Jump to content

Photo

99 idle drifts to very lean

- - - - - Idle AFR lean

  • Please log in to reply
18 replies to this topic

#1
HeavyRightFooT

HeavyRightFooT

    Member

  • Members
  • PipPipPip
  • 11 posts
  • Location:Bay Area
  • Region:California
  • Car Year:1999
  • Car Number:77

Hi All ! Need some help troubleshooting this issue. Searched the forum but didn't find anything closely related. 

 

Building a 99 hoping to join the NorCal spec Miata crowd this year.

Was hoping first posts would be of joy and excitement related to the first race weekends.. unfortunately the garage section it will have to be !

 

Just installed a new engine. Previous engine ran fine when removed, just old and keeping aside as a spare.

 

New engine went in without issues. Started fine other than idling high but adjustment of throttle cable and idle screw fixed that.

 

AFR at start looks fine, however after a few minutes my AEM wideband shows the AFR drifting to lean until the gauge cant even display anymore. (18+AFR !)

 

Sounded strange to me so first thought there must be an issue with the wideband.

Happened to have a spare Wideband kit so swapped the wideband 02 sensor and then the whole gauge itself but no difference.. it looks like it actually reads correctly. FYI Sensor is located on top of mid pipe around where the transmission fill plug is.

 

Since it appeared to start acting up while the engine is warming, I replaced the temp sensor at the back of the head. No change.

 

Thought maybe a big old vacuum leak, sprayed Brake cleaner around all the potential locations I could think about and no luck, vacuum hoses, intake, injectors, no change in rpm or signs of leak. 

 

Only thing I changed during the swap other than the engine itself (long block including intake) was the primary oxygen sensor which was completely rounded off by the previous owner and positively now one with the pipe. Procured a used pipe and installed new OEM sensor. Using Racing Beat extension to reach the plug in the front of the valve cover (Ex California emission car).

Could it come from that if it is defective ? Once the ecu goes to close loop idle, does it use the o2 readings ?

 

Still have the secondary o2 sensor plugged in to avoid engine light but zipped tied to the PPF by the trans. 

 

Otherwise everything looks ok to me. Checked all the hoses. plugs, contacts. Engine runs fairly fine, can hear a very slight change in tone as it warms up but nothing dramatic. Revs immediately if I touch the throttle, which I avoid to do much until I can figure out what exactly is happening here. I'm not going to put any load on it if it really idles at over 18 AFR... 

 

I mean, would it sound fine idling at that kind of AFR, or run at all ?

 

Once fully warmed if I turn it off and then back on hot, wideband shows AFR restarts around 14.7 and then again progressively drifts back to oblivion lean.

 

Fuel pressure is at 52psi. Idle at around 850. Timing at about 14 with the jumper on. When jumping GND and TEN idle barely raises.

 

Going to smoke test for leaks tomorrow evening, but would appreciate any tips or leads as to what to look for next !

This is starting to be a bit of a puzzle to me.

 

Attached is a log of the wideband reading on the top of the graph and rpm bottom for context. cold start. AFR crests at about the time the coolant temp passes 160.

 

Cheers

 

Oliver

Attached Files



#2
Steve Scheifler

Steve Scheifler

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 2,816 posts
It certainly sounds like faulty readings. Since you have attempted to eliminate sensor & gauge (but not wiring), what catches my attention first is that idle was off after install. If idle screw adjustments were large that may be a clue.

After hot start, does idle speed change as it drifts lean? It should if actually going that lean. Also with cold start vs hot start, sounds like it acts exactly the same in terms of AFR reading either way. For you “cold” isn’t particularly cold but still it should read a bit richer initially when starting cold, vs hot, does it?

You quote fuel pressure, if you have an adjustable regulator turn up the pressure a few PSI to see if that’s reflected in the AFR. If it doesn’t the sensor/gauge is likely wrong.

I’m at a disadvantage not having experience with CA cars but you lost me talking about “Procured a used pipe and installed new OEM sensor. Using Racing Beat extension to reach the plug in the front of the valve cover (Ex California emission car).” I have no idea how those two sentences connect. What plug in front of the valve cover? But the wrong sensor or bad reading from it may lead to the ECU slowly trim back fuel hunting for stoich.

