52 replies to this topic

[86 Wonder Wedge]

Before I get too deep into it, I know I'm probably gasping at straws, but off all the extensive diagnostics I've done, this is only thing that has be scratching my head.

First, the problem: At idle, I get an intermittent misfire like pop from the exhaust when warm, is somewhat hesitant at part-throttle (sub 3K RPM), gas mileage has been down, and at 10k miles, the old plugs have got a white powder (like burning coolant/too lean, but no coolant loss or vac leaks (21 in.Hg at idle)), hunts at idle, studders and I can hear the injector "hiss" randomly in sync with the popping in the exhaust.

Now, the CTS, IAC, injector, MAF, O2 (just put a new one in anyway, 126K) TPS (new Hitachi unit) ALL check out. System voltage @ 14.02V at idle and holding.

I did have an IAC code when I bought it in May, but new throttle body gaskets solved that code.

Now when I test the MAF, it gets the 14.02V, has a varying sensor voltage and starts at 0.125V, and the ground, KEOE (key on, engine off) reads 3.6 ohm (from connector to body ground).

NOW. When the car is running, and MAF unplugged, ground back probed at the connector and to the body ground with my DMM, it reads 3.5, 3.6 ohm, then every 4.5/5 seconds, it flips to -16.5 ohm for a second, then back to 3.6. but if I measure it with just the ignition on, it holds steady at 3.6 ohm. WTF? Also, when the MAF is unplugged, the idle smoothes out, but smells rich,

any thoughts? Is this affecting my running? is this normal?

Edited by 86 Wonder Wedge, 14 December 2012 - 04:45 PM.

[AKghandi]

you have the flapper style maf correct?
like this one?


if so use a screw driver and pop the top cover off

after you have it off look at the contact area(will be self explanatory when you get it apart.) and see if you can see where the contact arm has worn through the contact area.these are called "dead spots" which would cause your symptoms.

as far as fixing it goes you might be able to use some window defrost repair paint(never tried this as my maf is the hot wire type) or just get a new one.

Thats all i know about that so i HTH:)

[djellum]

cant really tell from your post, but if its a flakey white buildup on the plugs it can be ash from contaminated fuel or burning fuel additives.

have you or do you put anything in your fuel?

Like i said, cant really tell if you just meant white plugs from running lean, or if there was actually some powdery ash on them.

[86 Wonder Wedge]

Thanks for the replies guys!

Nope, its a hotwire maf... 91 loyale, spfi.

And the white stuff was on the arm itself, and powdery white.. Every once in a blue moon..

[Fairtax4me]

White powder = lean mixture.
Are only certain plugs coming out white?
If its only one or two it could mean a couple fuel injectors aren't spraying properly. Vacuum leaks are also possible. You replaced the throttle body gaskets and that cured one issue, the manifold gaskets could be bad as well.


MAF should be grounded through the control unit and it probably will screw around with the ground circuit when it's running. Its kind of a pointless test anyway since the connector is unplugged. Resistance may change when the circuit is loaded, so to get an accurate picture of what's going on you should back-probe the connector while its plugged in.

The signal voltage that gets sent back to the ECU is what you should be more concerned with anyway since thats what the ECU uses to determine what change it should make to the fuel mixture. Can't tell you exact specs, but I'm sure that info is around here somewhere.

[86 Wonder Wedge]

That's what I was wondering as well about that ground circuit, but was unsure.

Also, the white power was on all 4 plugs and I cant see the injector's spray pattern directly (before it hits the throttle plate, SPFI). The intake gaskets did look slightly wet around the port, but my vac reads 21 in.Hg. shouldn't it be lower?

Also, when I unplug the MAF, the idle stops hunting and the vac gauge holds a rock hard 21 in.Hg. With it plugged in, the vac would dip about 1-2 in.Hg when the injector would "stumble"...

[AKghandi]

Ah now hot wire mafs I know a lot more about!

I assumed from your name it was an 86 xt.. and I guess they are right about assuming

when you unplug the maf it will go into a sort of limp home mode. where it runs off a set fuel map, instead of using the maf to regulate.

first make sure your intake pipe or "snorkus", Doesn't leak at all. ye olde carb cleaner will work for deducing leaks. dont spray the maf directly though.

