OBD-II data logs (2 Viewers)

[landtank]

I had some time today and have been wanting to post up some data logs for people to review. In the past there have been a lot of talk about how our ECUs deal with controlling the fuel delivery. Hopefully these will allow people to see what is going on inside the ECU and how it adapts to changes.

Since we can now attach pdf files I think this is the best way to go. I have 6 data logs, 3 from different stock MAF sensors and then 1 from my MAF w/vacuum, 1 wo/vacuum and then a MAF that has been sleeved w/vacuum.

I'll start by attaching the 3 stock files. People can print them out and then reference them in this thread. I broke them up in groups of 10 line segments to help identify where abouts on the log you are talking about.

 

[94SRUNNER]

Thanks Rick! Love the data!

 

[landtank]

first, I don't know what happened to the pdfs but there are only 7 pages of data in any one file. If you print the whole document you get 14 pages, 7 are blank.

I should point out here that the way the software records the data it's one item at a time, from left to right. So while two lines might have the same Thr Pos value, the LTFT% value might be different because the Thr Pos moved since it was recorded on that line.

And all the logs were driven over the same route to help with comparisons

Since the only thing that has changed is the sensors them selves any differences between the data logs is actually differences between the sensors.

Some basics from my understanding. The raw signal from the MAF sensor is used to calculate a base injector timing. The ECU uses that initaly to start and run the engine. Then when the engine reaches a certain temp it begins to incorporate the O2 signal and "trim" the injector timing to get the proper fuel mixture.

Initially it adjusts the STFT. But as time elapses it will begin to migrate that portion of the trim that is always there to the LTFT value. This period is referred to as closed loop. Feedback from the O2 sensors are used to adjust the injector timing

You can see this by looking at identical Thr Pos's. The LTFT will be constant while the STFT will fluctuate.

By comparing the three logs you will notice that at particular Thr Pos's you don't always see the same LTFT. The ECU seems to be mapping out the individual sensors and how they uniquely see the air flow.

IF you look at the MAF A log on page 2, lines 27-30 you will notice what I believe to be open loop. This refers to where the O2 sensors are not referenced and no adjustment is made to the STFT value.

 

[landtank]

Well I had a PC crash and lost all the PDF files of the logs. Then went on a week's vaction so I just had time to remake the PDF files from the original data logs.

These are the last three files form the run and they are my housing with the vacum line attached, one with it removed and one modified with a sleeve.

First you can see how the ECU reacts to the change is the air flow out put of teh MAF sensor and also the increase in fuel pressure when the vacuum line is removed.

Lastly, the sleeved unit with the vacuum line still attached seems to be the one that most closely follows the data of the stock units.

 

[turbocruiser]

Rick, are you selling the sleeves yet?

 

[landtank]

Rick, are you selling the sleeves yet?

I sent off a drawing that will hopefully be the final design of the sleeves yesterday.

I need to confirm an at home install and then will make them available.

I also want to do another run of logs just between a single stock MAF and the sleeved unit giving it more run time between changes to ensure that the log is accurate to an ECU that has fully learned the housing. I kind of rushed these a little doing 6 changes in a single afternoon. I was trying to get as close a sample as I could without the enviroment changing too much.

But these are good for understanding what the ECU is doing and how it is adapting to changes in the system. And with the exception of WOT, the system seems to be adjusting STFT% to maintain AFRs.

 

[turbocruiser]

What's happening at WOT?

 

[landtank]

What's happening at WOT?

open loop, no STFT adjustment. Look for the columns were the TPS reports a position of 70% and you'll see that there is nothing being logged in the STFT column.

Up until then the STFT is being adjusted.

 

[turbocruiser]

open loop, no STFT adjustment. Look for the columns were the TPS reports a position of 70% and you'll see that there is nothing being logged in the STFT column.

Up until then the STFT is being adjusted.

Sorry, what I meant was what are the AFR's at WOT. Thanks.

 

[landtank]

Sorry, what I meant was what are the AFR's at WOT. Thanks.

I don't have a wideband setup on the truck so I don't know.

 

[turbocruiser]

I don't have a wideband setup on the truck so I don't know.

What!!! Man, you gotta get a wideband working; it doesn't make much sense to develop the sleeve without wideband AFR's, especially at wide open throttle. Its only an afternoon of working on the exhaust tube and about $300 and it will indeed help. Anything less is just a guess at getting the sleeve setup right.

 

[landtank]

I have an innovate wideband already. Nobody, regardless of what setup was used has had anything other than a rich condition at WOT. What is it you expect me to be able to do?

 

[turbocruiser]

I have an innovate wideband already. Nobody, regardless of what setup was used has had anything other than a rich condition at WOT. What is it you expect me to be able to do?

