legacycentral bbs

Hey all,

I'm having serious hesitation problems, but they seem contrary to most of the stuff I read about on here. I thought that it warranted its own thread.

First off, my car had minor hesitation problems when it was stock. I ran new Ground wires, and the car seemed to run better. Mostly, when it was cold it would hesitate. As soon as the weather heated up, the car would run fine. The new ground wires seemed to have fixed this.

Here's the wierd part. When STOCK, the car seemed to have 2 boost profiles. Sometimes, the car would only reach exactly 5psi and sometimes it'd top out at 9psi. I tried but I could not determine any sort of pattern as to when it would choose one of these "profiles" over the other. I know my boost gauge isn't suspect because vrg3's scantool reported the exact same thing.

I chalked this up to a broken Boost Control Solenoid, so I bought a Hallman Manual Boost Controller. We first plumbed it in as a replacement to the BCS. The hose from the turbo compressor went into one end and then the hose to the wastegate went into the other end (we actually had to replace the wastegate hose, as it was too large to fit on the MBC). I know the MBC worked, because we were able to adjust the boost. I put it out at 5psi first, then adjusted it all the way up to 12-13psi (felt the fuel cut on the freeway), then adjusted it back down to 9psi.

Well, after about an hour of driving, the car would start hesitating BAD. The boost is consistant, but the car would barely pull up to speed. As soon as I put the BCS back in, it ran great again (even though most of the time it only hit 5psi).

Things we tried
-Plumbing the MBC in as a replacement to the BCS
-Plumbing the MBC in series with the BCS
-Originally my boost gauge was plumbed into the line going to the blowoff valve. In case that caused any problems, we remove that (didn't change anything).



I'm at my wit's end here. I can't think of anything else that would cause this kind of hesitation. The two boost profile problem COULD be related, but it could also be completely unrelated.

What do you guys think?

Just out of curiosity did you leave the Original Boost Solonoid Plugged in electrically?

Yes. I understand that it throws up an error code if I disconnect it.

I also left that vaccum hose connected to the bottom of it on there.

There are times when the ECU stops driving the boost control solenoid.

I do not know for sure, but I believe one of these times is when it detects abnormal amounts of detonation.

It also does it when it is unable to properly control boost with it.

The former is certainly also associated with timing retard, and it's possible the latter is as well. So if the MBC is making the ECU think it's failing to control boost, then the ECU may be doing things to the timing to try to protect the engine. That may be some of what you're experiencing.

Also, Nigel recently noted that his ECU seemed to occasionally stay in closed loop even when it started to go on boost. He didn't say if he has stock boost control or not, but I have noticed on my car that if I use aftermarket boost control I can sometimes end up in the same situation, which certainly causes a hesitation-type effect. My guess is that the ECU wasn't designed to expect boost at such small throttle angles.

Hm. Maybe I need to come up with a cost-effective version of a certain tuner's ridiculously expensive "EMI adapter."

That makes sense.

Another question I had in relation to this... I was using your scantool and noticed my timing was getting retarded a bit. Up to -6 degrees or so. I didn't know if this was at all common at the time, and it's slipped my mind until then.

For what reason am I getting detonation, if this is the problem? I've run nothing but 92 octane in it since I've owned it. I should try some octane boost with the MBC, see what that does.

If you're seeing that much retard, then the ECU's definitely seeing some knock.

Have you ever replaced your knock sensor?

No I have not. Nobody's ever really explained to me what the symptoms of a bad knock sensor are.

EDIT: After reading that technical service bulletin, the reason I thought my knock sensor is fine is because I've never seen a CEL.

The symptoms can vary, but among them are hesitation and excessive timing retard.

The original knock sensor design was prone to failure, so they redesigned it. I highly recommend you get an updated knock sensor and install it.

So do you think a failing knock sensor would account for both the hesitation with the MBC as well as the two different boost "profiles"?

I'll get on replacing that as soon as possible.

It may.

You can check if your sensor is the old design or the new design by looking at the connector on the sensor. The original one was gray and the redesigned one is white.

If you replace the sensor and you're still seeing lots of retard, your engine may in fact be detonating, and we'll have to figure out why.

Ouch!

I just called the dealership (Carter Subaru in Seattle) and they quoted me $70.87 for that knock sensor! (he had 9 in stock).

Anyone know of somewhere to get it any cheaper?

Uh, that might have been a mistake. Did he give you the part number he was looking at?

The sensor you want is part number 22060AA031, and list price is around $65. So it should be around $50 from a discount dealer.

