Alrighty.. I have a feeling this will be a long post. Most of it is just a list of what I've checked, the problem itself is simple. The car is an 86GT, 2.8 auto. The computer chip is stock 86 auto CA (AATC chip), the only change is the EGR is disabled. I will put the EGR together later, but first I want to sort this issue out.
Problem: I need to pass a "two speed idle" emissions test. In part of the test they will check emissions while revving at 2500rpm - the engine is stumbling every 2 seconds when it attempts to do this.
Details: The engine idles fine when it's cold. I can increase RPM with the throttle and it will rev steadily. When it gets above about 160F coolant temperature, it begins to stumble when revved. It stumbles about once every 2 seconds. When it gets completely warmed up after driving, it then starts to stumble at normal idle.
I don't know if the "hunting/stumbling hot idle" problem is connected to the warm stumble at faster RPM, but maybe it is.
- If I drive against the brake, it does not stumble. Only when revved in park/neutral. - However, I do believe I can sense a slight miss when driving at very light throttle and low load.
- Log says that 160F is not the onset of closed loop operation, so that is not what's triggering it. Closed loop happens much earlier. - If I unplug the MAP sensor, the stumble at fast idle problem goes away. However, the driveability gets worse. - If I unplug the TPS, it doesn't fix anything. - If I swap my MAP sensor with the sensor from an 84, the problem stays the same. - If I connect the MAP sensor with a spare hose, directly to the manifold without the T, the problem stays. Tried this on both sides of manifold. - If I disconnect the vacuum to the fuel pressure regulator, the problem stays. - Suction at the vacuum hose feels as strong as it does at the manifold. - The reference voltage at the MAP sensor connector is steady 5.01v with the engine running. - The signal voltage at 1500rpm warm was about 1.2V. At this speed it stumbles, and when it does, the voltage changes *after* the engine speed changes, not before. This is on a digital multimeter though so for all I know, there could be a delay in the reading. The sensor info can also be seen in the WinALDL log.
I found a major leak at the fitting to the EGR valve. That has been fixed but the stumble issue remains. The EGR valve itself was checked out of the car, it holds suction and does not allow you to blow through it. EGR vacuum is currently capped off and disabled in the chip.
A log when driving shows BLMs of 120 in many cells of the RPM vs MAP table. But it appears that for the cells which apply at fast idle the BLMs are in the 125+ range. The 120 cells are mostly those with a higher MAP.
Logs ======= Here's an ALDL log showing what happens as I rev the engine and it warms up. I hold at ~1500rpm, and when it gets above 160F it begins stumbling. Later I drive against the brake and there's no stumble: http://armos.fastmail.fm/ph...20428_165745_LOG.txt
One simple question: Is the MAP sensor a wear item? As noted above, I tried swapping in the MAP sensor from an 84 and it behaved the same. But both sensors are old and say "GM" on them so probably originals. Is it likely they both have the same subtle fault? The readings seem to behave "normally", but I'm no expert at that.
I appreciate any ideas of what might cause this problem. It just repeatedly stumbles when revved at idle, but seems to run fine under load. I have new plugs on hand and plug wires on order. I also have a new TPS on hand which I don't expect will help.
The next post will contain some sensor graphs from the log data so it's easier to see. It might take me some time to get them up. Thanks for any ideas.
I had trouble with the width of these images. Hopefully they aren't annoyingly wide on older monitors, I shrunk them down a bit but this made them harder to read. Underneath each picture is a link to the full size version. The values on the left column of each graph correspond to the item appearing in red. The other values can only be inferred by looking at the range for that sensor shown at the upper right.
full size image In this graph, first you see the RPM held at 1450rpm. After things start to heat up, it begins to stumble. Throttle opening is steady through all of this. Then I tried a faster engine speed and it still stumbles. At the right edge of the graph, I put the transmission in Drive and run it against the brake. Here it does not stumble, which I find interesting. The ECM was in closed loop for this whole graph.
Below is a graph of the same thing, but showing the MAP readings: full size image
full size image Same thing again, but with more sensors shown together
Now see what happens when the MAP sensor is unplugged: full size image MAP is unplugged and there's no stumbling anymore. But the logs above don't seem to show a problem with the MAP sensor. Or do they?