Well I would expect you to use something other than just the OBD-II data to setup the sleeve correctly. Anything less is just a guess. I'm not saying or suggesting we all won't still run rich at WOT, that's not the point, the point is you need your AFR's at all ranges anyway and I was just wondering how the sleeve skewed the WOT AFR. If you've got a good wideband stop screwing around already and install it.

 

[landtank]

during the times where there is a STFT value anything I do with the bore of the housing will be negated by the ECU doing whatever it wants to do in order to please itself.

If you go look at all six logs you'll see that regardless of which MAF was used or even if the vacuum line was attached or not the LTFT% is always 4.68.

So the only way to alter the WOT AFRs is to lower the output of the MAF sensor or reduce the fuel rail pressure during those time.

Enlarging the bore of the MAF housing will lower the output during WOT but will also do it during idle producing high LTFT%.

Unplugging the vacuum line from the fuel pressure regulator puts the system into a inversely proportional differential verses a constant one. For those who might not understand this, it basically means that as the air flow increases the fuel pressure differential decreases by exactly the same amount. So at idle the fuel differential is higher than if it was a constant situation but gradually gets closer to being the same and converging at atmospheric pressure. Beyond that the inversely proportional setup has a lower differential.

So these two situations can compliment each other. The inverse proportional differential offsets the lower airflow during idle and then converges at atmosphere where the two work together to lower WOT AFRs.

Now Shawn, you and quite a few others have already used multiple WB sensors to verify that this setup still produces a rich condition but not as rich as the stock setup.

But now everyone wants the vacuum line attached. Which means we now also have to shrink the size of the housing to get the air flow back up at idle. This will also increase the air flow reading during WOT. Now with a constant fuel differential it also will be increased during those times as well.

Increase fuel pressure and increased readings from the MAF sensor leads me to believe we'll be more rich than before.

I'll see what I can do about getting the wideband installed to do more comparisons. But I first want to make sure the ECU reacts similarly between the sleeved MAF and the stock one.

 

[turbocruiser]

during the times where there is a STFT value anything I do with the bore of the housing will be negated by the ECU doing whatever it wants to do in order to please itself.

If you go look at all six logs you'll see that regardless of which MAF was used or even if the vacuum line was attached or not the LTFT% is always 4.68.

So the only way to alter the WOT AFRs is to lower the output of the MAF sensor or reduce the fuel rail pressure during those time.

Enlarging the bore of the MAF housing will lower the output during WOT but will also do it during idle producing high LTFT%.

Unplugging the vacuum line from the fuel pressure regulator puts the system into a inversely proportional differential verses a constant one. For those who might not understand this, it basically means that as the air flow increases the fuel pressure differential decreases by exactly the same amount. So at idle the fuel differential is higher than if it was a constant situation but gradually gets closer to being the same and converging at atmospheric pressure. Beyond that the inversely proportional setup has a lower differential.

So these two situations can compliment each other. The inverse proportional differential offsets the lower airflow during idle and then converges at atmosphere where the two work together to lower WOT AFRs.

Now Shawn, you and quite a few others have already used multiple WB sensors to verify that this setup still produces a rich condition but not as rich as the stock setup.

But now everyone wants the vacuum line attached. Which means we now also have to shrink the size of the housing to get the air flow back up at idle. This will also increase the air flow reading during WOT. Now with a constant fuel differential it also will be increased during those times as well.

Increase fuel pressure and increased readings from the MAF sensor leads me to believe we'll be more rich than before.

I'll see what I can do about getting the wideband installed to do more comparisons. But I first want to make sure the ECU reacts similarly between the sleeved MAF and the stock one.

I think I'm missing something ... how would shrinking the size of the housing increase the airflow? I do agree that with the fuel pressure regulator referenced we will have higher fuel pressure at WOT and therefore we will have richer running at WOT. I just can't capture the shrinking size of the housing equaling an increase in airflow? I still say you need all the data that's available, not half of it and relying on narrowband O2 sensors is only going to give you what the computer thinks is happening ... not the net effect of everything together. When "we" were developing your LT MAF v1.0 you were all about all my wideband readings at all ranges of operation, why are you avoiding that important information now?

 

[sleeoffroad]

Rick, thanks for taking the data. It seems like the sleeve is getting things closer to stock.

I wanted to see what effect your sleeve has on things. You state

during the times where there is a STFT value anything I do with the bore of the housing will be negated by the ECU doing whatever it wants to do in order to please itself.

From the plot it seems like this is not quite true.

I plotted your data to get an idea of what is going on. It showed some interesting observations. Below is a graph of some of the LTFT plotted against sample number. Obviously the series are not synchronized with respect to each other, but you mentioned that you did the same course and the files contained roughly the same amount of sample numbers.