Mine is a turbo, but when I had a maf issue, first the intake pipe was leaking, then my maf went bonkers and it never threw a code. It caused my car to studder, cut out, and stall, all bad. And the maf would click like crazy. from all my diagnostics it was working.. I bought a JY one and all is well. I have never heard the new one click. If you can find one in a junk yard for cheap i would try that. dont clean it though, these mafs are self cleaning.

[naru]

Also, when I unplug the MAF, the idle stops hunting and the vac gauge holds a rock hard 21 in.Hg. With it plugged in, the vac would dip about 1-2 in.Hg when the injector would "stumble"...

This pretty much proves you are suffering from a lean misfire.
(ECU enriches the mixture w/MAF unplugged)
What does the O2 sensor have to say?

Could be bad MAF,injector,metered air leak,vacuum leak or low fuel pressure.

I`m leaning toward a bad MAF or metered air leak,but,I would check fuel pressure early in the diagnosis.

An engine leak down test w/attention paid to the exhaust might be interesting.

[Cougar]

Try spraying some WD-40 around the intake manifold gaskets while the engine is running to see if you have some air leaks there.

[ivans imports]

loyale spi ? have seen a bunch pound out the center of the injector pintle on thiss injector the more worn the LESS feul it moves there wierd like that mabee a new injector

[86 Wonder Wedge]

This pretty much proves you are suffering from a lean misfire.
(ECU enriches the mixture w/MAF unplugged)
What does the O2 sensor have to say?

Could be bad MAF,injector,metered air leak,vacuum leak or low fuel pressure.

I`m leaning toward a bad MAF or metered air leak,but,I would check fuel pressure early in the diagnosis.

An engine leak down test w/attention paid to the exhaust might be interesting.

I was on the same track with the O2 (it was the original one at 126k miles) and it would read, at idle, 0.210V approx. And vary as the RPM would spike or decel. I changed it since it was cheap, AND it seemed like it couldn't be doing its job cold when it should be hot before it starts working.

And fuel pressure is a rock steady 24psi at idle, 5k rpm and holds 20psi overnight. Thought about that too. Really, I'd like to test the MAF signal voltage at the ECM and/or find a CFM,RPM vs. Signal voltage table before I go writing off this MAF.

And food idea about the injector pintle.. That sucker seems to have gotten louder over the last 15k miles from when I first got it... I might just pull that thing out from its housing and take a look and make sure none of the orings or that small filter isn't causing problems...

Thanks again for the ideas guys! I'll keep us posted on results.

[skishop69]

The ground circuit is important and should be checked as you did. Fluctuating resistance in the ground circuit WILL affect your signal voltage back to the ECU because it will cause a reduction of voltage in the unit. Resistance on that circuit may change with the ECU powered up, but it should not fluctuate. If it's fluctuating when the MAF is unplugged, your problem is either one of the grounds to the ECU or the ECU itself. If it only fluctuates when plugged in, the hot wire is going bad. Also, it's a misconception that hot wire MAFS are self cleaning. If oil or other 'heavy' liquid contaminants get on it, it does not generate enough heat to sufficiently burn it off. This the causes the MAF reading to skew low. In other words, the MAF will show less air coming in than there really is, and the ECU will cut the fuel back and you will end up with a lean condition. You CAN clean any hot wire MAF as long as you use a designated MAF cleaner and you do it first thing before turning the key on. It has to be cleaned cold or you run the risk of damaging it.

[naru]

O2 sensor is telling you your lean.
Since it is responsive,it is functional.

Ohmmeter readings on a powered circuit are nearly meaningless.
Try a voltage drop test instead w/the MAF plugged in.
Put the 2nd lead on battery negative or an ECU ground.
Should be a steady 0.00 volts in a perfect world.

[Cougar]

The resistance to ground reading seems a little high to me. I suggest you clean the engine and chassis ground connections with a wire brush to see if you can get the 3.6 ohm reading closer to zero. The -16 reading you got intermitently indicates there is a voltage drop occurring on the lead which shouldn't happen if the wiring is correctly grounded.