Try increasing RPM and watch AFR, you won’t hurt it in neutral. If fresh air is getting to the gauge the higher exhaust flow at 4500 or so may give a better reading.

Try driving it at minimal load. If it’s really that lean and stays there it won’t even pull away without stalling or having a serious lean surge. At those minimal driveway loads you won’t hurt it.

It needs to run a bit before plugs are likely to show much but how are they looking?
  • HeavyRightFooT likes this
Instigator - Made a topic or post that inspired other Broken record - You are starting to sound like a broken record.

#3
HeavyRightFooT

HeavyRightFooT

    Member

  • Members
  • PipPipPip
  • 11 posts
  • Location:Bay Area
  • Region:California
  • Car Year:1999
  • Car Number:77

Thanks for the reply and suggestions ! Much appreciated. Ill investigate these avenues this evening.

 

To answer your questions:

 

-Hot starts idle doesn't really change much in speed. I'll double check later today.

 

-Cold start AFR starts with 12-13 for a few seconds then spends a few minutes around 14.7 then drifts to lean. Hot starts AFR starts at 14.7 for a shorter amount of time and drifts to lean.

 

-CA emissions cars have a pre-cat in the header and the primary o2 sensor on the header before said converter, close to the head. This primary o2 sensor plugs on a connector situated on the front of the valve cover.

My car was converted to a federal emission exhaust system by the previous owner, with the primary o2 sensor now located in the front of the mid pipe close to the bottom of the header like most other miatas. However the primary o2 sensor plug is still on the front of the valve cover and so requires moving that connector or using an extension to be able to plug the sensor.

When I installed the new engine I tried to remove the primary sensor from the pipe. Sensor was stuck in the pipe and completely rounded so just bought another used mid pipe and installed a brand new primary o2 sensor.

Apologies if this was confusing, I was trying to provide as much context as possible.

 

I'll report back later on the points you mentioned.

 

Thanks !



#4
gerglmuff2

gerglmuff2

    Member

  • Members
  • PipPipPip
  • 284 posts
  • Location:St Paul MN
  • Region:LOL
  • Car Year:1993
  • Car Number:81

these small engines dont push a lot of exhaust at idle, pretty common i think to stop getting results at idle. my 1zz-fe and 2zz-fe mr2s with fat exhausts both went to full lean after a few minutes at idle, never and issue. my spec miata at idle will drop in and out of reading. 

you probably have some exhaust leaks. 

 

i frankly wouldnt worry about it. as long as the readings are good at 5k+ and WOT, thats all that matters. 


Gordon Kuhnley: Driving miata's in all conditions, courses, and motorsports that I can. 


#5
Steve Scheifler

Steve Scheifler

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 2,816 posts

these small engines dont push a lot of exhaust at idle, pretty common i think to stop getting results at idle. my 1zz-fe and 2zz-fe mr2s with fat exhausts both went to full lean after a few minutes at idle, never and issue. my spec miata at idle will drop in and out of reading.

you probably have some exhaust leaks.

i frankly wouldnt worry about it. as long as the readings are good at 5k+ and WOT, thats all that matters.


I generally agree with this IF the exhaust is potentially leaky which is why I suggest holding up to 4500 or so to see if AFR come back around 14-15. Similarly the light load test drive to see if AFR looks more appropriate, but his concerns about going straight to 5+k WOT are reasonable. (Though at an actual 18:1 it should fall on its face before it can hurt itself)
Instigator - Made a topic or post that inspired other Broken record - You are starting to sound like a broken record.

#6
Ron Alan

Ron Alan

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 3,732 posts
  • Location:Northern CA
  • Car Year:1995

I've been doing this 10 years and I can tell you I suck as a trouble shooter...I rely on smarter people than me to help...they are on this forum. That said...I just dealt with the exact same issue!

Because I don't think these cars just go lean`...I for sure thought I had a bad O2 sensor. Because the unit was new the manufacturer actually sent me a replacement...which wasnt the problem :(

 

The problem was I suck as a welder! My O2 bung looked like I had gotten it welded all the way around...piled up nice and high :) But when I sprayed some wd40 around it I could see it seep inside the pipe! Mine is located between the flange and the resonator on the SD exhaust. I separated it and watched...well, out it came, cut off the bung, a little grinding and welded in a new one! This time I did it correct...drilled a hole big enough so the the relief on the bung dropped in...didn't sit on top! Yes...live and learn  :(

 

I guess enough outside  air was coming in to show up on the sensor! Though I always thought  it would go out? Regardless...solved the problem.

