Reading MLV Data Logs

[whittlebeast]

I would like to try to put out a thread that will help people learn how to read data logs. Over the last 15 years of tuning EFI, we have learned patterns that indicate problems. Here are a few of the tricks.

First, I set up the top section of the Log Viewer to look at the three most basic things, RPM, TPS and MAP. The second section has the remaining sensors CLT, MAT and Battery Voltage, In the third trace I plot MAPxRPM (an optional calculated field if you have a registered version of MLV) and Duty Cycle. In the forth trace I plot MAP and PW. In the last trace, I plot Lost Sync Count, Lost Sync Reason and VE.

http://www.nbs-stl.com/tuning/Sample%20 ... 0Check.png

This happens to be a well behaved motor doing a warm start. I will go thru what I am looking for in each section.

RPM, MAP and TPS.
I want all three of these to be relatively smooth. I never want to see the RPM to drop to 0 or spike real high. This is almost always an indication that the code somehow lost the location of the crank and reported the wrong answer. Spikes in the TPS and MAP are almost always a wiring issue.

CLT, MAT and Battery Voltage
I want to see smooth changes in CLT and MAT. No changes as the motor cranks. Changes in temperatures wile engine cranking is almost always caused by sensor grounding issues. A drop in battery voltage wile cranking is normal and makes it easy to see when the starter was under load. Once the motor starts, the voltage should rise to 13.5 to 14.2 volts and hold steady.

MAPxRPM and Duty Cycle is very similar to Pulse Width and MAP. I want to see both of these traces to run generally parallel. Big spikes in Duty Cycle or Pulse Width are almost always too much Acceleration Enrichment. Constant spikes are too sensitive AE triggers. I like to see about a 10-15% increase in fueling when AE set in. This is all a rule of thumb but it really helps spot issues.

Lost Sync Error and Lost Sync Count
The simple answer is if the ECU is loosing where the crank is, the motor will never run right. See the orange oval where the ECU lost the crank location and the yellow circle where the engine RPM dropped to 0. The battery voltage is an indication that the starter was still spinning the motor. Big trouble.

VE
I simply am looking for a smooth curve. If the top two are running flat like on the highway or at idle, this trace should be fairly flat. Spikes are an indication of issues in the VE table,

Here is an example what a motor with issues may look like.

http://www.nbs-stl.com/tuning/Sensor%20 ... Issues.png

Feel free to ask questions and have fun tuning.

Andy

[whittlebeast]

Here is a motor at idle. In this calibration, the Acceleration Enrichment was set up as MAP based. The hunting idle was triggering AE. Look at the PW jump from 2.5 ms to 4.5 ms every 1.75 sec. The motor was floating from 850 to 1450 and back.

http://www.nbs-stl.com/tuning/Hunting%20Idle.png

If you look at about 10 sec, the motor also had an RPM drop out due to a Lost Sync Error.

It is amazing what you can see when you simply take the time to look.

Andy

[whittlebeast]

Here is an example of a hunting idle caused be having a big change in VE in the idle area. Flatten the VE and the hunting will clear right up. Notice that the timing has been added and is flat enough that it is unlikely the cause.

http://www.nbs-stl.com/tuning/Hunting%2 ... 0Based.png

Look how the AFR follows right along, just as you would expect with this huge change of fueling. I find it interesting that the AFR is about 1 sec delayed at these engine speeds and exhaust flow rates.

Andy

[whittlebeast]

Here is the same hunting idle in scatter plot view with the AFR synced for the time delay.

http://www.nbs-stl.com/tuning/Hunting%2 ... Synced.png

Andy

[R100RT]

Hi Andy, thanks for taking the time to present this information - really puts the icing on the cake for our favourite control system and the great people associated either directly or via the forums.
Regards,
Lorne

[gjls]

Thanks for posting it all makes more sense now

Sent from my SM-N910V using Tapatalk

[turbo conversion]

Good Stuff if I weren't so technically challenged I would do a video on AE

David

[whittlebeast]

Here is what AE looks like on my ITB Honda race motor. It is the bump up in the red fueling traces. Notice the matching MAPdot and TPSdot.

http://www.nbs-stl.com/CRX/AE%20Testing%202.png

Andy

[whittlebeast]

This is what a few neutral throttle stabs looks like on that motor.

http://www.nbs-stl.com/CRX/CRX%20Thrott ... sponse.png

Check out the RPM Per Sec in green at the bottom.