I'm not opposed to buying a new MAP sensor but I can't see much evidence that it's bad. I also tried swapping an old MAP from an 84 and it didn't fix the problem. It makes me think that something else is happening when I unplug that sensor which is masking the problem. I can of course look at specific things in the logs if anybody suggests it. For those who wish to look at them, the raw log files are at the bottom of the first post. Thanks for any ideas.
[This message has been edited by armos (edited 05-09-2012).]
I'm no expert, but first, recognize that MAP sensors have a colored plastic insert visible in the connector cavity. Make sure that it's green (or the same color as your connector). Sensors with the same color insert are calibrated identically so using the '84 MAP may not have given you reliable results.
Next, when the MAP sensor is unplugged, the ECM substitutes a fixed value for MAP (as your graphs show) and ditches the usual equation for calculating the fuel injector pulse width by relying instead on the more crude signal from the TPS to control fuel delivery. Given that your stumble goes away in this mode, I'd say there's a good chance your problem is related to the MAP sensor. The other indication is that under load (trans in gear), the MAP sensor is forced to operate in a different range of its operation and the idle smoothes out. To me, this indicates the erratic MAP sensor readings in your graphs could be the cause the engine stumble.
If that's not it, then the formula for GM fuel injector pulse width in closed loop operation is as follows. It indicates that an unstable alternator output could also result in erratic BPW:
BPW = BPC * MAP * T * A/F * VE * BVC * BLM * DFCO * DE * CLT * TBM
1. faulty spark control: either your coil, pick up, or ICM might be a source of the problem, though that wouldn't explain why it smoothes out under load; or
2. faulty fuel delivery: erratic or low fuel pressure caused by a faulty pump, pressure regulator, or partially clogged filter could cause most of the symptoms.
3. Unlikely, but what the heck... faulty IAC: it may be trying to adjust itself in off-idle conditions throwing the MAP for a loop. If you unplug it, it will remain at the position it was at the time of disconnection preventing it from cycling. Try unplugging it while running the engine at the higher idle speeds to see if the stumbling goes away.
Thanks for the feedback. I looked at my MAP sensors today and I don't see anything but black plastic. But the 86 does have what looks like it could be a faded greenish dot of old paint, the 84 doesn't even have that. When I looked up the part online it seems the same part number is sold for both cars. Still, I am near the point of getting a new MAP.
Interesting point about the battery voltage. The current alternator is a CS130 installed last September. I haven't noticed any pulsating lights or that type of thing. The original alternator I inherited with the car *did* have that problem though. I have no idea how long it was driven like that. I wonder if the BVC value is based on raw +12v, or if it's filtered. I suppose excessive noise on 12v could have killed some filter capacitors or other parts in the ECM.
quote
1. faulty spark control: either your coil, pick up, or ICM might be a source of the problem, though that wouldn't explain why it smoothes out under load; or
The ignition coil was replaced sometime earlier last year (I know that doesn't necessarily mean it's good). The distributor was pulled last August. I sanded some rust and replaced the pickup coil. I kept the old ICM. I did notice the previous owner left an old ICM in the trunk so it may have a history. At the time I was fighting a severe misfire which would occur a few seconds after starting the engine. It kept doing that after the distributor job, but then "mysteriously" fixed itself and never came back. It did have the stumble when revving though.
quote
2. faulty fuel delivery: erratic or low fuel pressure caused by a faulty pump, pressure regulator, or partially clogged filter could cause most of the symptoms.
In the August episode I also checked fuel pressure. It was 43psi with key on, and dropped to 39psi after 20 minutes. I haven't checked it since then. Fuel filter was replaced about a year ago. I'll have to look into the fuel pressure regulator, I don't know how to test that. I did notice it ran worse when I unplugged the vacuum to it, but that doesn't prove much. It does seem to me I can smell fuel underneath the intake plenum, but to some extent maybe that's normal. Dropping only 4psi in 20 minutes seems like it must not have any significant leak.
quote
3. Unlikely, but what the heck... faulty IAC: it may be trying to adjust itself in off-idle conditions throwing the MAP for a loop. If you unplug it, it will remain at the position it was at the time of disconnection preventing it from cycling. Try unplugging it while running the engine at the higher idle speeds to see if the stumbling goes away.
Today I tried unplugging the IAC while holding at about 1500rpm with the engine stumbling. It kept stumbling after I unplugged it. I forgot to mention that about 1-2 months ago, I did remove the IAC and sprayed carburetor cleaner through everything liberally. I also scrubbed the sides of the inlet hole with a rag as best I could. I cleaned off the throttle plate, which was filthy. I replaced the IAC with new, but it didn't fix the hunting idle or rev stumble.
