Oxygen sensor graph - trying to solve rough idle
#47
Racer
Thread Starter
From a private message I received, it's believed that the PST2 does support
Mode 6. However, I haven't been able to confirm this since I myself don't
have access to a PST2 scan tool.
The question then is if it is able to decode the hex codes into English. Turns
out this is something that can't be taken for granted.
I did find that in the ATS EScan user manual that when looking at Mode 6 data,
the monitors must be complete for related Mode 6 info to be accurate. Also, if
DTCs have been recently cleared, there will be no Mode 6 data available.
ATS has a new Mode 6 decoder that can be use to translate Mode 6 hex codes
to English:
http://www.automotivetestsolutions.com/dtc-mode-6.html
I have no affiliation with ATS, just like what they're doing.
-bruce7
Mode 6. However, I haven't been able to confirm this since I myself don't
have access to a PST2 scan tool.
The question then is if it is able to decode the hex codes into English. Turns
out this is something that can't be taken for granted.
I did find that in the ATS EScan user manual that when looking at Mode 6 data,
the monitors must be complete for related Mode 6 info to be accurate. Also, if
DTCs have been recently cleared, there will be no Mode 6 data available.
ATS has a new Mode 6 decoder that can be use to translate Mode 6 hex codes
to English:
http://www.automotivetestsolutions.com/dtc-mode-6.html
I have no affiliation with ATS, just like what they're doing.
-bruce7
i tried mode 6 with the pst2 and as you can see it is not supported. I took at look at the dme manual and there is no mention of mode 6, it talks about modes but stops at mode 5. I am guessing that the pst2 does support mode 6 but that the 5.2 dme does not. That would have save great if
it did work.
#48
Racer
Thread Starter
So if I follow you, the idea is to determine which cylinder is misfiring. And your
approach is to try and replicate what the DME does to calculate a misfire.
But is that really what the DME does? I thought what the DME did was to
calculate the angular velocity (rate of speed change) between successive cylinders.
Then it must apply a correction factor for whether the engine is accelerating or decelerating.
I don't have any reference to point to for how the DME does misfire detection
so this is just speculation. I did find though a patent abstract for misfire
detection that is interesting reading and I'm sure that there is a lot more
out there:
http://www.google.com/patents/US6023651
Other approaches that would be easier would include: spark line firing
waveforms, O2 sensor waveforms, pressure transducer waveforms,
and Mode 6 diagnostics.
-bruce7
approach is to try and replicate what the DME does to calculate a misfire.
But is that really what the DME does? I thought what the DME did was to
calculate the angular velocity (rate of speed change) between successive cylinders.
Then it must apply a correction factor for whether the engine is accelerating or decelerating.
I don't have any reference to point to for how the DME does misfire detection
so this is just speculation. I did find though a patent abstract for misfire
detection that is interesting reading and I'm sure that there is a lot more
out there:
http://www.google.com/patents/US6023651
Other approaches that would be easier would include: spark line firing
waveforms, O2 sensor waveforms, pressure transducer waveforms,
and Mode 6 diagnostics.
-bruce7
take a look at the attached description of misfire monitoring. The dme is looking for differences in angular acceleration (i.e. the change in angular velocity) for each combustion event. What i was planning to do is similar, instead of measuring or calculating angular acceleration i was planning to measure the time for each segment. Where a segment is 120 degrees of a crank revolution or every time a cylinder is at TDC. I know that the time for each segment is directly related to rpm and thought that an average value of each segment over a period of time would hopefully show which segment is slower that the others (i.e. the misfiring cylinder).
After looking at the links you posted i think it would be easier to use a pressure transducer in the exhaust and compare that data with the hall sensor. THere would be a lot less data to sort through.
Regarding your comment on testing the DMF, i did rock it back and forth and didn't think there was too much play. Also, i can see how a bad DMF could cause a rough idle and throw a fault code but i don't think it would cause the "phut" sound in the exhaust. I think that is a combustion misfire, the fuel hasn't ignited because there is too much or too little air.
looking forward to your comments.
#49
bruce,
i tried mode 6 with the pst2 and as you can see it is not supported. I took at look at the dme manual and there is no mention of mode 6, it talks about modes but stops at mode 5. I am guessing that the pst2 does support mode 6 but that the 5.2 dme does not. That would have save great if
it did work.
i tried mode 6 with the pst2 and as you can see it is not supported. I took at look at the dme manual and there is no mention of mode 6, it talks about modes but stops at mode 5. I am guessing that the pst2 does support mode 6 but that the 5.2 dme does not. That would have save great if
it did work.
Damn! That's a bummer. Now we know. Thanks for digging into that.
-bruce7
#50
bruce,
take a look at the attached description of misfire monitoring. The dme is looking for differences in angular acceleration (i.e. the change in angular velocity) for each combustion event. What i was planning to do is similar, instead of measuring or calculating angular acceleration i was planning to measure the time for each segment. Where a segment is 120 degrees of a crank revolution or every time a cylinder is at TDC. I know that the time for each segment is directly related to rpm and thought that an average value of each segment over a period of time would hopefully show which segment is slower that the others (i.e. the misfiring cylinder).
After looking at the links you posted i think it would be easier to use a pressure transducer in the exhaust and compare that data with the hall sensor. THere would be a lot less data to sort through.