Andy

[whittlebeast]

Here is a trace of a very well tuned motor hitting the Target AFRs. See the bottom of this trace. This motor happens to be running in open loop fueling.

http://www.nbs-stl.com/MLVDemo/325%20Hi ... Target.png

Andy

[whittlebeast]

This is the same plot as I posted a few posts back but what I did was invert the AFR so that rich is red and lean is blue. That way the colors are more similar to large VEs are red and small VEs are Blue. The data is still corrected (Offset) to account for the O2 delay. Here is what shows up in the same plot.

http://www.nbs-stl.com/tuning/Hunting%2 ... %20AFR.png

Andy

[SwedCharger-67]

Andy, I like what you have done here. I'm going to set up my logger the same way and compare my graphs to yours, maybe I spot something new that I'm not aware of. My engine is running fine, but not perfect so I still have some finetuning to do.
Thanks for the ideas,
Martin

[whittlebeast]

The real fun is when you start plotting MAPxRPM vs DC and MAPxRPM vs Spark Advance.

On most typical motors, MAPxRPM is essentially air flow being processed thru the motor.

Andy

[SwedCharger-67]

Thanks Andy, that opens up a new field to me, will be interesting to explore further!

[whittlebeast]

I just looked at the data coming of my motor and the AFR is delayed about .25 sec. I happen to have fairly long tube headers.

[rukavina]

was trying to play with this and I cant figure out how to formulate mapxrpm in the logs. I attempted to do a custom calculated field and FAILED!

[rukavina]

nevermind..... found it

[Rebellpntball]

I would like to try to put out a thread that will help people how to read data logs. Over the last 15 years of tuning EFI, we have learned patterns that indicate problems. Here are a few of the tricks.

First, I set up the top section of the Log Viewer to look at the three most basic things, RPM, TPS and MAP. The second section has the remaining sensors CLT, MAT and Battery Voltage, In the third trace I plot MAPxRPM (an optional calculated field if you have a registered version of MLV) and Duty Cycle. In the forth trace I plot MAP and PW. In the last trace, I plot Lost Sync Count, Lost Sync Reason and VE.

http://www.nbs-stl.com/tuning/Sample%20 ... 0Check.png

This happens to be a well behaved motor doing a warm start. I will go thru what I am looking for in each section.

RPM, MAP and TPS.
I want all three of these to be relatively smooth. I never want to see the RPM to drop to 0 or spike real high. This is almost always an indication that the code somehow lost the location of the crank and reported the wrong answer. Spikes in the TPS and MAP are almost always a wiring issue.

CLT, MAT and Battery Voltage
I want to see smooth changes in CLT and MAT. No changes as the motor cranks. Changes in temperatures wile engine cranking is almost always caused by sensor grounding issues. A drop in battery voltage wile cranking is normal and makes it easy to see when the starter was under load. Once the motor starts, the voltage should rise to 13.5 to 14.2 volts and hold steady.

MAPxRPM and Duty Cycle is very similar to Pulse Width and MAP. I want to see both of these traces to run generally parallel. Big spikes in Duty Cycle or Pulse Width are almost always too much Acceleration Enrichment. Constant spikes are too sensitive AE triggers. I like to see about a 10-15% increase in fueling when AE set in. This is all a rule of thumb but it really helps spot issues.

Lost Sync Error and Lost Sync Count
The simple answer is if the ECU is loosing where the crank is, the motor will never run right. See the orange oval where the ECU lost the crank location and the yellow circle where the engine RPM dropped to 0. The battery voltage is an indication that the starter was still spinning the motor. Big trouble.

VE
I simply am looking for a smooth curve. If the top two are running flat like on the highway or at idle, this trace should be fairly flat. Spikes are an indication of issues in the VE table,

Here is an example what a motor with issues may look like.

http://www.nbs-stl.com/tuning/Sensor%20 ... Issues.png

Feel free to ask questions and have fun tuning.

Andy

Super cool thank you!

Sent from my Nexus 7 using Tapatalk

[whittlebeast]

I got this log from a thread here that was great at showing what the motor was doing on the fuel side of things. I scatter plot RPM, MAP and EGO Correction. This is what I get.

http://www.nbs-stl.com/tuning/EgoControlResults.png

In the lower left area is a red box where the motor idles. You can see on the right where I locked the Ego Correction limits to a range of 90 - 110%. Notice that the dots in this idle box are clearly red or averaging around 110% indicating the tune is lean in this area. Also notice the blue box where the dots are clearly a blue pattern. The tune is rich in this area. Some parts of all tunes have to be hand tuned in some of these areas.

Next, I will show some of the trick you can do in MLV HD to make all of this easier to tie all of this together.

Have fun tuning

[whittlebeast]

MVL HD brings the ability to roll this data up onto buckets that match the way you have your tune set up.

http://www.nbs-stl.com/tuning/EGOContro ... togram.png

Andy