Today also installed the new Carter TPS which I had ordered a while back. It didn't help which isn't really a surprise. The old one was marked 0286 which makes me think it was original. That said, the new one had some other number that doesn't look like a valid date code, so I may have misinterpreted it's meaning. I also wiped and sanded lots of grime and corrosion off of the ground connection between battery and frame (the ground connection just below the battery). Put it back together tight with a new bolt and painted over it. I don't know if it did any good, I haven't run the car through a heat cycle yet.
A couple days ago I added a connection from an upper intake bolt to the psg side hinge, so it should short to the factory ground strap that's already there. I have not checked the condition of the other end of that factory strap - I presume it attaches somewhere on the firewall side of the engine.
I'm anxious to get my new plug wires from Summit, who keep moving back the ship date with no end in sight. I may go ahead and buy a MAP sensor without waiting for that.
Here's what the color coded MAP insert should look like:
It should match the color of the actual connector:
I checked three of my engines though, and only two of them had inserts (green), so it's hard to know for sure that your MAP is the right one or not given that the inserts obviously can fall out.
@Bloozberry Interesting.. my 86 MAP is on the car but I looked at the 84, and it's just black on that part which is green on yours. It doesn't look like anything has fallen out either. I do think my 86 had a deeper cavity so maybe something fell out of that one.
Thanks for that link to decoding the numbers dodgerunner.
The sensor on the car is marked 460 7304 The 84 sensor is marked 039 5285
So according to that site, "460" and "039" are both 1-bar. I wonder why they have different numbers, but supposedly they interchange. If the date code information is correct (Y DDD=day of year) It looks like both of them are 2 years newer than their respective car. I guess dealer mechanics liked to swap them out. Of course I'm just assuming the "7" is 1987. Maybe it's 1997, or 2007, who knows? I just love 1 digit year codes.
Anyway, it's getting replaced. From searching I haven't seen any brand preference but I'll order a Delphi MAP sensor.
Fiddling with grounds hasn't given any results, even though I am still getting dimming lights when I turn the fan on high. I guess I need to try to find the factory grounds on the firewall side of the engine and see what shape they're in.
I seem to have an "extra" vacuum hose under my plenum which is broken off at both ends. From looking at the vacuum routing diagram somebody posted elsewhere on the site, I can't find anything to account for that hose. It's almost like somebody replaced it and didn't take out the old one. I hear a whistling noise but idle speed is normal and it stalls if I block the IAC inlet. It might just be the inlet whistling. I'm tempted to order the RD steel hoses. Of course money makes me hesitate when I have immediate issues to resolve and there doesn't seem to be an obvious problem with vacuum. Still, those chincy plastic hoses don't make me happy. I may talk myself into them.
I'm suspicious about the firewall exhaust manifold, I hear ticking noises as I drive. Would that cause these symptoms? Stumbling every few seconds under no load, but runs steadily under more load? My BLM values are also mostly low, and INT is low.
[This message has been edited by armos (edited 05-11-2012).]
IP: Logged
03:26 AM
phonedawgz Member
Posts: 17104 From: Green Bay, WI USA Registered: Dec 2009
All Fiero MAP sensors are the same. I wouldn't spend the money if both work the same.
You can buy aluminum tubing for model aircraft to make your vacuum lines from, if you wish to save some money. You can paint them black if you want them to look like the plastic ones that came on the car.
Does the idle speed issue go away with the O2 sensor unplugged?
All Fiero MAP sensors are the same. I wouldn't spend the money if both work the same.
You can buy aluminum tubing for model aircraft to make your vacuum lines from, if you wish to save some money. You can paint them black if you want them to look like the plastic ones that came on the car.
Does the idle speed issue go away with the O2 sensor unplugged?
How are your engine to frame grounds?
I'll have to try the O2 sensor experiment. Not sure about engine to frame grounds. I added a ground from the EGR bolt to that ground point below the battery, it didn't seem to affect anything but it's just a 10ga wire, not as heavy as I read some people using. I probably need to find all the original grounds and see what shape they're in.
Had a random event today, one of the fuel injector fuses blew. I replaced it with another 5amp and so far it's stayed intact.
