You got PM.

 

SJSchafer I made the changes you suggested in another thread about the injector scaling for 60s and copied those numbers into my flow rate tables im still not sure what I should be using for short pulse limit and min PW, im using stage 2 values right now, also I dont understand the purpose of the short pulse adder table

 

The short pulse adder table is used to try to compensate for the lower, non-linear area of the injectors. Injectors have a linear area where they can pulse acurately as well as non-lier areas where they pulse inaccurately. They go non-linear up top and down low. The 60s lose the ability to stay linear at about 3.125ms. The errors are so small there as to be almost non noticable, but the lower you go in PWs, the worse it gets. The adders add PWs to the non linear areas of the injector to help make them linear. I would start by entering .6 on the adder table at 0ms and with the rest zeros, highlite all the cells from 0 to about 1.25ms and intropolate along horizontal. That should give you a good starting point. Then set up your VE histo (but stay in MAF) to see at what PWs you are having errors. Set your low PW limit to 1.125 and min PW doesn't really matter. Just put in a number that is less then .5ms...

 

i still dont see how im supposed to notice these errors this just isnt making sense

 

Okay, lest say that you tune your MAF for everything that commands 1ms and higher. They are all dialed in, right? Now log on the two VE histos that I told you about (while still in MAF). You will notice that in low PW areas the AFR may be running way lean. You can see that in your VE AFR error histo. You look at some of those same cells and see that you are consistantly running (for example) .685ms in those lean areas. Go to your tune and look at the adder table. Do some math and figure out where you are commanding that PW (cell value + adder value) and bump that cell a touch. You will likely see it between 2 so what for trends so you don't adjust the wrong one. So, for instance, I will look at my logs. I know that at 3400 htz on my MAF, I am good, because I tuned it up where the injectors are fairy linear, but my graph shows lean, so I look at my VE table that shows PW and see that I am running .978ms there. I look on that histo for more PWs in the high .9s and see a few. I check my VE AFR histo and see that everywhere I command high .9s I am lean. I look at my adder table from my tune and see that because of my adders, I am commanding .99ms at .875 (.875 cell has an adder of .115. .875+.115=.99). So I bump the adders at .875 just a touch. Then I relog. With persistance you can get 14.7 across the board, even during decel. There are easier and faster ways to tune these, but that way is correct. imo

 

For you plot filtering try useing this:

You would want to enter.

STFT's low freq= [SENS.121] >14.5 AND [SENS.121] < 14.9

This is to block out PE and DFCO. This is good if you are tuning off of Fuel Trims. But this in both your STFT and LTFT MAF Histogram Tables

 

Which is fine, except that it won't really help with tuning the MAF only at 1ms and above.

I am at home now, so I have access to HPT. Try this to eliminate everything below 1.2ms
[PID.6200] > 1.2

 

starting to make some sense now

 

Woohoo!

 

What was originally turning me off to the 60s is the fact that they couldnt be scaled properly but you seem to have found the workarounds for this

 

Yea, honestly, it's not the 60s that is the problem. It's our returnless fuel system. I plan on going to return style when I get a few $$ together. That will make these injectors behave much better at low PWs. Should make tuning MAF a lot easier too. Right now you can hit a lot of the MAF cells in different conditions and be lean or rich depending on how much VAC you're drawing. With a return system regulated to MANVAC, you will command the same PWs for a given MAF hz no matter what. 2 or 3 tuning runs and you're done.

 

I need to take a vacation, I have so many damn parts sitting at my house I need to install which should make everything easier.
Got a completely new fuel system sitting in my closet just itching to be installed, plus my clutch, hahn intake manifold, new FMIC, rods, pistons, head gasket, well actually have another complete motor just waiting to be put back together (should handle upwards of 600hp no prob lol)

 

I hear what you all say about tip in knock, but I didn't get any of that either. I also drove differently than most people. [Clutch in, slam gas, dump clutch] No tip in knock. And one hell of a burst.

 

It's official. I hate you.

