2004 Subaru Forester Throttle Hesitation

[MRC01]

I recently did the 100k service for this old car, timing belt, pulleys, tensioner, coolant flush, water pump, etc. Clean UOA, no coolant in the oil, no oil in the coolant, head gaskets are original and OK. Running like new with good power & fuel economy. Stick shift with new clutch installed @ 77k miles.

Just recently an intermittent problem started: sometimes if you apply more than 1/2 throttle at around 2,000 RPM the engine surges and hesitates. While it's doing this, if you let up on the throttle a bit the problem stops and it runs smooth. If you push down on throttle a bit, it suddenly "kicks in" and pulls hard like it should. This problem doesn't happen all the time, it's intermittent. Doesn't depend on whether the engine is cold or warm.

My initial guess is the throttle position sensor has developed a dead spot. I am going to inspect it, see if I can open it up to expose the internals, clean contacts, etc. Any other ideas?

 

[Lubener]

That, fuel filter and fuel pressure is what I would check. If your TPS has a dead spot, replace it. Don't bother cleaning or disassembling it.

 

[MRC01]

Wow, prices on replacement throttle position sensors for this car are all over the place, ranging from $25 to $175!
The high end of that range is a lot of $$ for what amounts to a simple variable resistor.
Any suggestions on good brands: Beck Arnley? NTK?

 

[wb2]

As a former owner of a 2008 Outback, if this only occurred after the timing belt change and all the other work was done, I would be looking for a loose vacuum hose or a loose clamp on one of the air intake hoses or throttle body. Or maybe an electrical connection not pushed in all the way. A decent amount of things need to be disconnected / removed when changing the timing belt. Only after checking the simple things would I start testing the TPS or looking at fuel pressure. Has the car been checked for any pending fault codes?

 

[MRC01]

...Has the car been checked for any pending fault codes?

Yep, my OBD-II reader says no fault codes. But then, intermittent problems don't always throw the appropriate codes. I've seen bad O2 sensors, bad IACVs, etc. that didn't throw any codes. Also, it's quick & easy to test the TPS, so I'll at least try that.

 

[MRC01]

I backprobed the TPS and measured the voltage with the engine on (but not running), slowing moving the throttle manually from idle to max. It doesn't have a dead spot, but it was not adjusted right. It's supposed to read 0.5 V closed (idle) and 4.5 to 5.0 open (full). It was reading 0.2 at idle and 4.0 at full. I adjusted it so it reads 0.5 and 4.6 respectively.

Now the car idles way too fast, and it has serious rev hang. Due to my adjustment, the ECU thinks the throttle is open when it's closed. I disconnected the battery for a few mins, then reconnected it to reset the ECU. Now the engine is gradually bringing the idle back down to normal (700-800 RPM). The good news is I can't reproduce the hesitation & stuttering. But it was intermittent, so it will take time to know whether it's really fixed.

Any tips to force the ECU to re-learn the idle quickly?
Something like this actually works, it's not mythology?

 

[MRC01]

I answered my own question and I'll post the results for the benefit of others.

After adjusting the TPS, the engine had the symptoms described above and needed to relearn the idle. I drove around town for 30 minutes but this did not help, the symptoms didn't change. Then I followed the procedure in the above video. The idle is now perfect and the rev hang is gone. So it works.

Maybe the ECU would eventually have re-learned the idle on its own. But it hadn't done that in 30 minutes of driving, who knows how much longer it would have taken. Yet this procedure did it in half that time.

 

[MRC01]

The engine still has this intermittent problem, so I'll check the fuel filter next, per the above suggestion.

 

[Rand]

do those even have a filter beyond the sock in tank?

 

[MRC01]

On this '04, the fuel filter is a black metal cylinder next to the windshield wiper fluid tank. Super easy access. On the newer cars it's inside the fuel tank, much harder to access.