You stated you hear a hissing sound and I have to wonder if you really hearing a leak in the intake manifold gaskets. This is a common problem for those cars. I suggest you spray some WD-40 around them while the engine is running to see see if that changes the idle speed that would point to a leak and a lean condition.

[skishop69]

Acceptable resistance on a circuit is 7 ohms or less. Voltage drop on a 12-14v system with 5 ohms is less than .2 of a volt. Negligible in the scheme of things. Just pointing that out so you don't take extra time away from the search doing something that will at best yield intangible results.

[naru]

Acceptable resistance on a circuit is 7 ohms or less. Voltage drop on a 12-14v system with 5 ohms is less than .2 of a volt. Negligible in the scheme of things. Just pointing that out so you don't take extra time away from the search doing something that will at best yield intangible results.

I agree with the sentiment,but,not the math.

Voltage drop on a 12-14v system with 5 ohms is 12-14v, not .2.
Voltage drop on a 12-14v system with 5000000 ohms is 12-14v too.

Voltage drop across any individual circuit element depends on how much other resistance is in the circuit,

For a 12 volt system to have a 0.2 volt drop across a 5 ohm resistance requires that 5 ohms to be in series with another 300 ohms.

V=IR will carry you far.

Subaru wants 10 ohms or less on the MAF and ECU to body grounds.

[skishop69]

Agreed. I have a degree in electronics and 20 years in auto diagnostics for electronics. I apologize for the typo as it should have been less than .02 volts. DOH! I also should have clarified 'voltage drop'. You check the circuit by placing one lead of the meter on one end of the live circuit and the other lead on the opposite end. Set to volts, this gives you the voltage drop on that circuit. I can tell you from experience that it doesn't always follow Ohms Law to the letter, but it's so close it's not worth ripping your hair out over. lol There are a few factors that play into it not being exact and it's just more math.

[86 Wonder Wedge]

I can tell you from experience that it doesn't always follow Ohms Law to the letter, but it's so close it's not worth ripping your hair out over. lol There are a few factors that play into it not being exact and it's just more math.

This is why I'll stick with my BS in Physics... you electronics guys believe in imaginary numbers. Something fishy there...

I'll have to wait till tomorrow to test the other grounds for the TPS, and CTS and for a vac leak at the intake gaskets. Propane (unignited of course) would work just as well, yeah?

But I know this may not help with the vac leak vs. MAF, but when DEAD cold, it has a considerable bog mid throttle, 1500>2500 RPM, but goes away if I WOT, then has an intermittent flat spot/bog above 4K RPM (at full throttle, no variance) but MOSTLY disappears after it's warm.

[Cougar]

Thanks for the info on the grounding specs guys. Even though the spec may be up to 10 ohms I usually find the real numbers are close to zero generally when the grounds of the engine and chassis are in good condition.

As far as Ohm's Law goes, it isn't an approximation and there's nothing fishy about it, it's THE LAW. If something doesn't measure what the calculation says it should be then something in the circuit or the measurment process hasn't been accounted for.

I still think the intake manifold gaskets are causing the trouble, but I could be wrong.

Edited by Cougar, 17 December 2012 - 07:59 PM.

[skishop69]

Yeah! Nothing fishy! It's the LAW! You tell 'em Coug! Physics.... Phbbbbbbttt! BS is right! Newton was a wuss. Ohm could've taken him with one arm behind is back in a static state! Roflmao Sorry, couldn't help but take a friendly jab back atcha Wonder.:-p

Edited by skishop69, 18 December 2012 - 03:22 PM.
auto correct

[86 Wonder Wedge]

Lol that's understandable. At least you're not a chemist.. Can't stand those guys..

On to business. Tested the intake manifold gaskets today, no dice, even cold. Rock solid vac reading and no idle fluctuations. I even bench tested the idle solenoid and that thing worked flawlessly. Little dirty on the pintle, so I cleaned it up and reinstalled, no love.

