Hesitation Blues...does a MAP sensor fail like this? Can I test/clean it?

[skizix]

Ok, I've gone over a lot in my '02 OBW (2.5, 5MT; bought it this past spring with 100k miles), and dealt with most of my driveability issues. However, one (or possibly two) remains.

 

Main symptom is hesitation on takeoff. Intermittently. Sometimes I'll go a whole day without it, other days it's almost every takeoff, other days it's ~50/50. Hesitates from 1-3 seconds, with very little power, then takes off like someone poked it in the rump roast. Occasionally it drops out so bad that I stall trying to get going (usually this happens when it's been good for a while, so I've gotten used to not feathering the clutch). Does not seem to happen on a fully cold engine, but kicks in way before full op-temp (like half way between C and the middle). Not entirely sure it never happens cold, but I think that's right.

 

The other symptom I've got is rough acceleration and deceleration. Tough to be sure, but I think it's likely the same cause as the hesitation. Not a harsh buck, but the power does not ramp up or down smoothly. Like I'll be cruising along, hit the gas (not necessarily floored), and it'll tug and let go and tug and let go. Gets up to speed in a reasonable amount of time, just not in a smooth curve. Oddly, it does this when I let up on the gas a bit and slow down, too. Feel the engine brake more then less, and guh-guh guh as it decellerates. These both follow a similar intermittent pattern -- some days it doesn't do this at all, some nearly always, some 50/50. Not a devastating driveability issue, but super annoying.

 

Also, once in a blue moon, I'll get what feels like miss, usually while at maintaining a slow, constant speed, say in a parking lot. Not convinced it's a miss (no miss code), but I do have a little oil on the plug terminals -- need new rocker cover and pipe gaskets, fixing this weekend -- so it could be a miss, but I'm not convinced. And I've never been able to percieve any knock/ping myself, even when I occasionally grind 'er too hard at low rpm's. Also, occasionally when at idle, rpm's dip a bit. Not a lot, but a bit, sometimes for several seconds at a time.

 

So, having replaced plugs and wires (both OEM), front 02 sensor (OEM), IAC (with a used OEM that looked brand new), neutral switch (OEM), fuel filter (fram, like an idiot), and air filter (K&N drop-in). I also got an ECU flash from a dealer, as their was a TSB (still somehow cost me $100) for this, to deal with hesitation by dampening response to front O2 sensor. Also: throttle body cleaned, IAC passage cleaned, TPS recalibrated (wasn't reading 100% at WOT). Coolant temp sensor and thermostat cleared by trusted mechanic. Mech also said: no intake leaks or exhaust leaks, and cat was not obstructed. I also ran a can of Techron and a can of BG-44K-or-whatever-it's-called injector cleaner (the good stuff, suposedly).

 

No codes, except Catalyst Efficiency (420). Mech says I need a new cat, after looking at rear O2S output. Not sure if I buy that (going to replace rear O2S this weekend and see if that works). Could be cat is overwhelmed by bad mix from something else.

 

So, remaining likely suspects would be: MAP sensor, knock sensor, coil, and.......you tell me. Coil tested fine for resistance, looks undamaged/no cracks, did not arc visibly in the dark, even when sprayed with water mist. But one terminal seems to develop a bit of corrosion/carbon slowly. Mechanic said he looked for knock sensor problems, but I think he was looking for defects in the circuit, rather than oversensitivity. He also said the MAP sensor looked ok, but...how did he determine this? BTW, seems as if a funky MAP sensor could definitely cause the 420 code.

 

So I'm starting to think it's the MAP sensor, based on the intermittent behavior, and because the problem exists with **decelleration** too. Problem seems to pop up when load **changes**, not just when load increases.

 

Questions:

1) what do you think this could be?

2) if MAP sensor...how can I test it?

3) if MAP sensor...can I clean it? What should I use?

4) anything other to suspect?

5) how can I zero in on the problem without blindly replacing parts?

 

thanks everybody.

[skizix]

Oh, and (as if I haven't been long-winded enough)...could my MAP sensor fail gradually like this? Googling seems to suggest that failure of a suby MAP of this vintage would be more likely to cause a no-start condition. However, honda, toyota, etc. forums talk about MAP sensors going south and causing symptoms much like mine. They also suggest that this is gradually eating my engine (preignition, etc.) if that's what it is.

[Manarius]

When my MAF went, it did the exact same things you describe. Replaced it with a known good one and has worked ever since. I don't advise cleaning them since they are very fragile.

[NOMAD327]

Knock sensors get blamed quite a bit for this problem on Subarus and don't seem to ever cause problems on other brands. I have a 99 with hesitation just on initial rolloff, but nearly all the time. I am going to replace front O2 sensor and plug wires just because it's time, and has helped some other people. My knock sensor failed about 6 months ago and set CEL, I replaced it with new Subaru part, and on mine at least, it didn't help hesitation. On my car it is a MAF sensor right after air cleaner and I did clean it, no help. If your K&N filter is too oily, or over time, it will put oil film on the sensor, but that is probably not your problem.

[xrturbo]

When my MAF went, it did the exact same things you describe. Replaced it with a known good one and has worked ever since. I don't advise cleaning them since they are very fragile.

 

This car will have a map and no mass air. It sounds like the same prob we had on an 02 forester. For us it was that airflow metering front O2 sensor.

[Manarius]

This car will have a map and no mass air. It sounds like the same prob we had on an 02 forester. For us it was that airflow metering front O2 sensor.

MAF and MAP do the same thing, so I think their failure symptoms would be somewhat intertwined.

[skizix]

Thanks for some feedback.

 

Pretty darn sure it's not the front O2S. Replaced mine with an OEM unit < 10,000 miles ago, and then a few months later, briefly swapped out the new one with a known good one, just to make sure I didn't by chance get a bum unit, since I was still having suspicious symptoms. Did not change things.