Yellow line is stock. - Seems like that is very small variation around 10%. That is what one expects.

Purple Line - LT MAF with Vac connected. It is pretty obvious that the LTFT is pegged at 19.5 for most of the run and it is obviously not doing it's job. This is the condition that we warned about all along. Why you are not getting a check engine light and why the value does not go over 19.5% I do not know. However >20% should trigger the check engine light.

Red - LT MAF with VAC disconnected - The LTFT is showing that for the majority of the run, the truck is running lean and it is adding fuel. It does not seem to be pegging out, but still has a much larger swing than a with the stock. Not sure why that is. Do you have any theory as to why the variation is so much more with the LT MAF than stock?

Blue - LT MAF with Sleave/Vac connected. Overall the trend on the LTFT is lower, but the variation is still much more than stock. Again, not sure why that is.

Just thought I would put this out there.

 

[sleeoffroad]

The second plot I made is of the calculated AFR values in LBS. I wanted to see if I can see any difference in the air flow measurements. From the plot below one can not determine much in terms of accuracy of the new MAF or higher resolutions. It would be interesting to plot the actual voltages against each other. Again one should not compare series to series based on the X axis. They are not totally synchronized other than assuming they are roughly the same run.

 

[landtank]

First, all the MAFs (that includes the stock ones) call for a lean condition initially. It's just the way they behave at sea level in cold weather from what I've seen. People at higher elevations in warmer weather have reported -LTFT% which would mean they are calling out for a rich condition initially.

Shrinking the size of the chamber doesn't increase the air flow it increases the output signal of the sensor. So the total airflow is the same but since it's traveling through a smaller opening it's traveling faster (for lack of a better word) and the sensor sees more air at any given time. The sensor only sees a sample of the air and then extrapolates what the total is. So increasing the sample increases the total.

In your plots you only tracked LTFT which isn't that accurate on these logs. The natural tendency is for the ECU to migrate the STFT% to the LTFT% until the STFT fluctuates around 0%. This definitely happens for idle as I've sat through a long enough period to see it. At other times it's been too hard to get into that long of a constant period to see it happen to know for sure.

If you look through the logs you will see that this hasn't happened on a few of them at idle. I rushed the process and the ECU didn't always have time to adjust. If you look at the Stock A log you can see this migration on the last line of the initial idle segment. If it had idled longer you would have seen the STFT fluctuating around 0 with the higher LTFT%. Looking at the Stock C log, the LTFT% is about half that of the rest but the STFT is also greater. But when you add the LTFT and the STFT together the total FT% is about equal. That also goes for the Sleeved MAF housing, not enough time elapsed for the ECU to fully migrate the FT%s.

Knowing how much the vacuum line altered the FT% I calculated the change in size and had the sleeve made. I then compared idle voltages from the sleeved MAF to a stock one which showed about .05v difference, to the best of my ability as the signal fluctuates quite a bit.

My next step is to confirm similar FT% between the sleeved MAF and the Stock ones. This is imperative to avoid CEL faults and to keep the values within designed limits.

Then finally a Wide Band.

Right now I'm having another sleeve made that would be more of a finished product and then will do head to head comparison with a single stock MAF.

I was asked by someone to post this info and so I did. This is still an ongoing process and not meant to be anything more than an insight into how the ECU behaves. It's not meant to be a sales tool or validate anything. If it comes across as such it can be deleted.

 

[sleeoffroad]

In your plots you only tracked LTFT which isn't that accurate on these logs. The natural tendency is for the ECU to migrate the STFT% to the LTFT% until the STFT fluctuates around 0%. This definitely happens for idle as I've sat through a long enough period to see it. At other times it's been too hard to get into that long of a constant period to see it happen to know for sure.

I agree, it is possible that a reason for the wide swings might be that ECU (LTFT) have not "learned" enough. It would be interesting to see the data for MAF with the sleeve, after several weeks of driving, over the same run to see if the fluctuations have settled down.

 

[landtank]

I agree, it is possible that a reason for the wide swings might be that ECU (LTFT) have not "learned" enough. It would be interesting to see the data for MAF with the sleeve, after several weeks of driving, over the same run to see if the fluctuations have settled down.

When I did the 6 logs I started with the LT wo/vac then the LT w/vac, then the sleeved one and then the three stock units.

Not a very smooth transition as the LT w/vac was the furthest away from the stock units and since the stock units came after the sleeved one and then in concession that probably helped the learning some for those.

The biggest disadvantage I have is that the truck sits for weeks and then I drive it only on the weekend and not every weekend so it's hard to get a good stable log. The weather around here changes drastically so each start is yet another long cycle to get there. And then throughout the day it can change 30*.