Tested the MAF again and got 3.6 ohm on the ground to body, but 0.2 at the ECM plug.. Hmmm.. Also got 0.3V KOEO at the sensor voltage, and 1.35V at idle up to 2.95V at 4K RPM. Even with the green "freeze frame" connectors plugged up,it still was popping and hunting... And I'm still looking, but can't find voltage ranges for this MAF.. 1.3V seems a little high for me..

Also check for actuation of the EGR to make sure it wasn't stuck and it isn't, and the throttle body, valve cover and intake gaskets are all solid.. Still holding at 21 in.Hg with slight dip randomly... And unplugging the MAF doesn't COMPLETELY solve the popping.. Its 80% less, but still happens and then smells real rich.

Now what.. :/

[Cougar]

Well so much for the intake gaskets, I really thought that was it.

Can you run a temporary ground jumper from the battery to the MAF sensor to try and eliminate that extra resistance to ground? It may not help but hopefully it is a simple thing to try out and eliminate that as a potential problem. Also, if you haven't yet cleaned the engine and chassis grounds I suggest you do that. Another simple thing to do. Faulty grounding can cause various issues. The trouble is beginning to sound to be either due an electrical problem or a fuel problem. The hissing injector makes me wonder also. Could that be due to air leaking into the fuel delivery somehow?

Edited by Cougar, 18 December 2012 - 10:33 PM.

[pressingonward]

I think the negative resistance reading you get when it's running is because that ground wire is tied into other sensors, and when it's running you get a tiny amount of voltage running through your meter to ground, changing the resistance reading.

In my experience symptoms of bad MAFs aren't that intermittent. They're usually good, bad, or somewhere in-between, but usually steady-state. Plus you say that it still pops a bit when unplugged. That suggests to me that the MAF is reacting to the problem and exaggerating it, but isn't the root cause.

A bad injector is an interesting idea, but again they tend to be either good or bad or somewhat bad, but not intermittently bad.

My bet is that you have a bad distributor. Distributors rotate so they love to do things intermittently. The camshaft position sensor inside the distributor is the primary signal to the ECM to inject fuel. If it sends out a garbled signal it will cause the coil and the fuel injector to fire at the wrong time or for the wrong duration (fuel injector), giving you a nice popping noise, hesitation, and fluctuating idle.

Make sure your timing belts didn't jump a tooth. If it's only one tooth on one camshaft it can make the engine run funny but not horrible.

[86 Wonder Wedge]

I thought about a short-cut ground, but I'm worried that the ECM needs to ground it for it's own reason... but worth a shot.

And I also wondered about the distro as well. I had replaced the cap and rotor @ 114K when I did the T-belts (and bought the car) with BWD stuff and just several weeks ago, I went to double check the distro shaft for play (since GD had an EA82 that was doing the EXACT same thing and this was mentioned) to account for a less than dead on 20 BTDC even with the green connectors plugged in.

The shaft has ZERO play in it, but I found the screw holding the rotor was loose, so I blue loc-tite'd it in there, and it improved greatly, but still not perfect.

I've yet to 2x the belt timing, but will do tomorrow...

Think it might be the CAS on its way? I was considering the coil as well, but the resistances check out and it has a steady light when I put the timing light on the coil wire. But you know what, I have seen it "cut" in and out a few times, but dismissed it since the popping and the "blanking out" were independent of each other. But if the CAS is sending a dirty/bad tach pulse or open/close signal, I imagine it would affect both the transistor AND the pulse width of the injector... Hmmm....

And might be related, but at dead cold, the car takes 2x as long to initially fire than it does when warm or semi-warm. The motor kicks about half a revolution and fires when warm, but takes 3 or 4 revs before it catches when cold, then it rocks real hard for a split second while it climbs to 2200 RPM then knocks down smoothly to 1500 then 800 when it's warm on it own.

Edited by 86 Wonder Wedge, 19 December 2012 - 11:21 AM.

[Fairtax4me]

Vacuum sounds fine. A little flutter here and there is considered normal. Certainly sounds like its going lean.

One fuel injector and it has been replaced, kinda rules it out in a way. Vacuum readings rule out a leak.
Fuel pressure seems low to me, but may be normal for these older cars.
How many ECT sensors do these things usually have? Seems like there would be two?