Regarding your comment on testing the DMF, i did rock it back and forth and didn't think there was too much play. Also, i can see how a bad DMF could cause a rough idle and throw a fault code but i don't think it would cause the "phut" sound in the exhaust. I think that is a combustion misfire, the fuel hasn't ignited because there is too much or too little air.
looking forward to your comments.
take a look at the attached description of misfire monitoring. The dme is looking for differences in angular acceleration (i.e. the change in angular velocity) for each combustion event. What i was planning to do is similar, instead of measuring or calculating angular acceleration i was planning to measure the time for each segment. Where a segment is 120 degrees of a crank revolution or every time a cylinder is at TDC. I know that the time for each segment is directly related to rpm and thought that an average value of each segment over a period of time would hopefully show which segment is slower that the others (i.e. the misfiring cylinder).
After looking at the links you posted i think it would be easier to use a pressure transducer in the exhaust and compare that data with the hall sensor. THere would be a lot less data to sort through.
Regarding your comment on testing the DMF, i did rock it back and forth and didn't think there was too much play. Also, i can see how a bad DMF could cause a rough idle and throw a fault code but i don't think it would cause the "phut" sound in the exhaust. I think that is a combustion misfire, the fuel hasn't ignited because there is too much or too little air.
looking forward to your comments.
stuff. I agree that the approach using pressure transducers would be a
good next step. It's interesting that in the 1993 SAE paper that JasonAndreas
linked to they mention the use of pressure transducers as an option for
monitoring misfires to comply with OBDII but that the cost was too high.
The costs for putting together a pressure transducer diagnostic system for
shops is not too bad but it is still a bit high for most DIY owners. I would like
to get into it when I can. I haven't done a top-end yet and was thinking
that it would be great to do the before/after analysis with pressure transducers
when that time comes.
#51
Racer
Thread Starter
just a quick update. I got a some help on Excel and was able to manipulate the data. There are 300,000 samples data for about 70 seconds of running. To make a long story short, the highlighted section in the Excel image below shows a summary for each cylinder.
This test did not find the smoking gun i was looking for but does show that cylinder 1 is on average a little slower than the rest.
This test did not find the smoking gun i was looking for but does show that cylinder 1 is on average a little slower than the rest.
#52
As John Thornton illustrated in the pressure transducer case study video
back in post #42, it's good to have as many tests as possible to confirm
a diagnosis.
You could do a cranking compression test and see if #1 is low relative to
the other cylinders.
You could do a spark firing line test and see if there is turbulence in
#1 relative to the other cylinders.
You could do pressure transducer tests to see if there is a mechanical
issue with cylinder #1 relative to the other cylinders.
If there is any kind of unusual noise you could do a NVH test.
I don't have an example video of this type of test, but PicoScope have a
test kit similar to the transducer test kit that can be used to characterize
noise, vibration and harshness into a waveform for analysis.
Keep at it. I'm sure you'll find whatever it is.
-bruce7
#54
Racer
Thread Starter
the fuel injectors have been tested and cleaned (twice) and swapped from bank to bank (twice) and have not seen a difference in trim values or "phut" sound. the only correlation i have noticed is the intake rockers with lifters. it's hard to believe that new parts are bad. i would guess that something with the rebuild/valve job is root cause. other than a small rough idle, the engine is silky smooth and sings...
i think my next step is pressure measurement of the intake and exhaust. trying to find a reasonably priced transducer.
#55
Racer
Thread Starter
#56
problem it would cause a compression issue and result in turbulence in
the cylinder and the spark voltage waveform will be erratic or noisy.
#57
Primary waveform
Do either of you have a primary ignition waveform to compare?
This is the 12V signal coming from the computer to one of the coils.
I need to figure out how to hook up to the secondary to capture that as well - any tips?
BTB
Secondary (12V) waveform between computer and coil.
This is the 12V signal coming from the computer to one of the coils.
I need to figure out how to hook up to the secondary to capture that as well - any tips?
BTB
Secondary (12V) waveform between computer and coil.
#58
Do either of you have a primary ignition waveform to compare?
This is the 12V signal coming from the computer to one of the coils.
I need to figure out how to hook up to the secondary to capture that as well - any tips?
BTB
Secondary (12V) waveform between computer and coil.
This is the 12V signal coming from the computer to one of the coils.
I need to figure out how to hook up to the secondary to capture that as well - any tips?
BTB
Secondary (12V) waveform between computer and coil.
https://rennlist.com/forums/993-foru...lug-wires.html
Let me know if you have any questions. I'll try to help.
-bruce7
#59
There is this DIY I did. The secondary ignition stuff is towards the end.
https://rennlist.com/forums/993-foru...lug-wires.html
Let me know if you have any questions. I'll try to help.
-bruce7
https://rennlist.com/forums/993-foru...lug-wires.html
Let me know if you have any questions. I'll try to help.
-bruce7
It may take me a week to digest it.
What I am finding is that my rough idle -almost- completely disappears when the engine has some load on it. For example, when the engine is warm enough to run the oil cooling fan, or if I switch on my high beam headlights. My theory is that the additional electrical load results in additional mechanical load from the alternator. And, like others, the idle gets better when the AC is switched on (but this also boosts the idle, whereas the oil cooler fan and high beam headlights don't seem to).
BTB