 

I would put the car on the dyno and see what it reads at the tailpipe! I bet its normal! Start your FP at 55...you can always take back out!

 

I'm in the East Bay...hope to see you out soon! Maybe we have already talked...I just don't know you as heavyrightfoot!


Ron

RAmotorsports

 

Donor - Made PayPal donation Bona fide - A bonafide Spec Miata driver

#7
HeavyRightFooT

HeavyRightFooT

    Member

  • Members
  • PipPipPip
  • 11 posts
  • Location:Bay Area
  • Region:California
  • Car Year:1999
  • Car Number:77

Hi all, thanks for all the help !

 

Tried all the suggestions but nothing changed anything... added fuel pressure, nothing, tried keeping it at 4k, nothing, swapping o2 sensors, nothing. Smoke tested for leaks and a little some came from what appears to be the EGR valve but a tiny thing. 

 

All this makes me think its back to the Wideband itself.



#8
HeavyRightFooT

HeavyRightFooT

    Member

  • Members
  • PipPipPip
  • 11 posts
  • Location:Bay Area
  • Region:California
  • Car Year:1999
  • Car Number:77

I've been doing this 10 years and I can tell you I suck as a trouble shooter...I rely on smarter people than me to help...they are on this forum. That said...I just dealt with the exact same issue!

Because I don't think these cars just go lean`...I for sure thought I had a bad O2 sensor. Because the unit was new the manufacturer actually sent me a replacement...which wasnt the problem :(

 

The problem was I suck as a welder! My O2 bung looked like I had gotten it welded all the way around...piled up nice and high :) But when I sprayed some wd40 around it I could see it seep inside the pipe! Mine is located between the flange and the resonator on the SD exhaust. I separated it and watched...well, out it came, cut off the bung, a little grinding and welded in a new one! This time I did it correct...drilled a hole big enough so the the relief on the bung dropped in...didn't sit on top! Yes...live and learn  :(

 

I guess enough outside  air was coming in to show up on the sensor! Though I always thought  it would go out? Regardless...solved the problem.

 

I would put the car on the dyno and see what it reads at the tailpipe! I bet its normal! Start your FP at 55...you can always take back out!

 

I'm in the East Bay...hope to see you out soon! Maybe we have already talked...I just don't know you as heavyrightfoot!

 

Hi Ron ! Not sure if we have me in person yet. My name is Olivier. I rented a spec Miata in 19 for a race at Sonoma and had met a few people there but its been a little bit. Hope to meet you in person soon.

 

Used to race bmw e30s and spec e46 up in the northwest. Moved to the bay and took a little break but its time to go back racing :)

 

Well, Replaced the wideband gauge, sensor, an now wiring, and thought I had eliminated this, but After all this, it does look like it must be the wideband. 

As I read your post I realize I did weld the bung in myself as well, using a shaped aftermarket bung with a curve instead of the step, so as to simplify the mating of the bung to the pipe. Have had success with these on other cars in the past so thought it would apply here as well.

Just like you I'm not a pro welder but its on and no leaks at the welds. However, I did notice that the sensor had just a little bit of play in the thread while screwing it in. But didn't think about it twice as the sensor went in fine and tightened correctly.

 

I'll check that this weekend, maybe just plug that one bung and weld the one that came with the kit just to check that box off the list... 

 

Live and learn as you say !  I'll report back this weekend. 

 

Thanks !

 

Olivier



#9
Ron Alan

Ron Alan

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 3,732 posts
  • Location:Northern CA
  • Car Year:1995

Ahhh...we have met and emailed recently! I run NASA Norcal SM. My contact is in one of those emails! 


Ron

RAmotorsports

 

Donor - Made PayPal donation Bona fide - A bonafide Spec Miata driver

#10
chris haldeman

chris haldeman

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 928 posts
  • Location:Mckinney
  • Region:texas
  • Car Year:1999
  • Car Number:73
Potential causes......
Egr stuck open
Idle air control valve failure
Throttle body leak
Intake gasket leak
Not many other items that could cause this. A few points I want you and anybody else reading this thread to know....

