw, or whatever hack fix you can come up with.

[brysawn]

50% denatured alcohol in the tank - retard the timing till it barely runs. It will pass.....

GD

 

I'd like to know more about this. What does this option fix? Or in better words, what tests does it make it pass? and how?

[eulogious]

I'd like to know more about this. What does this option fix? Or in better words, what tests does it make it pass? and how?

 

I have never done this but logic would say that burning a different type of fuel would give you different reading out your tail pipe since you are burning a different fuel. My guess is that the alcohol burning emits the correct HC readings to fool the machine into thinking it's running correctly, when in fact it's just burning different fuel. Retarding the timing probably makes it burn a tad less fuel? making it pass or something... Just a guess on this, but it makes sense to run a different fuel to fool the computer at least. I can see that helping for sure.

 

Good trick to know just in case. Luckily I don't have to worry about emissions, but in case I ever do, good trick!

[GeneralDisorder]

Dentatured alcohol is (typically) 90% ethenol and 10% methenol (to make it undrinkable). Just like E-85 it dramatically reduces HC emissions and reduces all your other emissions by a significant amount. It will run, but not all that well. You won't like how it runs.

 

Retarding the timing causes the engine to run hotter - further reducing emissions. There's lots of speculation on exactly why this works, but ultimately emperical evidence sugests it does work. Base timing for the SPFI is 20 degree's - knock it back to 10 to start with.

 

Fixing it right is prefered, but I've worked with engines that couldn't pass due to worn internals - this works on them. And if you don't have time to fix it - it's an option.

 

GD

[brysawn]

Dentatured alcohol is (typically) 90% ethenol and 10% methenol (to make it undrinkable). Just like E-85 it dramatically reduces HC emissions and reduces all your other emissions by a significant amount. It will run, but not all that well. You won't like how it runs.

 

Retarding the timing causes the engine to run hotter - further reducing emissions. There's lots of speculation on exactly why this works, but ultimately emperical evidence sugests it does work. Base timing for the SPFI is 20 degree's - knock it back to 10 to start with.

 

Fixing it right is prefered, but I've worked with engines that couldn't pass due to worn internals - this works on them. And if you don't have time to fix it - it's an option.

 

GD

 

That's cool and good to know. But wouldn't doing this cause an odd idle? and wouldn't that make it fail?

[GeneralDisorder]

That's cool and good to know. But wouldn't doing this cause an odd idle? and wouldn't that make it fail?

 

It will idle - just won't have much power.

 

The more ethenol you add the worse it will get. At 50% you are still running 50% gasoline so it's like E-50 at that point. Most of the gas stations sell E-10 anyway as regular unleaded so it's not *that* big of a difference.

 

It just helps a lot with emissions and in conjuction with some timing retard, oil/filter/air filter change.... it's often just enough to bring them under control. Do this to a good running car with good cats..... you'll blow basically zero's across the test and the testing guy's will be amazed .

 

DO NOT buy isopropyl alcohol - NOT the same stuff. I told a co-worker to buy denatured and he bought the isopropy by mistake..... big problems - not flamable enough I guess.

 

GD

Edited April 2, 2010 by GeneralDisorder

[Skylar]

i know this is an old post but by engine coolant sensor do you guys mean thermostat ? i have an 88 wagon with spfi that failed emissions yesterday. dont really wanna pay the 150 at a shop but i probably will have to.

 

Hc limit was 220, emissions was 390.

Co limit; 1.2. idle emissions; 5.24

any info would help. thanks !

[zukiru]

it's the coolant temp sensor for the ECU

it's located in the coolant crossover pipe that runs by the intake.

it has the same plug as a fuel injector.

the wiring/ harness is more often bad than the sensor is.

 

autozone, etc. sells these connectors with 6-8 inches of wire for less than 10 bucks.

 

also have seen the wires soldered right to the sensor, eliminating the easy to corrode connector.