 

[QUOTE=SJSchafer;2019705]Set up a MAF histogram that shows your plot as the MAF hz and your data as ARF error. Then go to plot filtering in your histogram and set the filter to only allow the data that is above 1.2ms. That way you can tune most of your MAF without worrying about low PW errors giving you bad data. Once you have the MAF dialed that way set up 2 VE table histos, one reading AFR error and one injector PWs. By looking at the VE AFR error histo, you can see where your rich/lean errors are and by looking at the corrisponding cells in your VE injector PW histo, you can see what PWs are being commanded during those errors. That lets you know where to tweek the adder table to eliminate your low PW errors.


Once the weather gets a little warmer I plan on tuning my car. I have set up the injector control tables for 42lb injectors. I have a few questions for you.

1. Should I tune MAF or VE first. I am getting conflicting information on this. I am thinking I should tune VE first as the MAF checks the VE periodically for corrections. Is this right?

2. Should I copy the injector control tables in then start tuning VE + MAF or should I tune it before then paste the tables and tune again?

3. I have been reading about your explination on plot filtering. For tuning the MAF I believe from what I read that you set up your histos MAF vs AFR error% to plot filter out any low PW errors below 1.2ms to tune your MAF. Then you set up your tune for VE and create two histos. One logging VE vs AFR error % and the other logging VE vs PW adder error. You use the VE vs AFR error % to tune your VE table and the VE vs PW adder to tune the PW adder table in the corresponding cells where you recieved an error on the VE table. Am I correct?

4. Now when you are tweeking the PW adder table, do you take the value in the cell you are working with and paist special or do you just bump up or down the PW by hand? If it is commanding a certain value +/- my understanding would be take the existing value and +/- the diference to adjust the PW adder table. Is this correct.

5. Now when tuning the VE table, should I smooth out the table everytime I make an adjustment or can I fully tune the table then smooth out all the peeks and dips. Then once I have smoothed out specific parts of the table should I then smooth the full table once, then take that finished table and paist it into my tune and flash to car? Now with that being done I then do a log once the car has learned the new MAF, VE and fueling tables to see how she runs.

6. Is there anything I left out or any other info you can give me that I have not thought about.

Sorry for such a long PM

Thanks Kiaso

bump

mid day bump

 

Hey Kaiso. I just returned your PM, but since you posted, I will post the reply here for others to read as well:

"Hey. I'll see what I can answer for you right quick. You are partially correct. You filter the MAF to get a good MAF without low PW errors. You create the 2 VE histos and you STAY IN MAF. You just use the VE histos so you can more easily see the lean and rich areas and what PWs you are running at the time. You need to stay in MAF so you can compair your AFR errors and only change the ones that occur in the MAF cells that you already tuned. Does that make sense? Really, when you are tuning adders, you are looking at 3 histos. You look at the MAF histo while going through your graph to make sure the error is in a MAF cell that you have already tuned to make sure it is not a MAF error. Then you use the VE PW histo to see what PW you are running there. You then check the other areas of that histo to see where else you are running those PWs. Then you cross referrence the VE AFR error histo to see if you have the same errors in those cells. You can also use those histos to see what happens at PWs just above and below the PW you are having probs with.

You cannot multiply by percent error. In oder to get the proper percentage you would need to know the commanded PW before the adder, the adder value and the offset for that voltage and VAC level, then adjust your percent, subtract the PW and what's left is your adder. The problem with that is that it won't work because with adders, if you are not enough, you will be at 16.xx-17.xx AFR and you can, with a very small adjustment be at 13.xx-12.xx AFR. This is due to the non linear nature of the injectors at those PWs. It is also due to the fact that the ECM looks at the adders tables as they relate to each other. You can adjust 1 adder too far up and make the adders before and after that one too lean while that one is too rich.

Allways tune in the following order (keeping in mind that some timing may need to be adjusted along the way to prevent knock):

1 - Set your IFRs and Offsets
2 - MAF tune normal driving (no WOT) above 1.5ms with 60s or above 1.2ms with 42s.
3 - adders working your MAF down untill your MAF and adders are perfect (still no WOT)
4 - VE normal driving (payin attention to adders that may need a little fine tuning that are more apperant in VE)
5 - VE PE
6 - double check you MAF and go ahead and tune MAF PE
7 - adjust your timing and PE adders for max power with no knock.
8 - re enable your COT and DFCO and check for consistancy in your MAF when COT kicks in (if COT runs rich or lean when enriching, check your IFRs and offsets and start all over from the beginning).
9 - enable trims and check for consistancy.