Idle afr means nothing if on track is good
Also remember the o2 sensor reads burnt fuel so what that means if there is a slight miss fire at idle it will read lean but actually be fuel rich.
X-factorracing.com
3 podium finishes
2 2013 NASA nats
1 2013 Scca runoffs
Donor - Made PayPal donation Bona fide - A bonafide Spec Miata driver BFG Supertour Winner - Circuit of the Americas Winner - Majors Winner - Make it Rain - Made Paypal donation of $100+ We have a Winnah! - Won their 1st race... Congratulations!

#11
Steve Scheifler

Steve Scheifler

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 2,816 posts
The consistent/repeatable (as told) drift to lean makes me think it’s likely not just a vacuum leak, but yeah, maybe something weird with the iac valve. But it would require a large amount of extra air getting through which should be audible both in the sucking sound and in the idle speed.

But here’s another thing to monitor. On the NBs you can connect to the OBD port and watch major parameters including timing, throttle position and fuel trims. It would be interesting to watch those as the AFR is steadily increasing to see if the ECU is attempting to correct for it. If the O2 sensor is working it will be continuously on the lean side and the ECU should be adding fuel in the short-term trim corrections. Or there may be another parameter specifically for responding to the O2 reading (rather than the trims). Maybe someone else will know for certain or look up the PID list.
Instigator - Made a topic or post that inspired other Broken record - You are starting to sound like a broken record.

#12
HeavyRightFooT

HeavyRightFooT

    Member

  • Members
  • PipPipPip
  • 11 posts
  • Location:Bay Area
  • Region:California
  • Car Year:1999
  • Car Number:77
Thanks for all the inputs ! Much appreciated !

So here is where I stand at this point: some progress but no fix yet.

To summarize what I tried so far:
-having the complete old engine that came out of the car at disposition, which ran ok before I pulled it out, I swapped egr, throttle body including iac and tps, coolant temp sensor, one by one, no difference.
- At the wideband, one by one, swapped sensor, gauge power wiring harness and o2 sensor harness with a spare identical set and nothing. Per previous post checked exhaust leaks at sensor, went ahead and plug first bung I had welded and installed a new one being extra thorough, nothing. Smoke tested it for vacuum leaks, nothing.

-looking at the posts on this thread again and considering that the engine sounds fairly fine, I though maybe there is an issue with the power or grounds I am using for the wideband (connector for cigarette lighter, feeding both the wideband and my data system). I ran the power and ground directly to the battery to see, and it worked fine hovering right around 14.7 !!

So I wired the wideband to a different switched power under the dash thinking that might do it, and ran a separate ground to the frame of the dashboard.
This time it drifted lean again after a while, but settled around mid 16afr instead of maxing the gauge again like before...

Any idea what might be going on ? Why would it run fine directly from the battery but not from a 12v switched power source ? Every other gauge or accessory is running perfectly, including when powered from the same locations as mentioned before. (Cigarette lighter connector, radio connector, power side view mirrors connector).

Olivier

#13
Dave D.

Dave D.

    Member

  • Members
  • PipPipPip
  • 199 posts
  • Location:New Jersey
  • Region:NASA Northeast
  • Car Year:2002
  • Car Number:14

If you did not look at the gauge, does the engine run any different from another car that would give you an indication that something was wrong? How does it run in general? If it was that lean while driving, it would fall on it's face on accell. I think you are seeing normal readings at no load idle. It will run rich after start as the ecu gave a decent shot to the injectors to get started, then will scale back at a warm idle. I don't remember if the 99 has PIDs for fuel trims(I'm at home, not at the shop), but look for them plus injector Ms times to see if the ECU wants the motor to lean or rich. Drive with the scanner on and watch the values on accell and coast down to see the change.

   A vacuum leak large enough to make the idle very lean will almost always give a very rough idle(shaking) and a CEL,however sometimes you can have a small leak with no symptoms other than a CEL(small vac hose off). The fact that you have swapped many parts and changed gauges and senders with no difference and no outward driving symptoms should be and indication that those might be normal readings. I only look at AFRs at WOT on track or dyno.