If Carter won't give you a discount on it, try somewhere else... I know subaruparts.com has it for about $50, but I think they're actually pretty near you so you might be able to pick it up directly. (Tacoma Nissan-Subaru?)

No, the guy on the phone verified the Part # with me (and it was the same number you listed).

I'll check around.

After the knock sensor I'd say fuel pressure. Sounds close to what I had with a bad FPR. I had the bleeder valve up and running long before that so I knew it was fuel. I had a sometimes fast sometimes fall flat if I got in the boost thing going on. I hooked up a fuel pressure gauge so I could see it while I drove and found that base FP was good but it didn't always rise with boost.

I'll keep that in mind

When it comes down to it... smooth power delivery can either be one of two problems. Fuel or Air

I'm shopping around trying to find a better deal on a knock sensor. Carter Subaru wants $78, Bow Wow auto parts wants $65. Subaruparts.com wants $49 (plus shipping, plus at least 2 weeks wait).

One of three problems: fuel, air, or fire.

OK, it's almost definitely the knock sensor.

I drove around a bit running vrg3's scantool. The hesitation comes hand in hand with Timing Correction. The correction would go as far as -4 degrees on the stock Boost Control Solenoid. With a MBC cranked to 8psi, it'd get up to -6 degrees.

also, the "Pinging" sensor would occasionally flicker to "on"... so I think all signs are pointing to a knock sensor malfunction.

I HOPE it's the knock sensor, and my engine isn't really detonating. Any reason it would be detonating?

Yeah, it does seem like the sensor might be getting false positives.

Our engines can detonate without it being audible to the untrained ear (or even to the trained ear without special listening instruments), but it shouldn't happen much with a mostly stock engine... But there are some things that in general can cause pinging. Excessive carbon buildup can cause an increase in compression ratio which can result in high cylinder pressures. Excessive blowby can lower the effective octane of the fuel. A failing fuel pump or pressure regulator, or a clogged filter or bad MAF sensor can cause occasionally lean fuel mixtures. Stuff like that.

My car is doing exactly the same thing. I'll try to change the knock sensor soon, i'll let you know the results.

Knock sensor didn't really do it for me.

The false positives on Detonation are gone, and the car doesn't hesitate AS MUCH. But it still hesitates.

I even changed the inlet on the MBC, it no longer comes from the turbo compressor outlet (like the BCS did) but it goes from the manifold.

D'oh.

Have you tried switching back to stock boost control though? That'll probably work pretty well now.

The manifold probably isn't a great source for the manual boost controller... You can end up with weird things like full boost at part throttle.

When I replaced the knock sensor originally, I had it running on the stock BCS. Only after a day or so did I switch it to the MBC, where it got worse.

It was after driving it for a day or so that I swapped it to the manifold (as it says in that HOWTO posted in the library).

Tomorrow I'll put it all back to stock and try to track down the problem.

Is there any easy way to tell if fuel pressure is my problem?

I had a plain old fuel pressure gauge. What I did was put a brass T fitting after the filter with a extra 4in of fuel line to go to the rail. (so I wouldn't have to cut the stock fuel line) I then put about 2ft of fuel line on the T fitting and ran it to the gauge. Snaked it out under the back of the hood and used the wiper and some tape to hold the gauge on the windshield where I could read it.

My new idea has to do with seeing if the ecu gets funny when it can't control boost. I have a bleeder valve on my MAP sensor so if I set it so the ecu never sees over stock pressure that would do it. You could do the same with a FCD.
I can't do it (in Iraq) but would be interested if someone did. If you have vrg3's scan tool you could see what timing looked like at different boost levels (to the ecu) The stock timing must be very conservative with no IC. Might be some good power there...

Yeah, I disagree with Matt's instructions in that HOWTO, unless you're using a big turbo and running a lot of boost maybe...

Hooking up a fuel pressure gauge (make sure you get one designed for EFI systems!) like -K- describes is the cheapest way to check fuel pressure. It should always be about 36.3 psi above manifold pressure.

Oh, are you still seeing two boost profiles under stock boost control now?

Have you gone through all the vacuum hoses to check that they're still flexible and have no cracks?

-K- - I want to determine that too, but it's hard to collect data about multiple things to correlate them. You're right that stock timing is pretty conservative, but the ECU does have the ability to be both conservative and aggressive with timing, which is one reason a lot of people feel like their cars accelerate well sometimes and badly other times.

I just hooked up the stock BCS today (re-hooked it up actually).

I'll let you know if I'm seeing the "2 profiles" thing later tonight.

I'm suspecting I don't have a vaccum leak, as the vacuum gauge gets good solid readings... somewhere around -20 at idle and as low as -26 when cruising on the freeway.