An observation

[Gelf VXR]

After looking at the hi octane maps of most other LS2's I noticed that they were all the same, except the corvette tune that comes with EFIlive, its much more aggressive. So I copied over into my tune and did some data logging, AutoVE/BENs and knock.

To my suprise, I observed that the difference between the BEN values of both banks tightened right up accross the VE map, which led me to thinking can this difference be used an an indicator for MBT timing???

Stock LS2
http://i5.photobucket.com/albums/y164/andysa1966/MONAROSPARK.jpg
EFIlive Corvette
http://i5.photobucket.com/albums/y164/andysa1966/CORVETTESPARK.jpg

Timing can effect AFR, if some one could explain why, I'm not certain, I've tried searching for a paper on the web?, then when you think about a v8 it can be thought of as two four cylinder engines bolted together, with slighty offset timing MBT curves.

If and when both banks are at the peak of MBT curve +- 2 degrees, then BENS/AFR's will be very close. If one bank is running a little off the peak of the curve, then the BEN difference will be greater???


Instead of advancing until knock is detected, you could retard until BENS start to drift, then advance a few degrees??

[joecar]

Very interesting...

TOL: might be the reason why BEN's don't converge for everyone.

[Gelf VXR]

http://i5.photobucket.com/albums/y164/andysa1966/BENsvsMBTTiming.jpg

[joecar]

A picture is worth 1K words... and an Adobe document picture at that... cool...:cheers:

Please post it... :muahaha: ...I have to add it to my collection of notes. :D

[Blacky]

Timing can effect AFR, if some one could explain why, I'm not certain, I've tried searching for a paper on the web?,

Not sure if this helps...
http://www.innovatemotorsports.com/resources/myths.php

Regards
Paul

[Gelf VXR]

Looking at the ben difference in the two excel pics in the first post, the collumns for 4400 rpm remain unchanged by the increases in timing, probably gone to far (not far enough??) should try a reduction in timing for APCYL/RPM for those VE cells first?

A good starting point might be the low octane map, log, advance 2 degrees, log and compare, advance 2 degrees, log and compare etc

If and when the BENS converge your at MBT, if you advance further and they diverge????, you have gone to far.

My lack of understanding of why timing affects AFR leads me to question and need to clarification on this point, but If i am correct, when timing is not far enough advanced not all the oxygen will have been consumed in combustion before the exhaust gas reaches the WB sensor. I suspect If you advance too far the BENs will stay close because on both banks all the oxygen has had time to be cosumed in combustion??

[Gelf VXR]

Not sure if this helps...
http://www.innovatemotorsports.com/resources/myths.php

Regards
Paul

Thanks, but not really. Good information about MBT, whats required is a method of trying to locate MBT using EFIlive and a wide band.

What requires clarification is that if for example a engine has been tuned on a dyno for maximum torque at a given AFR, what is the effect on the AFR when timing is advanced or retarded from the MBT setting?

[The Alchemist]

hi, I have found that with tuning timing that AFR varies with timing changes at a given load and RPM point for sure. Just like you have found :)
Once you have the fuel close enough for jazz so to speak you will find that as you move the timing closer to MBT (keeping the fuel tables unchanged) the mixtures tends to lean slightly as the engine is burning the fuel more efficiently I guess. Once you reach mbt and start to over advance the engine the AFR starts to go richer again even with 1 degree change. So I often use AFR AND TORQUE as a guide when setting best timing on power runs.
Setting best fuel for best torq is another story altogether :)
Hope this helps.

Mike

[5.7ute]

The way I see it MBT gets the peak pressure at the point in the cycle where it will have the greatest influence on the crankshaft. Hence the increase in torque. Being a degree or two either way will not have a major impact on torque output but if it is too far advanced can cause knock.
Since the fuel is totally consumed within approximately 70 deg of crankshaft rotation retarded timing should still read a stoich AFR as it still has enough time to burn completely.
It would be interesting to see some empirical data off a dyno though to see how far retarded you would have to go to see any change in the resultant AFR,s.
As for the BEN,s converging. It may be as simple as a slight change in the cylinders filling ability with the change in timing.
Great topic for discussion Gelf VXR.

[The Alchemist]

It would be interesting to see some empirical data off a dyno though to see how far retarded you would have to go to see any change in the resultant AFR,s.
Great topic for discussion Gelf VXR.

trust me, 1 deg you will see a change, 2 degs a blind man in a snowstorm would see it :)
Next tune I will prepare a couple of runs with timing changed but fuel kept "the same" and will post it up on this thread :notacrook:
Mike

[Gelf VXR]

How long before we can get 5 gas exhaust analyser to replace the WB O2 in the NB bung? And unexpensive

[Gelf VXR]

I am putting the finishing touches to a spread sheet (beta if you like) that will graphically display the effects of altered spark on BENs, BEN difference, grams/cyl, actual spark and knock from a base log of the stock spark advance with up to three alternative spark logs.