#14
Steve Scheifler

Steve Scheifler

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 2,816 posts
Dave, you echo sone if my own comments above, but I think the biggest question still unanswered is how it drives under light load. I think there is undue concern about risking damage, but it may just not be practical to drive at all due to noise, neighbors, liability risks.

HRF, if your wiring test is consistently repeatable it pretty much proves that what you first suspected was in fact broadly correct, you are getting false wideband readings. Unlike narrow band sensors, widebands have a preheater which draws pretty significant current. In theory a typical cigarette lighter socket is wired and fused more than adequately (modern “accessory” sockets may be less so if never intended to have a lighter in them). However, you are already running some stuff off that circuit, and it’s a 20+ year old car and wiring. Question: Did you by chance reduce the fuse size to one more appropriate for the low current required for the data system? I can envision a situation where a marginal fuse, whether too small or making poor contact, gets hot enough to increase its own resistance but doesn’t quite blow. Pull it out and put a new 20 amp one in. Also, the contact surfaces inside an old lighter socket may have significant oxidation preventing a good connection. For one reason or another I’ve seen countless issues with old OEM sockets. And not all widebands are created equal, the key difference being the controller not the sensor and how well they deal with voltage. Some are really picky, and some are notorious for blowing sensors because they can’t deal with voltage fluctuations or spikes, particularly some of the newer faster sensors required for getting on line extremely quickly during engine warmup. (I’ve commented on this before, but in short the attributes that are attractive for smog certification of a street car are not necessarily a benefit to us, particularly if it results in frequent failures because aftermarket controllers aren’t as sophisticated as the OEM ones.).

Anyway, broadly speaking I think you have your answer, faulty AFR readings. For various reasons including those noted above you should never run a wideband off an accessory socket as a permanent solution anyway, so hardwire it then clean things up for your data system and consider hardwiring it as well if you don’t need it to be portable.
Instigator - Made a topic or post that inspired other Broken record - You are starting to sound like a broken record.

#15
Ron Alan

Ron Alan

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 3,732 posts
  • Location:Northern CA
  • Car Year:1995

14-17 during warm up is normal on my cars.


Ron

RAmotorsports

 

Donor - Made PayPal donation Bona fide - A bonafide Spec Miata driver

#16
Steve Scheifler

Steve Scheifler

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 2,816 posts

14-17 during warm up is normal on my cars.


But to be clear, he says it always starts around 14 and drifts to off-scale lean and stays there. Warmup really should be richer, then move up to ~14.7. If yours is really leaner, ~17 during warmup it’s not doing something right or air is getting into the system.
Instigator - Made a topic or post that inspired other Broken record - You are starting to sound like a broken record.

#17
Ron Alan

Ron Alan

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 3,732 posts
  • Location:Northern CA
  • Car Year:1995

But to be clear, he says it always starts around 14 and drifts to off-scale lean and stays there. Warmup really should be richer, then move up to ~14.7. If yours is really leaner, ~17 during warmup it’s not doing something right or air is getting into the system.

Starts richer and drifts towards 16-17 at temp. Both my 99's. And yes...leaks are very possible :(

 

But my wideband is within .1-.2 at WOT between 5500-7000rpm to the dyno reading out the pipe.


Ron

RAmotorsports

 

Donor - Made PayPal donation Bona fide - A bonafide Spec Miata driver

#18
Steve Scheifler

Steve Scheifler

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 2,816 posts

Starts richer and drifts towards 16-17 at temp. Both my 99's. And yes...leaks are very possible :(

But my wideband is within .1-.2 at WOT between 5500-7000rpm to the dyno reading out the pipe.


WOT in the powerband is what matters unless it idles or cruises way rich risking fouled plugs and washed down cylinders.
  • Ron Alan likes this
Instigator - Made a topic or post that inspired other Broken record - You are starting to sound like a broken record.

#19
HeavyRightFooT

HeavyRightFooT

    Member

  • Members
  • PipPipPip
  • 11 posts
  • Location:Bay Area
  • Region:California
  • Car Year:1999
  • Car Number:77

Thanks for all the help !

 

Car went on the dyno and runs fine ! ... my OCD will just have to get to terms with funky idle AFR readings then.

 

Cheers

 

Olivier


  • chris haldeman likes this




1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users