If I unplug the O2 sensor, it stops stumbling, just as it did when I unplugged the MAP. I have noticed that my BLM values are low and INT is low, so unplugging the O2 is making it run richer than it normally does. However, under low load conditions my BLM aren't quite so bad, they start around 125 and end up at 120 as the MAP increases. INT is low as well, it's in double digits under idle conditions.
With MAP unplugged, supposedly the only adjustment the ECM makes is to substitute a fixed MAP value, which is a pretty low pressure value. So it seems to me that with MAP unplugged, it would actually run leaner than normal (until the O2 compensates for that anyway). So if I'm getting this right, with O2 unplugged it gets richer, with MAP unplugged it would get leaner or stay the same. Somehow, under either of those scenarios, the stumbling stops. I still hear an isolated misfire every few seconds though.
The O2 sensor is a Denso that I installed April 2011, so it's 1 year old.
[This message has been edited by armos (edited 05-11-2012).]
I tried sucking gas from the vacuum line to the fuel pressure regulator but I couldn't suck anything. It's sealed tight from the vacuum end of things.
I put a small amount of seafoam through the manifold vacuum. I was hoping to look for an exhaust leak without making a ridiculous amount of smoke in the process. Looks like I underestimated it, it only smoked for about 3 seconds and I didn't have time to see much. I did find a leak on the trunk side of the exhaust using some tissue paper, I can feel some air there on my hand if I rev the engine fast enough. It seems to be in the area of the cylinder #5 exhaust port. But I don't know if it's the manifold or just the gasket. I'm also in doubt that it has anything to do with the stumbling idle problem, and those manifolds are expensive.
One problem I have with explaining this by leaking exhaust is that it should cause the O2 sensor to read lean, giving me higher BLMs. I'm getting low BLMs, and that's mainly at higher RPM and MAP values.
I guess I really need to dig into the grounds. I'll have to see if I can find them.
I ran a larger amount of seafoam through the engine, the amount they suggest for treatment. It didn't smoke nearly as much as in the videos I've seen, so either I did it wrong or this engine is pretty clean. Anyway, I do have an exhaust leak near cylinder #5. Hard to tell if it's the manifold or just the gasket. I'm doubtful that has anything to do with the stumbling idle though. The only other leak I found is the muffler, which I knew about. I didn't see any smoke coming from the firewall/front bank of the engine, which is a very pleasant surprise.
After looking at the ALDL logs again, I noticed that when I unplugged the MAP sensor, the O2 sensor voltage was constantly high. So even though the service manual doesn't say so, apparently it runs rich when MAP is unplugged. When the MAP is unplugged, or when the O2 is unplugged, I don't get a stumble. So the common thread here is that when the engine is rich, or cold, it doesn't stumble. If it's stoich and above 160F then it stumbles.
Today I swapped the ignition module to a spare that was in the trunk. I don't know it's condition, it was there when I bought the car. I sanded the contacts and anything that would interfere with grounding the module. It still stumbles, but it's an old module so that doesn't prove much. Found out my plug wires have shipped so I'll do plugs and wires this weekend. I should have done plugs a long time ago but wanted to wait for the wires.
It sounds like a lean stumble from your description. When you change plugs look at the color of the old ones. Lean is whiter than the normal brownish. Make sure all plugs are at least alike. If one is off then investigate that cylinder further.
Can you monitor your fuel pump voltage? I'm wondering if your stumble happens when the voltage drops to the fuel pump causing a lean. When mine did that is was more of a pulsing than a stumble though. This is one of the reasons for adding more/better grounds.
That's an interesting idea about the fuel pump. Would it be just as well (maybe better) to monitor the fuel pressure instead? Then if I see a pressure fluctuation I'll try to monitor the voltage on that circuit as it's running.
I'm starting to wish I had kept the old O2 sensor. It was running a bit rich with that sensor (mileage improved after I replaced it), but it might have been just enough to stop the stumble and pass emissions. Oh well.
I switched back to the newer white ICM that was in the car at the start of this thread (module B). Previously I had swapped to an old used one which didn't fix anything (module A). Since I inherited both of these modules with the car, I don't know how good either of them really are. It did seem to me that the car misfired a bit worse with the older module "A", but I didn't keep it long enough to be sure. I switched back to "B" after a day of driving. I made every effort to have clean grounds at the ICM screws.
I changed the plugs and wires, new plugs are ACDelco R42TS. I let it warm up on the fresh plugs and found the problem still remains.