Its very high resolution and spreads all of cell data right across the MAP and RPM ranges. Corrosponding maps need to be set up in the scan tool to enter logged data.

MAP ,0,1,2,3,4,5,........,105 RPM ,400,800,900,1000,1100,........,6800

TAB 1 -Instructions to be completed
TAB 2- Log inputs with output for BEN difference of base and logs 2 to 4, inputs as follows

INPUT - BENS LC1 (WB 1)
INPUT - BENS LC2 (WB 2) logging both banks at the same time is manditory
OUTPUT - BEN difference
INPUT - grams per cylinder
INPUT - Actual Spark (hi octane map needs to be copied to lo octane in tune)
INPUT - Retard due to knock

TAB 3 - logged data outputs, as follows

OUTPUT - Difference BENS LC1 of log 2,3 and 4 with base log
OUTPUT - Difference BENS LC2 of log 2,3 and 4 with base log
OUTPUT - Difference of BEN difference of log 2, 3 and for with base log
OUTPUT - Difference of grams/cyl of log 2,3 and 4 with base log
OUTPUT - Difference of actual spark of log 2,3 and 4 with base log
OUTPUT - Difference of knock of log 2,3 and 4 with base log

TAB 4- AS above but with provision to only display data in the range of grams per cylinder as inputed by user.

I did tab 4 because the logged data is in the VE table format, MAP against RPM. The spark tables are in the format RPM against grams per cylinder.

For example if the range is set to 0.118 and 0.122, the output data displayed will be for grams per cylinder 0.12 for all rpms (that were logged) and easily referenced against changes made in that range of the spark table.

TAB 5 - HI octane tables for reference

I am unable to do any logging my self at present as I am abroad working, so if any one has any data loggs where the only changes being made were to spark advance with pids for BENs both banks, air per cylinder, spark and retard due to knock being logged that they could let me experiment with???

If any one would like to try out the spread sheet themselves excel sheet (30MB) I can send it as an email attachment, 2MB limit for attachments here.

What i am doing is using conditional formating to highlight cells green when there is a desired result, pink for undesired result, yellow for uncahnged result. So that with out having to stare at the numbers, you can graphically interpret the the output results as positive or negative.

What needs to be determined is what results are deemed to be positive or negative for each of the outputs??

For example, if the difference between LC1 and LC2 BENs is 4% in base when compared to 2% in log 2, the results are positive and highlighted green.

In the two logs that are entered into the spread sheet, this can be seen. How ever the BEN,s in log 2 indicated AFR went rich, BEN values decreased, which follows that VE decreased (negative??) but the grams per cylinder values increased, the PCM calculated that there was more air (positive???), and this is why AFR went rich, the PCM injected more fuel becase it calculated there was more grams per cyl of air. ???? Anyone?

[Gelf VXR]

Why timing affects AFR / BENs. Fueling is fixed agaisnt the VE map (VE=g*K/kPa), so the PCM calculation for grams per cylinder for base and subsequent logs assuming the VE map remains unchanged should remain the same or very close. If this value increases by 5%, then BEN value would reduce by 5% as 5% more fuel was injected, not as a result of timing change. If it remained the same and the BEN value reduced by 5%, this is as a result of less g per cyl fill / reduced VE as a result of the timing change.

Multiplying the BEN value by the APCYL value equals the actual grams per cylinder value. BEN values and APCYL values should not be independantly compared from one log to the next, only the product of these two sets of data.

BEN X APCYL (PCM's calculated grams per cyl figure) = grams per cyl (Actual)

grams per cyl (Actual) X (Injector Pulse Width X Injector Flow Rate) = COMMANDED AFR


When timing position is changed, the torque / VE characteristics of the engine change for better or worse, marginally (if close to MBT) or substancially (too far advanced or retarded). Its this marginal / substancial difference in the grams per cylinder fill thats throw out the BEN value as observed by The Alchemist, nothing to do with the exhaust gas compersition at peak MBT being different at stoich from any other timing point at stoich as i thought:bash:

If BENs tend toward lean after applying advance = moving toward MBT
If BENs remain unchanged after applying advance = very close to MBT
If BENs tend toward rich after applying further advance = moving away from MBT

With dual BEN readings, when BENs remain unchanged and BEN difference is at its closest and remains unchanged = MBT average of both banks:notacrook:

IMHO this represents the most practical way of locating MBT on the road, and I cant wait to get home and start experimenting

Edit:

I'm adding logged GM.TRQTRANS PID data to the spreadsheet as an additional indicator for MBT

[Gelf VXR]

The good news is TRQTRNS PID works a treat, I managed to do a 2.4MB base log, AND updated VE from BEN's. The following day i managed four <1MB logs, the first a base log and 3 subsequent logs where in each case i increased the spark table by 1 degree.

