RomRaider

Background

The purpose of this spreadsheet is to create a table to summarize knock data. It is meant to fine tune the timing table after it has been set up using WOT logs. This spreadsheet assumes that the ROM is set up such that an IAM of 1 can be maintained.

This is based upon the information that merchgod provided here: viewtopic.php?t=1840

and was inspired by varmint's request here:
viewtopic.php?t=2029

When romraider was down, I posted version 12 on nasioc and received some feedback from normally aspirated users.

http://forums.nasioc.com/forums/showthr ... ?t=1412169

This spreadsheet keeps track of fine learning knock correction. So far, I have been able to track it reasonably well when I used merchgod's learning viewer tool.

topic2772.html

This spreadsheet does not include knock from poor tip-in. I eliminated that on purpose because I didn’t want to subtract base timing due to throttle changes which really are not dependent upon rpm-load location in the base timing table.

Instructions

Instructions are embedded in the spreadsheet.

VE Added

Since all the parameters are logged, I added a VE table as well. This was recommended by Freon in his post here:
viewtopic.php?p=24679&highlight=#24679

This is useful for tuning AVCS.

There are some caveats with the VE table. The VE includes a rudimentary correction for temperature (absolute temperature ratio). Also, you have to be careful that some data points may be anomalies. It uses an average for each cell but each cell may have as little as one data point in it.

EDIT: Added a subroutine to VE spreadsheet that automatically color cells with too little data as light grey.

EDIT: The pivot tables in the VE spreadsheet were changed so that the fields are ascending.

EDIT: Added Intake Air temperature to air density calculation. IAT is set up for Fahrenheit. Note that if you don't log IAT, you can enter in a default value in the box provided.

EDIT: Added fine learning tracking.

EDIT: Uses merchgod learning view tool and FLKC offset parameter.

Don't know if this will help, but I'm adding the current fine learning table offset to the next logger def update. That is, you'll be able to log and determine when the applied fine correction is accessing a specific load/rpm cell and when it changes.

Right now, I have it set up to gather data from the time of ecu reset to the time that fine learning becomes active.

It'll be interesting to see how we can take advantage of this data.

Any comments? Works? Doesn't work? Etc...

I added a subroutine to the VE spreadsheet to color the cells with less than 5 data points as light grey. I also threw out any data with VE > 1.3. There are a handful of erratic MAF readings that are obviously not correct.

I'll see if I can get some logs in tonight and on the way to work tomorrow. Looks like a good tool!

Any feedback?

I think my VE calculation is wrong. It seems to work at sea level but when I tried it on a log from high altitude (12.2 psi atmospheric), it doesn't work.

Here are my equations:

MAP = MRP + P_atm

P_atm is the actual atmospheric pressure and entered in a box in the spreadsheet.

densitySTP = 1.2 kg/m3 = 0.0012 g/cc at 14.7 psi and 70F

displacement = 2457 cc for 2.5liter boxer engine

load is defined for one revolution of the engine which is 2 combustion cycles out of 4 cylinders

load_theoretical = (MAP / 14.7) * densitySTP * displacement * (1/2)

VE = actual load / load_theoretical

edit: added explanation of P_atm

mickey, this may help:

http://www.romraider.com/forum/viewtopic.php? ... +mrp#18213

pay particular attention to the part where i assumed ambient presure was 14.7psi due to being at sea level...

hth
ken

Hmmm... I've read this a couple of times now and I'm still not getting it.

For your equation at sealevel, you have



I think my equation is the same except that I have

MAP = MRP + P_atm where P_atm = actual = 12.2 psi. I kept the 14.7 in the denominator since that is for the conversion from STP.

The only other difference is temperature but I didn't think that was that big a difference since a 25C difference is about 9% difference in load. I'll throw that in and see what happens.

I was having some issues when using more than 2 separate logs. The spreadsheet would add in a header row and cut off the last column. I haven't had much show up in the fine learning or feedback knock correction yet, but I need to do more midday, stop and go driving.

BTW, as far as VE goes, what kind of numbers are considered decent?

Well, it appears that my VE calculation is wrong. Still trying to figure it out. It might be that the person who gave me the log was going up and down the mountain so his P_atm is not constant.

VE will vary depending upon where you are in the load/rpm range. So, it's more of a comparison type analysis between tunes. Theoretically, a VE of 0.85 to 0.90 would be good.

I don't believe the spreadsheet adds in a header row. You should double check your csv. Sometimes the logger will do that if you change a parameter midstream or turn the engine on/off.

I believe the spreadsheet will cut off the column of the 2nd log if it doesn't match anything in the 1st log.

In the off chance that you ever actually need some logs going up a hill I gathered some neat ones the other day. Both have lots of 4th and 5th WOT and are less than 5 minutes each. One starts at 10.74 and goes to 9.87 and the other one goes from 10.45 from 9.72.

Did you log manifold absolute pressure or manifold relative pressure?

I need MAP logged. If os, email them to me at my screen name at yahoo.com.

I logged manifold relative pressure.... but also logged atm. Sounds doubtful these will work but I'll email and perhaps they'll be of some utility.

Dammit. The person sent me a log with calculate load (2*g/rev).

Make sure to log load (g/rev).

I did discover that I had forgotten to change the pivot table field order to ascending. Updated that.

you probably already checked this, but make sure your temp calcs are done in absolute degrees.. so kelvin/rankine.