These are the old spark plugs: click here for full resolution picture (1MB) They are arranged with plug #1 at the lower right, #2 at upper right, #5 at lower left, #6 at upper left (same as if you're looking at the engine from behind the car).
Plug #2 got dropped, so the oil/grease on that electrode probably wasn't really there before I dropped it. plugs #2 and #4 had a lot of fresh oil above the threads, so I have a leak there somewhere, but the oil wasn't below the threads. The gap on the old plugs wasn't very good. The worst was #3, it measures 0.060. #1 measures 0.052. They were gapped 0.045 at installation 1 year ago on the same gauge. Plug 1 looks lean. Plug 6 looks also lean but not quite as bad. To my eye 3,4 looked maybe slightly hot but not much. 5 looks normal and it's gap was perfect. 2 was dropped through the engine compartment so it has debris on it, but underneath it's color also looked okay.
My exhaust leak at cylinder 5 was from a loose manifold bolt. After tightening it the leak seems to be gone, at least I don't feel it anymore.
I checked the fuel pressure with the engine running, and it seems normal. Pressure is 38psi while running, and it moves in a range of about +/- 1psi as the engine is stumbling. I think that would be normal with the change in manifold pressure. Fuel pressure also temporarily jumps up when you pop open the throttle, which again I think is a normal response between the fuel regulator and manifold pressure.
Electrically speaking, my lights dim when I turn on the radiator fan, which from what I've read is not normal. Most recently I took apart all the ring connectors at the +12v junction block near the battery and sanded them clean. My lights still dim after doing that so I don't think it helped. I previously added a 10ga wire between the EGR bracket and the ground bolt below the battery. I also cleaned up that junction below the battery. I have cleaned up the decklid side of the factory braided strap. The engine side of that strap still needs to be cleaned, I couldn't get to it yet. I also need to find where the main negative battery cable goes to and clean that. I'm a little unsure why a weak ground would cause such cyclical behavior where it stumbles every few seconds. But it doesn't cost anything to clean them up regardless.
To summarize where I am, the symptoms seem to indicate that the engine is stumbling under a light load stoich warm idle. If the engine is forced to run rich, it doesn't stumble. If coolant temp is below 160F, it doesn't stumble. If it's being driven against the brake, it doesn't stumble either. At moderate temperatures above 160F, it will stumble every few seconds when I rev the engine slightly. After it's been driven and it's hot, it will go into the "hunting idle" syndrome without being revved. I suspect these are both the same problem, just as it gets hotter it begins to occur at normal idle.
My thoughts at this point are:
1) Whether I should buy a new ICM, or if it makes sense for the ICM to cause these symptoms. 2) Can the O2 sensor be defective such that it's reporting a richer exhaust than reality? Does this happen? If so, the ECM will try to run too lean. The sensor is a Denso from 1 year ago. 3) It's apparent from looking at the plugs that the mixture isn't even between all cylinders, but I don't know if it's bad enough to explain this issue. Cylinder 1 does look pretty lean, 6 to a lesser extent. Since fuel pressure seems to be behaving correctly, perhaps I have bad injectors. That's really something I'm hesitant to guess at though.
Thanks to anybody who read through all that. The emissions deadline is getting painfully close and I'm running out of ideas.
I swapped in an old O2 sensor from an 84. No effect, the idle still stumbles when revved.
When I unplug a single plug wire from the distributor, it can hold a steady rev without stumbling. I tried this individually on #1, 3, and 6.
My guess is that having any cylinder unplugged causes excess oxygen to be seen at the O2 sensor so it ends up running more rich. This masks whatever the actual problem is. It's probably the same effect as when the O2 sensor is unplugged, or the MAP sensor is unplugged. It seems this car has to run rich to idle properly. As it warms up the problem gets worse.
I was watching the car idle in diagnostic mode with WinALDL earlier. It showed my BLM at 120, but I'm not sure that's valid if I haven't driven the car around, I think I read somewhere that it always starts at 120 when you hook up the ALDL cable. Not sure about that.
The idle is smooth when INT is 128. Then INT gradually gets lower and lower. As it gets into the 70s it starts to feel rough. Eventually it gets down to 20. By that time the engine is stumbling badly. Then the idle suddenly recovers and smooths out, and I see INT has reset to 128. Sometimes the INT settles into a range of 60-80 and stays in that range. In that case it manages to idle without hunting and just occasional misfires, though it doesn't feel as smooth as at 128. With 1 plug wire unplugged, it would usually stay in the 60-80 range.