 

[MRC01]

Replaced fuel filter. No difference, so I guess the original wasn't clogged or dirty. Same symptom: at low (1750-2750) RPM, if you apply about 2/3 throttle, it causes a rapid pulsing hesitation. It's a fairly light pulsing, not shaking the car, but it's definitely enough to feel. If you change the throttle setting (more, or less), the pulsing goes away. It still pulls strongly and smoothly at full throttle, so it must be getting enough fuel. No error codes.

It's not a serious problem, so maybe give it time. It either gets worse or it doesn't. If it doesn't, drive it another 100k miles and don't worry. If it does, it will be easier to troubleshoot. I'm about ready to chalk this one up to "It's an old car, it runs fine and is reliable. But it's developed a quirk or two in its old age."

 

[mk378]

Misfiring at low rpm and heavy throttle-- always think weak sparks first. No mention yet by the OP of plugs or wires.

Fuel starvation shows up first at high rpm and heavy throttle such as when climbing hills. That is when the demand for fuel is highest.

 

[MRC01]

I installed new plugs a year ago (also ensured gap was correct). The old ones were fine and came out fairly clean despite having nearly 100k miles on them. So the plugs are fine. Coils & wires are all OEM, but visual & tactile inspection shows no cracks or issues with the wires. All 4 are indeed firing with big fat sparks that are easily visible even in daylight if I remove one. High RPM from 3/4 to full throttle is strong & smooth. Idle and throttle from closed to about 1/2 throttle is smooth. It's the range from 1/2 to 3/4 throttle where it has some pulsing & hesitation.

 

[mk378]

How old are the wires? There's no way to be certain that spark wires are good by looking. If it's intermittent, mist some water on the wires from a spray bottle and test drive. Also unplug the coil end of the wires and check they are not burnt or corroded there.

 

[frisco]

There is a filter sock attached to the fuel pump. The access panel is either under the rear seat, or under the trunks carpet.

Fairly easy to remove the pump assembly, 4 bolts for the cover, then 8 for the assembly. Sorry i forgot the size of the nut.

My brother had similar, but not exact, (2006 forester) intermittent problems you described. His problem was melting/arcing of wires. Worth a check.

 

[MRC01]

Since the engine pulls hard at full throttle all the way to redline, I doubt the problem is inadequate fuel pressure or flow.
It could be the wires. It might also be the manifold pressure sensor.

I pulled that sensor. The car is applying 5 V to one of the pins, so it's "alive". The sensor's resistance measures about 5 kOhm or 10 kOhm depending on which pins. However, if I understand how the pressure sensor works, that resistance should change if one applies vacuum to the sensor nozzle. Atmospheric pressure is full throttle and applying vacuum would be like part throttle. But the resistance is always the same. If that is how the pressure sensor should work, then it's not working.

 

[MRC01]

PS: it's not the manifold pressure sensor. I just backprobed it while the engine was running. It reads about 1 V at idle and jumps to 3 V when I blip the throttle.
So... that leaves the spark plug cables. They look fine, exterior insulation is still in good shape, flexible, not cracked. But who knows...
If I ring out the wires, they should be about 10k ohm each, right?

 

[mk378]

Disconnect the wires at the coil. If any of those connections are burnt or corroded, you need new wires and a new coil. Otherwise just new wires. Resistance checks are a fool's errand, lots of false negatives. The problem usually is that the sparks leak out.

Not sure why you haven't done this yet, it's literally the first part you see after opening the hood.

 

[MRC01]

I did physically inspect the coil & wires. The connections are clean (nothing burnt or corroded) and the wires insulation is all in good shape. Also unplugged each in turn while the engine was running to confirm spark. Nothing obviously wrong. So I went on to inspect other things. Now those checked out so I'm back to the wires. The wires are inexpensive so I think I'll just replace them. I just hate replacing things "randomly" unless I can test/confirm they are the root cause.

 

[Trav]

You may have a crank sensor going bad. If you have a scope you can watch it, the bugger may not throw a code until it is totally shot.
Do not trust ohm readings on these, many times they will show the correct ohms until heated then drop out.