Setting up MAF

[durson1]

Running the new 3.2.1 code. Just upgraded from the B&G. Runs good an speed density. Trying to get it to run with a MAF sensor. 2.0L 4 Cyl. VW. Wanting to use a MAF off a Ford escort/ranger/ and others ranging from 2.0-3.0L based on what I've found. Does anyone have any curves, volts vs flow, or maffactor file for any of the Ford MAF's in this displacement range?
The one I have is a: F67F-12B579-EA.
Or is anyone running a MAF on a NA 2.0L, and what MAF are you running and could I get a copy of your maffactor file? I am willing to buy another sensor to get this to work. I've been playing with bits and pieces of numbers I've found and plugging them into the maf analyzer app but what it generates doesn't look right.

[techsalvager]

I have a few curves, I will be done generating a curve for a N60 nissan maf soon.
I may have a curve for that one you have. I will check.

Pop up a picture of what the maf program generates.

[durson1]

I can idle with it, but any throttle leans way out. The mafanalyzer program is on my home desktop. If you need a screenshot of that I can do that later. It will run with the ford V8 inc file. I have to set the correction table bins all to 40%, and in VEAL it drops the VE way low, and my fuel load doesn't get above 18%. I think I just need a better starting point. What MAF sensors do you have curvesw for? I am willing to change it out to get this to work. I can post an msq if needed. I forgot to log on my test run though.
Thanks for the help.

[techsalvager]

get a log of it, also throw up your tune and a pic of the maf analyzer with the maf curve you put out.
I will post a how to tune maf in, in a moment.
Just N\A or forced induction as well?

Also voltage\adc count at key on, engine off and idle.
Whats the grams a second reported in the datalog you are idling at?
Also what fuel injectors?

[durson1]

I'll get a log as soon as I can. It's just NA. 3 gps at 1100 rpm idle. 19# Bosch injectors. I'll get the rest of the info you need as well.

[durson1]

Heres a screenshot


That came out small. Here is a link to it.
http://i16.photobucket.com/albums/b31/d ... nshot1.png

I ran the mafanalyzer program a few times tonight and noticed if I run the numbers I have I get a max flow of 188 gps which seems resonable. If I include a zero reading of .17v which gets rounded to .2 the max flow it shows as 115 which seems way too low, which is close to the file I was trying to use earlier today. In a pdf included with the maf program I read that if you include a zero reading it will use linear least squares, any other number it will use a non linear fit. I guess thats why the 2 files I get are different. I guess I will try using the least sq. file next time and see how close it is. I used the one that maxes around 115 because it starts with a few lines of zero's and have heard thats how they should start. Maybe this all shows how little I know. Oh well.
As a side note these are the numbers I'm using. I wish I could remember the source so I could give them credit. Its in a MAF summary file I found while researching.

V g/sec
0.34 0.94
0.39 1.07
0.6 1.68
1 4.77
1.96 16.87
3.9 88.91

[techsalvager]

in your tune, turn off secondary fuel load, don't use one. Only use primary fuel alogthrim set to MAF.
In your AFR table, set all the idle and cruise targets for time being to 14.5 or 14.7 and leave higher loads around 13 or 12.5. Makes it easier to tune for commanded AFR if the area is the same mostly.
can you tune idle to the same afr as in afrtarget table or is it allways rich?
Also get a datalog if you can.
disable your wideband under EGO control, we do not want closed loop working or else we will be chasing our tale.
controller step size = 0, ego sensor type = disabled.

[durson1]

I will try those changes. Idle was always rich, 12-12.5 afr.
The setup I was following was the directions from "elaw" in this post.

viewtopic.php?f=101&t=40570

I wanted to use VEAL to help fine tune once I got a working maffactor file.

[techsalvager]

I will try those changes. Idle was always rich, 12-12.5 afr.
The setup I was following was the directions from "elaw" in this post.

http://www.msextra.com/forums/viewtopic ... 01&t=40570

I wanted to use VEAL to help fine tune once I got a working maffactor file.

You won't have to use VEAL with tuning if you tune the maf transfer function properly. Infact in maf mode you can't as VE table is useless.

If idle is rich allways reguadrless what you do then injector dead time is more then likely wrong. What injectors are you using?
Upload your tune after you make the changes.

[techsalvager]

I think I just need a better starting point. What MAF sensors do you have curvesw for? I am willing to change it out to get this to work.

Currently finishing up a transfer function for this one.
My high airflow is close expect past 4.5v, working on middle voltage airflow 2-4v range and my idle voltage area airflow is tweaked in well enough.
unfrountaly my car idles around 1.5v and I only see as low as 1v of airflow so I can only guess airflow under 1v, I do know dead flow is at .30v
I'm maxing this sensor out at around wastegate pressure up top near 4.4-5v, have to increase the boost to finialize past that to near 5v and it should be done. This sensor though is too small for me, I plan to go for a N62 which reads more air mass flow.
The n60 sensor should be able to be found cheap on nissan forums, its not a sensor that people want since its not usually a great upgrade from what they have. The n62 out flows it by a bunch and is what people usually go for, but for a n\a 2l the n60 may suffice.
I dynoed in at 148whp with my speed density setup and wastegate with a smaller intercooler, I have no idea what I'm hitting now but I would expect around the same. I expect maybe the sensor will be maxed around 200whp in a pull through setup like mine ( maf infront of turbo )

Picture of the Nissan N60

[techsalvager]

Here is a rough guide, I still need to edit and enter a few more items but give it and read and comment on what you don't understand. I need to make it better for everyone to understand.

https://docs.google.com/document/d/10h4 ... prli4/edit

[durson1]

Nice write up. My injectors are Bosch design 3 part# 0280155710? 210cc@3 bar. I am set for 1 ms open time. They are high imp. I figured the open time mightactually be higher, but it sounds like it may need to be lower.I haven't made those changes yet. I'll upload my tune when I get a chance to make the changes. Is there a gauge in TS and does MLV log MAF volts and ADC count? I know I've seen maf gps, and maf map.
That n60 maf you, do you have a pretty good transfer curve on it? I am good at mechanical and electrical, the software, tuning, computer side I am not so proficient at. I have a very good SD tune now, but ultimately would like to get it running with a maf sensor. During the week I don't get much time to work on it, but I'll do what I can.
Your way do tuning seems more direct. The other thread had me make changes to get the VE table 2 to fine tune maf with veal. Thanks for the help. I'llmake some changes and report back.