Everything seems to point to a lean misfire. Plug #1 looked the leanest of the old plugs, plug #5 looked good. Since those are in the same bank they should both be getting the same injector pulse. Fuel pressure only moves by about +/- 1psi when it's stumbling, and that's probably a normal reaction to manifold pressure. Last I checked, fuel pressure only loses 4psi in 20 minutes with the key on. So I don't think the injectors can be leaking. I had 2 plugs whose color looked good, and the other 4 were different degrees of lean.
=========== Is it possible for a faulty injector to spray too much, but not be leaking? Given that the plugs show an uneven mixture within the same bank, is there some other explanation for that besides the injectors? If injectors are unbalanced, does the ECM tend to go overall lean in that scenario, or rich?
IP: Logged
10:49 AM
May 30th, 2012
masospaghetti Member
Posts: 2477 From: Charlotte, NC USA Registered: Dec 2009
I'll be keeping an eye on this thread...I'm having a different issue but I noticed my INT values are also low, I was under the impression that they should average around 128 but mine are often in double digits...but I'm getting a spark knock in prolonged highway drives.
I have refurbished injectors in shipping, I'll post back with the results once I get them installed. I believe I've cleaned up all the chassis to frame ground connections with no effect. From what I've read there's another ground for the ECM which I need to look at, and the battery to engine ground.
The old injectors have the Rochester "RP" logo on them and appear to be Multecs, from searching it looks like this part number are multec injectors GM used on 2.8 and 3.1 engines: 5306 5235136 From the info I found, apparently there's some disagreement between sources of whether the flow rate is 15lb or 16.8lb/hr. So it looks like the injectors have been replaced before. I don't think multecs were original to anything earlier than 1988, and this car is a late 86.
The replacements I installed are 17lb/hr refurbished multecs from "Precision Auto Injectors". Intake gaskets were also replaced. Thus far it now seems to hold a fast idle at higher temperatures without stumbling, which is wonderful but I haven't run the car enough yet to be convinced it's resolved. I drove it once and the light throttle response was smooth and stumble free, but again I haven't driven it enough to draw final conclusions.
The erratic behavior at hot idle is still there though - it still goes into fits of stumbling from 600-1200rpm as it did before. I've noticed it smooths out when I flip on the radiator fan. It might be electrical but I've had no luck finding it yet. The exhaust is out of the car now so won't be able to test drive it further until that's back together. The muffler was a bag of catalyst chunks with a hole in it.
IP: Logged
08:31 PM
Francis T Member
Posts: 6620 From: spotsylvania va. usa Registered: Oct 2003
BTW: a faulty MAT -Mass Ambient air Temp or outside air temp the one before the TB- sensor can also cause this problem. I've seen it happen on our wideband, at first it was just in a specific RPM range, but then it drove ECU nuts.
[This message has been edited by Francis T (edited 06-05-2012).]
BTW: a faulty MAT -Mass Ambient air Temp or outside air temp the one before the TB- sensor can also cause this problem. I've seen it happen on our wideband, at first it was just in a specific RPM range, but then it drove ECU nuts.
Mine was reading very high temperatures about a year ago, but after cleaning the sensor the readings went back to normal. WinALDL is still reporting normal temperatures at the MAT. Something weird I noticed on mine is that it has some electrical test points inline with the sensor connector. Somebody before me was trying to measure the resistance in-circuit I guess. And yet it was faulty when I got the car, so I'm not sure what they accomplished.
Finished putting the exhaust back together. It needed a new muffler but then I found the catalyst was falling apart so replaced that too. The exhaust is fine but guess what.. fast idle is stumbling again. I thought that was fixed, or at least it was for 1 day.
Low idle seems to stabilize when the radiator fan is on. I think I'll just ground a wire under the front hood and make it run full time. If that makes it run steady then so be it. 195F should be enough to pass a test with a new cat. I don't think it's illegal yet in Kalifornia to modify your radiator fan. Might also play with the throttle body screw to get more air through the throttle instead of that black IAC tube (which I forgot to clean when I had the manifold off). I know you're not "supposed" to mess with the throttle but somebody did that long before me.
