Need help solving an intermittent miss

[AHTOXA]

While it doesn't sound like your case, indeed an early symptom of a slipping clutch is a brief disruption of power that may feel like a miss. Especially if one isn't too familiar with either. That Guy wasn't too far out there.

[bk7794]

I don't know what the guy was thinking...Im almost afraid to go back and look at the post.

I am beginning to wonder if its two things. The air temp sensor, or something completely wrong with the ECU. "

It only does it 96% of the time when the air gets hot around the engine bay, so when its warmed up. Or the o2 sensor, unplugging it and having no cel light kind of made me nervous.

[theholycow]

Happening when it gets hot intake air is another point in favor of knock/preignition/etc, though the IAT (Intake Air Temperature) Sensor could be at fault.

How about experimenting with intake air temperature? Hack up a DIY CAI and a DIY WAI (Warm Air Intake -- google for it, you'll find ideas on hypermiling forums) and see if manipulating IAT produces results consistent with your observation. Then just switch them in winter and summer...or just keep the CAI.

Also monitor your IAT. You can use a remote oven/grill thermometer somewhere convenient in your intake (before the throttle body, of course).

[bk7794]

Happening when it gets hot intake air is another point in favor of knock/preignition/etc, though the IAT (Intake Air Temperature) Sensor could be at fault.

How about experimenting with intake air temperature? Hack up a DIY CAI and a DIY WAI (Warm Air Intake -- google for it, you'll find ideas on hypermiling forums) and see if manipulating IAT produces results consistent with your observation. Then just switch them in winter and summer...or just keep the CAI.

Also monitor your IAT. You can use a remote oven/grill thermometer somewhere convenient in your intake (before the throttle body, of course).

What I am getting concerned with is the knock sensor gone on the car. But I would think that would throw a CEL. Today I am going to reset the computer, the first time after I hit that bump, just to see if that does anything.

I was going to at one time take it off to see if it was dirty.

I will google around for that. Thanks!

[theholycow]

Knock sensor failure can go undetected, even by modern computers. The knock sensor isn't some high tech thing, it's literally just a microphone picking up all vibrations in the block; the computer is programmed with the sound of knock for that particular engine with that particular sensor in that particular location. Anything changes (corroded mounting surface, reduced sensitivity, different major parts that change how knock sounds conducted through the block, whatever) and the computer may still hear standard background noise but fail to detect knock.

[bk7794]

Knock sensor failure can go undetected, even by modern computers. The knock sensor isn't some high tech thing, it's literally just a microphone picking up all vibrations in the block; the computer is programmed with the sound of knock for that particular engine with that particular sensor in that particular location. Anything changes (corroded mounting surface, reduced sensitivity, different major parts that change how knock sounds conducted through the block, whatever) and the computer may still hear standard background noise but fail to detect knock.

I would test that but I am a bit worried about even looking at that part of the engine funny. Due to the extreme heat that the back of the engine experiences I would be concerned with breaking the connector off...then I would definitely get a CEL.

When I did that O2 sensor test I was getting very bizzare readings of .06 itstead of .6 which I am supposed to get. I am thinking it might be that sensor even though no CEL is on. I was also reading in the book that when the ECU gets a bad signal it throws in a known good signal to replace that bad one. I wonder if that is what is happening. But I don't know for sure, I am going to see if I can read the voltages again.

Also when I take my foot off the gas the injectors turn off. Saves gas! woo!

[theholycow]

Yeah, I totally understand not wanting to disturb something that may be kinda functional but will probably go completely to hell when you touch it. BTDT.

When it gets a bad signal, most computers will substitute tables of prepared data; they don't actually simulate a signal, they just do the math based on the tables.

The fuel injectors shutting off is called Deceleration Fuel Cut Off and I'm a big fan of it. Unfortunately it is rarely as dependable as one wants it to be, often finding excuses not to happen.

[bk7794]

Yeah, I totally understand not wanting to disturb something that may be kinda functional but will probably go completely to hell when you touch it. BTDT.

When it gets a bad signal, most computers will substitute tables of prepared data; they don't actually simulate a signal, they just do the math based on the tables.

The fuel injectors shutting off is called Deceleration Fuel Cut Off and I'm a big fan of it. Unfortunately it is rarely as dependable as one wants it to be, often finding excuses not to happen.

The old car blues...I hate it.

Will it throw a CEL when it substitutes the original data with the tables of prepared data?

What would cause it not to happen? Load on an engine like A/C I am assuming, along with power steering and battery draw right?

[theholycow]

The blue old car...I have it.

In general, yes, it will throw a CEL when it goes into open loop (open loop = tables with prefab data, closed loop = reading from the sensors). Your car does seem to act a little different than I might expect.

DFCO conditions in Hondas are as you guessed: A/C, wipers, fan, headlights; those conditions will generally raise the minimum RPM required for DFCO but not totally cancel it. When I had the VW, having shifted recently would prevent DFCO entirely, but it didn't seem to care about accessories so much. My GMC doesn't seem to care about any conditions, but it waits 8 seconds before DFCO kicks in - and I'm usually back on the gas before then.

[bk7794]

The blue old car...I have it.

In general, yes, it will throw a CEL when it goes into open loop (open loop = tables with prefab data, closed loop = reading from the sensors). Your car does seem to act a little different than I might expect.

DFCO conditions in Hondas are as you guessed: A/C, wipers, fan, headlights; those conditions will generally raise the minimum RPM required for DFCO but not totally cancel it. When I had the VW, having shifted recently would prevent DFCO entirely, but it didn't seem to care about accessories so much. My GMC doesn't seem to care about any conditions, but it waits 8 seconds before DFCO kicks in - and I'm usually back on the gas before then.

You can tell when the DFCO kicks in?