The data captured is minimal, but the effects of spark can be clearly be seen against deliverd torque:rockon:

In some cases as marked in attached pic, two peak are visable, possible as a result of the theory that each bank peak torgue is slightly off from the other.

All the data is average apart from KR which is maximum, I attach the spread sheet (1MB) with all the logged data entered which compares the 5 logs by spark table cells, for example

0.12 @800
0.12@1200
0.12@1400

It is possible to tune spark table on the road:grin:

[Blacky]

Just a word of caution, the torque PIDs are only calculated by the ECM based on inputs from other sensors and values of commanded PIDs. It may* be possible that simply commanding a different spark or different AFR will change the calculated torque output - depending on the torque model that the GM calibrators programmed into the ECM.

* I do not know that for a fact, just something to consider. In some calibrations I think the torque PIDs' values are simply looked up in a table based on airflow and rpm, then modified slightly based on other values such as IAT spark and AFR - it may be that is the effect you are seeing/logging.

It would be nice to see the measured torque from a dyno based on the investigations you've been doing...

Regards
Paul

[Alex(AUS)]

The good news is TRQTRNS PID works a treat, I managed to do a two 2.4MB base log, with out updating VE from BEN's, the following day i managed four <1MB logs but in each case increased the spark table by 1 degree.

The data captured is minimal, but the effects of spark can be clearly be seen against deliverd torque:rockon:

In some cases as marked in attached pic, two peak are visable, possible as a result of the theory that each bank peak torgue is slightly off from the other.

All the data is average apart from KR which is maximum, I attach the spread sheet (1MB) with all the logged data entered which compares the 5 logs by spark table cells, for example

0.12 @800
0.12@1200
0.12@1400

It is possible to tune spark table on the road:grin:

It sounds very interesting Gelf. Well done!

I am not sure how the trans torque pid is calculated/measured. Have you noticed that if you add more timing past MBT that the trans torque drops off? Or, does it continue to increase? If it drops of, then cant you simply tune to that?

Thanks,

Alex

[Gelf VXR]

Just a word of caution, the torque PIDs are only calculated by the ECM based on inputs from other sensors and values of commanded PIDs. It may* be possible that simply commanding a different spark or different AFR will change the calculated torque output - depending on the torque model that the GM calibrators programmed into the ECM

Regards
Paul


Thats not all bad news thou, it just means were tweaking torque to GM mathematical model that they must have varified on the dyno??

I don't believe the model is any where near as that sophisticated from what I have seen so far, I suspect that the only sensors used are RPM and CURRENT gear and the PCM is calculating the torque from the rate of change of RPM against time, dependant on which gear is selected. A simple and sound formular. Similar to the way torque is reduced by gear as seen in attached pic of TC calibration

[Gelf VXR]

It sounds very interesting Gelf. Well done!

I am not sure how the trans torque pid is calculated/measured. Have you noticed that if you add more timing past MBT that the trans torque drops off? Or, does it continue to increase? If it drops of, then cant you simply tune to that?

Thanks,

Alex

Yes you can see torque increasing as spark is increased and decrease as spark goes beyonb MBT. You can also see the effects of engine braking giving negative torque

[Gelf VXR]

This is not meant to be all to end all process, just another tool to indicate MBT. For example maximum VE should be seen at MBT, so the results can be varified by monitoring BENS, any devation from MBT should see VE reduce?


The same method can also be used to see the effect of changing commanded fuel on Torque :)

[Gelf VXR]

Another important indicator I forgot to mention, I've had an autoVE tune loaded in the PCM for ages and average MPG was around 15.5MPG, never getting past 16, on the day I did the four logs I saw MPG rapidily increase to 19MPG after appplying 1 and 2 degrees of extra timing, 3 degrees saw no real change.

[macca_779]

Isn't the optimal timing table used for torque calculations? If it is then perhaps logging the trans torque PID has a direct relationship to this. One way to test it would be to leave your high octane spark table alone and match the optimal spark timing with it.. If torque goes up then it would prove it.

[Gelf VXR]

Updated the log info in the spreadsheat in earlier post. I had done a 2.4mb and updated VE from BEN's, the four logs on the same day where base, +1, +2, +3 to the HI octane map.