I'm thinking about making a chip to force it to run more rich, because it runs smooth that way. But I don't know if that will just guarantee failure at the emissions test or if it's better than stumbling. I swear the electronic controls on this car are defective from the factory. Why did no other car have the same ECM as a V6 Fiero? Because it doesn't flipping work, that's why. I can sit there in WinALDL and just watch it drop the INT leaner and leaner until the idle falls apart. The ECM is an idiot.
Although the idle is still screwed up, it does drive better than before. Light throttle response is smooth and steady now, not stumbley jerky as before. I assume the injectors are what did it.
I found out the idle for testing has to be below 1000rpm in neutral. It goes above that when it recovers from a stumble. So the only "fix" I can think of at this point is to artificially richen the mixture just enough to stabilize the idle, and hope it passes. I have a new catalyst so that should give me some leeway I hope. Looking in TunerPro there's tons of things I could screw around with relating to mixture, but I guess the most likely approach would be to change the "minimum BLM" value.
Is there any particular chip adjustment that's recommended for Multec injectors? That's what I found in the car and also what they were replaced with, so that's a common factor. I'll try to look into that.
If there's some way to control the range of INT values I'd love to adjust that instead of BLM, but I don't see a setting for it. It idles fine when INT is around 60 or above, but when it decides to head for 20 that's when the idle falls apart. Rigging the INT to a min of 60 or whatever would be a simple workaround.
IP: Logged
03:25 AM
Francis T Member
Posts: 6620 From: spotsylvania va. usa Registered: Oct 2003
We found with our headers that the engines idle better when cold if you swap over to a 3 wire heated o2 sensor. While youre not using headers it still may be something to look into. US Auto Parts has one for about $300 - USAP-3000
IP: Logged
07:43 AM
Jun 11th, 2012
StockGT Member
Posts: 386 From: O.C. , California Registered: Feb 2007
Is the cap missing from the throttle body idle stop screw ? Your post on 6-8-12 mentions the throttle body screw.
Looking at the winALDL Raw data, which I have not used much, so others may be able to confirm my understanding of interpreting the data. If I am looking at the correct column, the RAW:IAC is 0 much of the time in some of your logs. The Factory manual lists typical IAC count value in the range of 1 - 50. (at idle, engine operating temp., closed loop) From past posts on resetting the throttle body idle stop screw , a reading in the range of 25 - 40 is good at warm idle. I don't know if the IAC count returns to zero while driving ?
The values of the PROMIDA and PROMIDB seem to change over time. My understanding is these values should remain constant.
Is the cap missing from the throttle body idle stop screw ? Your post on 6-8-12 mentions the throttle body screw.
Yes, it's missing. I tried doing the adjustment procedure a few months ago but it didn't work for me at all. I could hear the IAC supposedly closing but it never stopped. After I got sick of waiting I continued with the procedure, and when I started the engine the IAC clearly wasn't closed, it was quite open. So I apparently did something wrong there. Much more recently I opened the throttle to see if essentially bypassing the IAC would smooth out the idle, but that didn't work. Then I closed it shut, that didn't help either. Unplugging the IAC mid-hunt had no effect either. The throttle body has been cleaned and IAC replaced. Having convinced myself that the problem wasn't related to the throttle or the IAC, I ultimately just left it open enough not to stick. I could try doing the proper adjustment procedure again but honestly I think it's fine where I unscientifically have it now. The IAC is in control, the throttle is just barely open.
quote
Looking at the winALDL Raw data, which I have not used much, so others may be able to confirm my understanding of interpreting the data. If I am looking at the correct column, the RAW:IAC is 0 much of the time in some of your logs. The Factory manual lists typical IAC count value in the range of 1 - 50. (at idle, engine operating temp., closed loop) From past posts on resetting the throttle body idle stop screw , a reading in the range of 25 - 40 is good at warm idle. I don't know if the IAC count returns to zero while driving ?
The values of the PROMIDA and PROMIDB seem to change over time. My understanding is these values should remain constant.
I looked at the above-linked logs in ALDLView (this is easier than reading it in Excel). I can't find anywhere that the IAC goes to 0, and the PROMIDA/B values show up as static values. If you mean they were changing from one log to the next, then that's possible. I have done some fiddling with chips but I think all the logs above were done on the same chip, at least I intended them to be. Stock CA chip but w/ EGR disabled, nothing funky that would complicate the issue. It definitely acts the same on the actual OEM chip as well, I've had it in the car at times over the past couple months.