This seems like a long and painfull process, and there are many shortcuts. This, however is the correct method. imo
Steve

so much for 'right quick'"

Sorry, one thing I missed. When tuning VE you should "hand smooth" the non changed cells according to the pattern that begins to show itself. I do it after each run. Otherwise your AFRs baounce a lot when your ECM referrences other nearby cells. In the end, you should have very little smoothing that needs to be done or you have fueling errors somewhere else.

 

Awesome Steve. Thats about the best I've seen of anywhere that tuning process is described.

In terms of tuning adders, I'm still having trouble understanding where to read, and where to adjust to modify the adders properly. Here is a pic with the pid I setup according to how you explained, but I am not certain that it is right, and even if it is, where to read for a pulsewidth error, and where in the adder table to adjust accordingly.

 

You should set the decimal to 3

 

Sweet...Thanks Steve.

 

Yeah. That looks alot better. Now I just need to learn what to do with these.


0.83051 0.90405 1.04294
0.80872 0.89056 0.99945 1.57166
0.72861 0.90027 2.34985
0.70724 0.92295 1.05243 1.30393 1.58691 1.91607 2.07214 2.12708 2.40784 3.45612 3.61633
0.71132 0.92369 1.06320 1.27138 1.51193 1.78375 2.02942 2.25830 2.50626 2.90527 3.10863 3.36456 3.74756 4.01476 4.40470 4.70734
0.57373 0.65528 0.91953 1.08756 1.24887 1.50539 1.71543 1.98766 2.27224 2.59261 2.90462 3.09041 3.41797 3.69263 3.90625 4.51660 4.44031
0.52513 0.65494 0.88312 1.10683 1.27270 1.48762 1.79121 2.02297 2.33459 2.61591 2.90059 3.41225 3.50274 3.86471 4.16260 4.46429
0.51081 0.72506 0.86691 1.07168 1.26371 1.56439 1.83437 2.06938 2.41743 2.57019 2.44141 3.22377 3.49325 3.90116 4.05693 4.99344 4.70988 5.50842 5.57454 5.54657
0.55440 0.99182 1.75985 1.80054 2.73132 2.94434 3.11890 3.82560 5.17273 5.68542 5.84412
0.49591 0.53406 1.52588 1.66321 1.87683 2.31425 2.58582 3.29045 3.23677 3.95203 4.77092 5.18036
0.44250 0.44250 0.86975 2.13114 2.41297 2.71454 2.86865 3.17891
0.42725 0.73242 2.62451 3.03040 3.14331
0.53406 3.44849 3.52287
3.45612

 

WT....^? Take the square root of the Hypotinuse devided by the circumfrence of the moon devided by the amount of seconds in the aztec calender and there you go...haha

 

Yeah... equals lean --> Boom.

 

Those are the PWs you are running at each given time. Do this: Set up an excell spreadsheet that you can use to add your adders to the adder cell. Put your histo on "L" instead of "A". Now scroll through your graph and see where your problems are and look at the histo to see what PW is being commanded. See which adder cell equals closest to that PW and bump it up or down.

I was going to attach my spreadsheet, but it seems I am not allowed to attach. I used to be allowed to attach...

 

Yeh...like a rocket!

 

looks kinda like this:

Labels 0 0.125 0.25 0.375 0.5 0.625 0.75 0.875 1 1.125 1.25 1.375 1.5 1.625 1.75 1.875 2 2.125 2.25 2.375 2.5 2.625 2.75 2.875 3 3.125 3.25 3.375 3.5 3.625 3.75 3.875 4
Pulse Width Adder 0.61719 0.5078 0.3984 0.2969 0.1953 0.1172 0.1484 0.0859 0.0391 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
0.61719 0.63281 0.64844 0.67188 0.69531 0.74219 0.89844 0.96094 1.03906 1.125 1.25 1.375 1.5 1.625 1.75 1.875 2 2.125 2.25 2.375 2.5 2.625 2.75 2.875 3 3.125 3.25 3.375 3.5 3.625 3.75 3.875 4