light load misfire

[gslender]

email me a log of the problem and tell me exactly where in the log it is missing. gslender at gmail

[gslender]

Just looking at the log and I'm initially very concerned.

1) AFR is mostly see-sawing between 13 and 15 - odd behaviour?

2) MAP is very odd and flat... not what I'd expect either... max is 30kPa and min 15 - WTF?

3) PW is 1.0 - !! 1.0 ms... that can't be right... but yup it is, max logged PW is 2.0 - so I can only assume this old firmware is broken in logging the correct PW

4) GVE rocks between 57 and 58.... that's it...

This log is bonkers and I can't make sense of much of what I'm seeing.

The tune MSQ looks ok.

It is going lean, and the map sensor isn't really showing the correct true manifold pressure - it can't be only between 15 and 30 kPA... impossible. There is a period in the log where RPM starts at ~1000rpm and then engine revs climbs to nearly 3000rpm in 6 seconds and MAP goes spastic hitting 30kPA in very a few spots, then settles at 22kPa where it remains mostly flat for the entire rest of the steady rpm (where you begin to have misfires and RPM oscilations).

Something is sick - I'd start with the MAP and get it reporting correctly. Take more logs and study them.

G

[gslender]

Checked the log you emailed this morning - I'd say it is also concerning.

You have 106kPa as being the highest MAP value reached. Unlikely in a turbo, and the curise manifold pressure is mostly in the low 30s.... again very unlikely.

My guess is that you have either a faulty sensor, or the ECU isn't calibrated for the sensor you have. Either way you need to get that fixed as we are trying to fix something that is already got something else broken.

Also, the AFR is generally all over the place. Really messy and as if you have major tune issues (it needs to be way, way more stable) so maybe it is spark (as that would defintely show up that way) but if it is pulling strong then that is odd and I would say it isn't - also if it idles ok, then I'd be against it being spark.

My vote is that it is probably running too lean all over, or you have a vacuum leak that is 1) making it run lean and 2) mucking up the MAP signal (but normally that goes in the other way, with values centred around atmos 100kPa).

Keep looking and fix up that map value.

G

[gslender]

What would you suggest a normal Map reading at cruise would be? i would have thought 30ish is about right?

In your log I am seeing closer to 20kPa as being the MAP during cruise conditions (ie steady state TP > 5% with RPM siting in a steady band around 3000).

I would normally see around 40kPa as being the normal for that type of cruise event.

Elsewhere in the log you hit 23% throttle and the MAP jumps to 28kPa... that's not right. It should be at least 60-100 kPa with that much throttle open.

On average, your map value was 28kPa... so extremly low that I'm just skeptical.

Also, I doubt very much that your MAP only ever got higher than 106kPa as during those 100% TP periods the turbo should have been producing some kind of additional air pressure, even if just 2 psi above atmos, that would be at least 112kPa so I'd be surprised if you didn't do better than that, but you are right, it would be techinically possible to not hit a hint of boost, though I'm just surprised that you didn't - at all, ever, anywhere.

G

[jerrah]

As Gslender knows I had a bad MPXH6400 mapdaddy sensor in my Megasquirt. We swapped it to a stock MPX4250AP which resolved my MAP weirdness. My MPXH6400 had an incorrect reading for ambient pressure (engine off) even after calibration.

Do you have a boost gauge? What does your megasquirt log when it's in boost? What reading do you get with the engine off?

[gslender]

As Gslender knows I had a bad MPXH6400 mapdaddy sensor in my Megasquirt. We swapped it to a stock MPX4250AP which resolved my MAP weirdness. My MPXH6400 had an incorrect reading for ambient pressure (engine off) even after calibration.

Do you have a boost gauge? What does your megasquirt log when it's in boost? What reading do you get with the engine off?

Good point!

Confirm your map reading is 101 when ignition on, but engine not running/cranking. If not then your map sensor is bad or calibration is out.

G

[jerrah]

66kpa with the engine off is very wrong and you can't trust what it's reading.

Either you're out of calibration or it's broken. It's why the readings don't make sense in your log file.

Have you tried recalibrating it? What sensor model are you calibrating it for in Tunerstudio?

[gslender]

I briefly ran hi res firmware last night and the miss was imperceptible. which leads me to believe it could be an injector issue?

I hate troubleshooting problems when something else is wrong.

The tune doesn't look like its scaled correct for 66kpa and it is running lean almost everywhere except wot.

Fix the map sensor or at least see what its like when it is working correctly.

We couldn't get jarrahs to tune and run right until we fixed the map sensor. As soon as we did it ran beaut!

G

[gslender]

How certain are you that the o2 sensor is correct? At 11afr it should be stumbling again... Too rich. If you're saying it's running ok, then maybe it's also reporting incorrectly. When was the last free air calibration?

G

[NitroDann]

The guage is wrong and the car is too lean.

So back to my original post.

Configure everything properly and tune it.

Dann

[NitroDann]

theres no such thing as no calibration

Calibrate it or use a different WB02

Dann

[manga_blue]

Seen this? http://forum.aempower.com/forum/index.php?topic=27324.0

[jerrah]

My AEM gauge is offset to the analog as per the link posted above but doesn't affect the accuracy of the unit as long as you apply the correct offsets in tunerstudio. That said it still would need to be wired in correctly.

It would see strange that bad injectors/plugs or wires would run fine but just have a miss at 3000rpm.

Obvious things: Have you checked your base timing on the motor?

Could you warm up the car and take a log and post the current tune you're using? I've never found the 'autotune' in tunerstudio to work 100% without some thinking of the operator to ignore the faulty tuning. Make sure you go through and ensure all your sensors are calibrated including o2.

Having a working MAP sensor will help tremendously when you get it dyno'ed again.

[gslender]

Further diagnosis....

When free revving, can you determine which cylinder is causing the miss?

In which case, you can track it by swapping either spark plug, lead, or fuel injector to see if it follows the move.

G