We found with our headers that the engines idle better when cold if you swap over to a 3 wire heated o2 sensor. While youre not using headers it still may be something to look into. US Auto Parts has one for about $300 - USAP-3000
My problem is with idling when hot. Cold is okay. It's always okay in open loop, probably because it's rich. It's still okay in closed loop for a while, but not after it gets sufficiently heated up. The stumbling rev starts at about 160-170F coolant temperature, and the hunting idle starts when it's fully heated up around the 200 range.
After the injector replacement job, the stumbling rev problem has been reduced but not eliminated. It's now somewhat intermittent instead of constant. It will stumble several times, then smooth out.
[This message has been edited by armos (edited 06-13-2012).]
Retard the timing. Stupid but effective solution to a hunting idle.
After some research, I found that the car has to idle below 1100rpm in neutral for the duration of a 30 second test. So that was the key problem I had to get through in order to take an emissions test. Running rich wasn't an appealing solution. Since my injectors are Multecs, I tried to copy some tables from a later car that uses Multecs but those tables are broken in TunerPro. GMEPro was able to edit part of what I wanted, but it didn't fix anything.
I loosened the distributor and waited for the engine to start hunting. Then I moved the distributor. As soon as I retarded it, the idle stabilized. Advance it back, and it hunts again. The timing was set correctly by a timing light. It's hard to know what angle you're supposed to be viewing from, but it was somewhere near 10BTDC.
California checks the timing, it has to be accurate within +/- 3 degrees. So I put the distributor back in the correct position, which looked like somewhere between 8-10BTDC depending on your preferred point of view. Then I retarded it with a chip instead. 5 was iffy, 6 was good so I went with 7 degrees retard vs stock. Having done that, it reliably idled smoothly. It still hunts in datalogging mode, but even more aggressive retard didn't fix that and they don't test in that mode so it doesn't matter.
In normal ECM operation 7 degrees retard vs stock still hunts if I'm rolling down the road in neutral, but if the car is sitting still it idles fine. I tried neutral at several stop lights and in a parking lot before the test, it was fine. The way it drives does feel retarded in this setup so I don't think it's just an inaccurate timing mark.
I was a bit amused to see the inspector had a hard time reading the timing, which comes as no surprise to anyone who has tried it on these cars. I lost sight of him for a few minutes and I suppose he was looking up some diagram of the timing marks. Then he returned and made another attempt with the timing light and was apparently satisfied. Earlier, on an F150 pickup, I saw him shoot the timing in 5 seconds flat.
The 2500rpm test made me nervous. I heard it stumbling every few seconds as he did this. After 1 minute the stumbling stopped. I don't know if it was stumbling during the interval that was tested, but the tailpipe results were excellent in any case. Luckily the revving stumble stayed within allowed limits, whatever they are. The low idle test was smooth, I didn't hear it acting up during that.
My test report shows negligible emissions at idle and 2500rpm. 1ppm HC at idle, and I think 2ppm at 2500rpm. CO was also extremely low. A new catalyst is an amazing thing.
Now that it's through emissions this idling problem has been reduced from a government crisis to a minor nuisance. I may try making a chip that retards timing only at idle, but I'm not sure if there's an appropriate way to accomplish that. This is a silly thing to do though. I don't think timing is the real cause, it just masks whatever is wrong.
My INT moves around a lot while idling, and from what I've read INT should be stable. I don't know what this indicates though.
[This message has been edited by armos (edited 06-13-2012).]
IP: Logged
04:01 AM
PFF
System Bot
StockGT Member
Posts: 386 From: O.C. , California Registered: Feb 2007
In looking at the RAW winALDL data, I was mistakenly looking at wrong columns. Using ALDLviewer makes it easy to review the Raw data !
Have you done an engine compression check ?
Not recently. My notes say I checked compression in Aug 2011 and it was good at the time. I don't have the numbers though. It looks like I first noticed the revving stumble problem about 1 month before that. The hunting idle had been there at least that long or longer as well.
This is an old thread but I want to update it with the final solution, in case somebody else reads this someday with the same problem. Previously I worked around the problem by retarding timing, just so I could get it to run an emissions test. The real fix (works with correct timing) seems to have been making a new chip to calibrate the ECM for the multec injectors (which I believe came from a 3.4L F-Body). I'm not ready to call it perfect but it seems to be basically fixed. Details were posted here: https://www.fiero.nl/forum/Forum2/HTML/128975.html