Ugh the other forum I posted on keep throwing me back to the ignition system. This is getting frustrating..

[theholycow]

There are a few ways to tell when DFCO kicks in...

1. Using equipment to monitor electrical status of a fuel injector's wire. I recommend sneaking a wire into a connector, rather than tapping through a wire's insulation. A duty cycle meter (or reasonable facsimile in the form of a dwell meter) does the job nicely, also giving you a relative measure of how much fuel is being used when not DFCOing. An LED or a volt meter could probably do the job just for DFCO.

2. Using OBD monitoring equipment. This generally doesn't work well, but people who haven't compared it to method 1 above by using both simultaneously don't know that and insist that it works perfectly. OBDII data updates are delayed and OBDII does not carry injector duty cycle or DFCO data so the equipment or user must guess when DFCO is engaged. The best way is by watching for open/closed loop which reacts more quickly and accurately than other data. I don't know anything about pre-OBDII OBD equipment and can't guess if there's anything for your car that will do it this way.

3. Butt dyno. When you think it's DFCOing, you turn off the ignition (carefully avoid turning it far enough to lock the steering) and see if the engine braking increases. If engine braking does NOT increase, DFCO was engaged. It can be hard to tell, as the increase may be very slight.

[bk7794]

There are a few ways to tell when DFCO kicks in...

1. Using equipment to monitor electrical status of a fuel injector's wire. I recommend sneaking a wire into a connector, rather than tapping through a wire's insulation. A duty cycle meter (or reasonable facsimile in the form of a dwell meter) does the job nicely, also giving you a relative measure of how much fuel is being used when not DFCOing. An LED or a volt meter could probably do the job just for DFCO.

2. Using OBD monitoring equipment. This generally doesn't work well, but people who haven't compared it to method 1 above by using both simultaneously don't know that and insist that it works perfectly. OBDII data updates are delayed and OBDII does not carry injector duty cycle or DFCO data so the equipment or user must guess when DFCO is engaged. The best way is by watching for open/closed loop which reacts more quickly and accurately than other data. I don't know anything about pre-OBDII OBD equipment and can't guess if there's anything for your car that will do it this way.

3. Butt dyno. When you think it's DFCOing, you turn off the ignition (carefully avoid turning it far enough to lock the steering) and see if the engine braking increases. If engine braking does NOT increase, DFCO was engaged. It can be hard to tell, as the increase may be very slight.

Number 1 I can see being plausible. That could become useful for other things aswell. Not sure I would ever do that thoug.

3. is interesting, I always wanted to turn my car off on this wicked hill but feared for power brakes. The car has no power steering so that doesn't matter too much.

So putting the car in neutral is that better or worse then just coasting down a hill in gear.

[theholycow]

3. is interesting, I always wanted to turn my car off on this wicked hill but feared for power brakes. The car has no power steering so that doesn't matter too much.

You don't lose power brakes when you shut the engine off (unless your brake booster or vacuum hose is broken). You lose power brakes after you step on the brake pedal a few times. At that point you have depowered power brakes; you can still brake. If you are a 97 pound out-of-shape stick figure you may not have the leg strength to brake well, but anyone else should be able to brake pretty effectively; the hard part is modulating the braking. It tends to feel like your options are head-through-windshield or no braking at all.

However, if you're in gear then it's a non-issue, as the engine will continue making vacuum as long as it's turning, regardless of ignition or fuel.

So putting the car in neutral is that better or worse then just coasting down a hill in gear.

This truly depends on your end-game.
- Are you planning to continue at your current speed or accelerate? Neutral wins. DFCO saves the fuel of idling but uses stored kinetic energy to turn the engine a bunch of extra revolutions.
- Are you planning to stop or otherwise use your brakes? DFCO wins. You are discarding energy either way, might as well use it to turn the engine and drivetrain instead of burning fuel for that.

[theholycow]

I was thinking about the power brakes thing and remembered a whole avenue we haven't explored...vacuum leaks. A vacuum leak could cause the symptom by making the mixture lean. Lean enough mixture for misfiring could cause weird O2 sensor readings, though I don't know just how weird.

One way to test for a vacuum leak is to get an aerosol can of starting fluid or carb/throttle body cleaner and spray it at potential vacuum leak areas while the engine is idling. If you can hear the RPM change then you've found a leak.

Another test, which can only tell you if there is a leak in your brake booster or the last section of hose going to it, is to shut the car off, wait a few minutes (not sure exactly how many minutes), then step on the brake. If it feels like power brakes then you're good. If it is hard like no power then there's a leak there.

[bk7794]

I was thinking about the power brakes thing and remembered a whole avenue we haven't explored...vacuum leaks. A vacuum leak could cause the symptom by making the mixture lean. Lean enough mixture for misfiring could cause weird O2 sensor readings, though I don't know just how weird.

One way to test for a vacuum leak is to get an aerosol can of starting fluid or carb/throttle body cleaner and spray it at potential vacuum leak areas while the engine is idling. If you can hear the RPM change then you've found a leak.

Another test, which can only tell you if there is a leak in your brake booster or the last section of hose going to it, is to shut the car off, wait a few minutes (not sure exactly how many minutes), then step on the brake. If it feels like power brakes then you're good. If it is hard like no power then there's a leak there.

Hmm that is an interesting concept for the power brakes. Interesting. Thanks!

I actually did that and found nothing thought I didn't douse the entire hose.

However, today I was doing some spirited driving and when I was high in the rev range I Was getting sort of like a fuel cut off feeling but the car would continue going u the rev range. However I do not know if I was hitting the fuel cutoff once but when I hit first I was just getting that buckign fuel cutoff over and over.

I am worried it could be a fuel pump, I Was told that they are aren't very likely to go out, but you know almost 300k its beginning to worry me.