After staring at the numbers in the sheet for a while, i have come to the conclusion that AutoVE BEN's can vary from log to log because actual spark may not always reflect the commanded spark. TTT is how to only include data when actual spark = commanded spark, a custom pid similar to BEN factor execpt that looks at the commanded spark and actual spark and returns a value. How to do:help2:


I think the tuning process should be

AutoVE to base spark
Log base, +1, +2, +3 increasing spark of each log
Note down actual spark value for when BEN factors indicates the leanest condition.
Make a mental note of MBT from the TRQTRANS pid
Enter the noted actual spark values into the commanded spark table
AutoVE to the new spark
Log base, -1, -.5, +.5, +1, +2
Note down actual spark value for when BEN factors indicates the leanest condition.
Note MBT from TRQTRANS pid, should be increasing in value and getting closer to leanest observed BEN condition
Repeat process until BEN values returned are <= 1.00 (Maximum g/cyl VE)
At this point MBT from TRQTRANS pid should be equal BEN = 1.00, all other cell BEN values should be less than 1.00 and thus lower torque.
AutoVE to optimum spark map for the current commanded AFR
Log base, +1, +2, +3 logs using alternative PE or lean cruise values for commanded AFR
Note the commanded AFR that yeilded MBT for PE or lean cruise from TRQTRANs pid and set PE lean cruise to suit.

Job done (or not):)

From the attachment, optimum spark may be more or less than the commanded spark

[Gelf VXR]

Isn't the optimal timing table used for torque calculations? If it is then perhaps logging the trans torque PID has a direct relationship to this. One way to test it would be to leave your high octane spark table alone and match the optimal spark timing with it.. If torque goes up then it would prove it.


I don't know to be honest, I'm hoping some one can say for certain.

From the calibrations i have seen in TC, the torque calculation is related to the speed of the input shaft, I suspect that the PCM monitors the speed of the shaft and references a look up table that predetermines the torque required to increase or decrease the speed of the shaft that was monitored in the next cell, with additional factors for the current gear selected.

[Highlander]

Well... after a few months of having my pops z06 where it should be I strapped it on the dyno to do a "fine-tune".

I gained 2rwhp by advancing 2º than what I had before. My bens did change in some spots to a maximum of 4% in a few areas. Now... Other than that.. I saw NO difference in bens by advancing the timing. Advancing the timing further yielded 0HP gain and/or changes in the VE or BENs.

BENs where DEAD on most of the time except those FEW 4% changes which where only a couple of revs. This means MBT is a range and 2rwhp on a 450rwhp car is not something to be statistically differentiable IMO to conclude where MBT is.

Now... my father says... it runs faster now.. JAJA

[Alex(AUS)]

Well... after a few months of having my pops z06 where it should be I strapped it on the dyno to do a "fine-tune".

I gained 2rwhp by advancing 2º than what I had before. My bens did change in some spots to a maximum of 4% in a few areas. Now... Other than that.. I saw NO difference in bens by advancing the timing. Advancing the timing further yielded 0HP gain and/or changes in the VE or BENs.

BENs where DEAD on most of the time except those FEW 4% changes which where only a couple of revs. This means MBT is a range and 2rwhp on a 450rwhp car is not something to be statistically differentiable IMO to conclude where MBT is.

Now... my father says... it runs faster now.. JAJA

You should have taken some timing out to see if you have already gone too far ... as you said the 2rwhp is nothing when you have a 450rwhp car and most certainly varying conditions between runs (ie IAT, ECT, CAT Temp etc) which all have correction factors and produce different torque/power even if they are zeroed ...

Alex

[Highlander]

Its a possibility, but given my experience with these engines. N/A from 24 to 28 I tend to gain about 6-9rwhp... I was seeing 27 overall and went to 29.. then 30... then went back to 29... no knock ANYWHERE... Coolant temp was where i wanted it and reproducible...

It actually was 3rwhp and 4rwtq in total.... So not bad after the car was already set and did 124mph on the 1/4.

What I will be doing is either getting an LM2 or using the analog voltage to log both banks and see how they change with a change in timing.

[Highlander]

http://i5.photobucket.com/albums/y164/andysa1966/BENsvsMBTTiming.jpg
I would have thought that timing would be totally relative to the piston going TDC and that this shouldn't apply. So 27º total advance would be 27º before TDC relative to that piston. But I can see how this could justify the difference in trims that one can also see in certain conditions between bankings... I like your abstractive nature when looking at it.... But.... I think the PCM compensates for the difference in bankings... although, since the difference is not that much maybe they don't? NICE PIC!

I think this could be verified with a timing gun?