 

Yes, my 2002 has no MAF, just a MAP. I don't quite fully understand the difference in what is being measured, except that the MAF measures the amount of air flowing past it, upstream of the throttle, by the air's capacity to conduct heat away from a heated filament, at any given moment. The MAP measures air **pressure** downstream of the throttle. No filament -- not sure what exactly it is, but I'm assuming either solid state, or some kind of thin diaphragm. Must be some qualitative difference in what is being measured, since some cars have both.

 

So the mechanism is completely different, possibly implying a different sort of failure mode, and different destructive or otherwise side-effects of cleaning it. MAF's reportedly fail gradually like this, but mostly I've heard that a MAP will either work properly or crap the bed completely so the car won't run at all -- and throw a code (a MAP code that is -- yes, my 420 could seemingly be caused by the MAP measuring incorrectly, throwing off the mix). BTW, I've heard of folks having successfully cleaned MAF's, and I've also heard plenty of advice not to try cleaning it. Can't find any info on cleaning a MAP. But since after my throttle body cleaning, the IAC passage was still pretty gunked up and cleaning it helped...I'm thinking the MAP my be as well, since it's in a similar position.

 

So, anybody had their MAP gradually go south? Anyone had success cleaning one? Any way to test it?

 

I'm leaning towards a faulty MAP, mostly since I would not think a funky coil or knock sensor would act up on decelleration, since engine load is minimal. But the MAP, since it **measures** load on the engine, just might. Does that sound right, or is it faulty logic? It would be good to get closer to some determination, since of the three, the MAP is the most expensive to replace.

[BitBasher]

I have a similar hestitation problem on my '01 mt Forester. No problems with rough accel or decel though. Replaced the front O2 sensor (OEM) - no noticeable difference. I did tweak the TPS slightly (can't remember which way) and that did help a bit. I had a look at the TPS resistance with an ohmmeter and it seemed to have perfectly smooth action, but perhaps it's acting up. Anybody had to replace a TPS prematurely?

[mtsmiths]

We had exactly the same problem, and after new plugs, wires (the O2 was 'fairly' new), new fuel filter, new air filter, checking fuel pressure,it was, in fact the MAP sensor.

 

Sokay, prolly needed all those things anyhow, sure needed plugs!

[BitBasher]

Thanks mtsmith, I'll be on the lookout for an MAP sensor.

[99lego]

my phase II 2.2 recently had a recall for the mass airflow sensor

[skizix]

99lego, we're talking manifold absolute pressure sensor, not mass air flow sensor, here. Similar purpose in life, I think, but a totally different device.

 

Anyhow, thanks for your replies. BitBasher's clinched it for me. Today I found out a MAP was only ~$100 (still a lot for what it is, but I fully expected two or three times that). So I just got one. Put it in just now (super E-Z), and after a short drive, it seems to have fixed things. Tough to be sure, as it's been intermittent, but so far so good. I'll follow up with results over the next week or so.

 

One thing: when I pulled the old one, there was what looked and felt like engine oil on it, and in the hole it mounts in. This thing pops into a hole right in the top center of the intake manifold. Did not look like K&N oil -- not pink, but a nice clean amber color, exactly like engine oil.

 

So......does this mean my PCV valve is hosed? Any other reason to see oil in the intake manifold??? I actually requested that my mechanic replace the pcv last time I was in, but he talked me out of it, saying, "those things never go bad."

 

Where is the pcv valve on an '02 OBW (2.5 5MT) anyway? Is it easy to replace? Is there anything else that could cause this? Should I put the old MAP back in until I fix this, for fear of ruining the new one?

[janewolff]

i was having the same problems with my forester (2000), even replaced many of the same parts that you did and it's still not running.  mine started out hesitating occasionally, usually after the engine was hot but it finally died all together.  the MAP was also covered in oil and looked pretty beat up so I ordered a new one, i'm hoping it fixes the problem. 

 

how is yours running now? did you check out the PCV? just curious how it's going as it sounds like the same issues at my forester

[Fairtax4me]

9 years old thread, doubt you'll get a response on that.

 

There is often engine oil in the intake manifold due to the nature of the way the PCV system works. The PCV sucks oil vapor out of the crankcase, and it tends to condense and collect on the walls of the manifold. Oil/gasoline vapors are generally not good for the MAP sensor, and can cause failure of the sensor with exposure. If there is an easy way to mount the sensor away from the manifold that can prevent future problems.

GM and Ford often have problems with manifold mounted MAP sensors. Subaru MAP sensors seem a bit more robust, but as these cars get older manifold mounted MAP sensor failure is seeming to become more common.

 

Subaru used to mount MAP sensors on the passenger strut tower with a long vacuum hose leading to it with a small filter. Those sensors rarely fail. 99% of the time when someone gets a CEL code for a MAP on those older cars, it's because of a split vacuum hose, which costs $2 to replace.

 

If the PCV valve is old, replace it. Subaru only on that, aftermarket a can cause other problems.

 

Another common cause of acceleration problems is a bad knock sensor. The housing cracks with age and allows moisture to get to the element inside, which causes it to corrode and send false or erratic signals. Easy to replace. Search for my knock sensor 101 thread for details on that.

[janewolff]

oops should have looked at the date on that one.

it has a new knock sensor that i replaced a few months ago (housing was cracked).  i also replaced the o2 sensors, plugs and wires, ignition coil, all of the exhaust which was crazy rusty and new the MAP sensor.  it's idling better but dies when i give it gas.  i'm going to check the PCV next 

[Trailcutr]

A bad knock sensor would throw a code but its not noticeable driving like a bad maf would be.