[techsalvager]

By default tunerstudio does not datalog maf voltage and ADC, I have to put it in the guide but for now here is what you do

Create a new text document and enter these into it

Code: Select all

[OutputChannels]
voltsADC7 = {adc7*0.0048876}
voltsADC6 = {adc6*0.0048876}

[Datalog]
   entry = maf, "MAF", float, "%.3f"
   entry = mafmap, "MAFMAP", float, "%.3f"
   entry = adc7, "adc7", int, "%d"
   entry = adc6, "adc6", int, "%d"
   entry = voltsADC7, "MAFvolts", float, "%.3f"
   entry = voltsADC6, "Knockvolts", float, "%.3f"


Save this as "custom.ini" and put it in your projectCfg folder.

Change ADC to point to the proper ADC that the MAF is attached to. JS4 is ADC7, JS5 is ADC6
You can remove knockvolts, but keep the raw ADC count datalogging.


These I take it?
Bosch EV6 injectors.

[techsalvager]

for your idle rich issue, start lowering the dead time, what was it at 1ms?
Slowerly go down by .05, so .95, .90, .85, .8, .75, .6ms, .55ms, .5ms
with a 2l N\A you should be able to idle it between 2-4 grams a second at 14.7 depending on load.
As you bring down the dead time, it should start leaning out the fuel mixture, use maf correction table to compenstate.
Set your commanded AFR table to 14.0 in this idle area for time being. Once you have a believeable grams a second at idle, you are probably close to your injectors true dead time.
With proper injecto dead time, your AFR shouldn't change much at idle with varying voltage.

Prior on my setup with injector dead time at 1.2ms with the maf and battery voltage at 12v it would idle around 13.5 afr, but soon as hit 14v it would idle at 11 afr. Now its way more solid with the varying battery voltage output.

[durson1]

Soif I do change maf sensors do you have a curve for that n60? Is that off a maxima?

[techsalvager]

Soif I do change maf sensors do you have a curve for that n60? Is that off a maxima?

yes I have a curve for it for my setup, which I can measure out for you. Though you should be able to tune a curve for the maf you have, takes time and learning but after learning to do one you can figure out how to do others.

I'm not sure the car its off of, I think its a v6 maxima or v6 altima?

[durson1]

I spent more time reading your writeup. Let me see if I get this. After I add that custom.ini, while running on my SD tune, I log :
ADC count, and MAF volts. And I do that with my best estimate mafanalyzer file burned? Or does that not matter? And is this just to see if my MAF sensor is correctly sized?
I think I then understand the tuning side. Towards the end of the document when you speak of percent changes if your rich or lean after you calculate you do that in the maf correction table, then make it a permanent change with new numbers in the mafanalyzer program. Correct? Maybe I am missing something. And this needs to be done with ego correction off? Where do I start though. My files I've generated with mafanalyzer haven't been close at all. I couldn't even get out of my driverway with ego correction on. Or is that file generated using info gathered from datalogging with my SD tune. I am missing something. and talking in circles I think.

And yes those are my injectors. They have the older EV1 connectors on them. I think there are some identical with EV6 connectors.

[techsalvager]

I spent more time reading your writeup. Let me see if I get this. After I add that custom.ini, while running on my SD tune, I log :
ADC count, and MAF volts. And I do that with my best estimate mafanalyzer file burned? Or does that not matter? And is this just to see if my MAF sensor is correctly sized?

To see if its correctly sized. Also if you know how many Grams a second your car normally idles say 3 grams a second, then you know say if the maf you are using is at 300 adc at idle, then thats probably 3 grams a second there.

I think I then understand the tuning side. Towards the end of the document when you speak of percent changes if your rich or lean after you calculate you do that in the maf correction table, then make it a permanent change with new numbers in the mafanalyzer program. Correct?

exactly

Maybe I am missing something. And this needs to be done with ego correction off?

All tuning should be done in Open Loop, EGO correction needs to be off until we have fully tuned in the maf transfer function, any usage of EGO correction will hurt tuning.

My files I've generated with mafanalyzer haven't been close at all. I couldn't even get out of my driverway with ego correction on. Or is that file generated using info gathered from datalogging with my SD tune. I am missing something. and talking in circles I think.

What did it do? go rich or lean?
Where you start is your injector dead time, start decreasing it.
Anytime you tune, you datalog!

[durson1]

It went lean. Way lean. I guess I need to get the dead time right and get it to idle at my specified afr first.

[durson1]

So what is the load index in the afr table now? It shows %. Percent of what. It still shows my MAP numbers I was using for load index.
If I switch EGO sensor type to disabled will it still give me an afr reading? Shouldn't I just set authority to 0?
With that custom ini will my log include MAF volts and ADC count even running in SD mode? I dropped it in my SD and MAF projects, project config folders. I created a new project for MAF to keep them seperate.
Here is an updated msq. I have a new curve I created in my .inc folder ready to burn when I get a chance to try next.