ECU / Airflow Meter question

[Fatty]

oops, very late i know but i just wanted to thank jbt and 93 clubman for the info. i didn't had any internet access for a few weeks therefore couldn't post earlier but better late than never i guess.

[Fatty]

ok so here are the results. 93 clubmans info is correct and i think it lines up with that wiring diagram that jbt posted. the red wire is the airflow signal. the red / green wire is +5v.
my afm is putting out about 2.45v idling at 1000rpm. i sat in the garage and slowly increased the revs up to redline at 6500rpm, as the voltage reduced down to a final reading of 0.8v at redline.
so anyway these results differ a bit from what i was expecting , based on what matty thought the voltages should be. so either matty's recollections were a bit off, or my afm is dodgy. given that i think my car runs a bit richer than it should, i think it might be the latter.

[The Pupat]

Nope I reckon matty guessed.

Now if you go to a MAP sensor, how are you going to take into account RPM or can that DFA thing do that. Also you've now proven you need to get 5V down to 0.8V at WOT at redline so you'll need more than 2.5V correction. Best idea for this I can think of is use resistors to drop the voltage (I've forgotten it's correct name but it's sorta like a whetstone bridge). You'll still need something that will correct based on RPM.

[Sheck]

Hrrrm wouldnt you be best off with a voltage inverter? Not sure how exactly it works or whats involved but i remember reading it in one of my Miata books and thinking of building one. (which i still might do as i can use the EManage to tune out any problem areas)

Unless the DFA has an RPM signal input and has some way of taking that into account then you'd have to figure some thing out.
Or just make sure that when the MAP is reading full flow then the mixtures are at an ideal number when your up in the redline (guessing 12-12.5 would be safe for most motors, cause that what i'm goin off heheh!), then hope for the best that its not too rich in the lower rpms.

Dave

PS ohohoooo i just remembered that the dudes in the Miata book used a different ramp rate on the throttle cable (as in a different linkage thingy to where the cable attaches on the throttle body) - but they were runing carbies so their voltage inverter was mainly for the spark side of things. They said it would be the same principal for EFI though..

[Fatty]

yeah still haven't decided which way i'm gonna go yet.
the 3 options seem to be:
1: stick with a \"flapper\" style afm like the stock unit or upgrade to an rx7 one. this is probably the simplest option. this may require relocating the temp sensor somewhere after the intercooler as well, to get an accurate reading.
2: use a \"hotwire\" afm / maf , as on the NB for example. this will require the voltages to be inverted, and a temp sensor fitted somewhere. this should be less restrictive to the airflow than a flapper.
3: use a map sensor. again this will need the voltages to be inverted and a temp sensor fitted. this method is the least restrictive to the airflow.

i thought of that idea about dropping the voltage too. i actually used a similar method while i was calibrating the dfa, but using a potentiometer instead of resistors. but i think the best bet to get the voltages right for the last 2 options is to use the voltage inverter schematic from the \"miata performance handbook\". someone was kind enough to post this in another thread. it's a simple little circuit using a couple of opamps and some resistors and apparently only costs a few bucks to build.

no the dfa can not take rpm into account, but as far as i can tell the \"miata performance handbook\" makes no mention of needing to take rpm into account when fitting a map sensor. i'm sceptical though. but i think it might be worth a shot to fit one (they seem to go for about $10 on ebay) BUT not connect it to the ecu. then tune it with the dfa over a week or so, to try to get it's output to corrrespond with the stock afm over the entire rev / load range. i'd just drive around with 2 multimeters hooked up, comparing and adjusting as i go. if i can't get it to work, then i just dismiss the idea and look at the other options.

anyway, thanks for your interest and suggestions. i appreciate it and welcome any help!

[Fatty]

haha sheck, yep the voltage inverter is the way to go. i was typing out my long winded post when you posted.

[Boags]

I didn't know that the afm sends a rpm value to the ECU...

Whatever does it on the stock system, will still be doing it when you get rid of the afm and use the MAP sensor. Remember the DFA is just a piggyback. All it does is alter the air reading from the afm.

I can't see why it wouldn't work, if the curve can be set up the same after the changeover, which is the difficult bit.

Boags

[Fatty]

yeah that's kinda what i figured too boags. i don't think setting the curve up will be too difficult either, as long as the map sensor is outputting a different / discrete value at each load point.

[Sheck]

If you want a load based signal - instead of the MAP - then go for something like a MAF from an RB20 or VG30DETT as they are heaps big enough to not pose too much of a restriction to descently modified SR20's or CA18's (for the smaller RB versions) and even for a modified skyline (for the 300ZX version).

Hrrrmmm didnt think about that ECU RPM thingy. How quickly does the AFM go into its lowest voltage (ie highest flow) reading? If its pretty quick then the MAP should work fine and the ECU will take care of the rpm side of things.

Very interesting... I think i might try it out on my car as the AFM is soo big it doesnt let me put the filter anywhere! My set up is Turbo-AFM-Filter and it just squeezes in there. Getting rid of the AFM will open up some options....

Keep posting Fatty!! I want to know how you go

Dave

[The Pupat]

I didn't know that the afm sends a rpm value to the ECU...

Whatever does it on the stock system, will still be doing it when you get rid of the afm and use the MAP sensor. Remember the DFA is just a piggyback. All it does is alter the air reading from the afm.

I can't see why it wouldn't work, if the curve can be set up the same after the changeover, which is the difficult bit.

Boags

Errmm okay back to sensor school 101 to you.

a MAP sensor stands for Manifold Absolute Pressure and outputs a voltage that give you the Absolute Pressure in the Manifold IE if you got the manifold to absolute zero pressure it will put out zero volts and will output full voltage at the top of it's scale (that's why you buy 2bar, 3bar etc MAP sensors they work on a different calibration scale). An AFM is a Air Flow Meter and outputs a voltage based purely on how much air the motor is sucking in at the time. A MAF sensor stands for Mass Air Flow and is basically the same as the AFM sensor in that it measure how much air is flowing into the motor except that the hot wire sensor have a built in correction for air temp in that measure the Mass of Air flow past instead of the volume like a AFM does.

So lets take a few cases.
At idle
MAP: will output a low signal as there is vacuum in the manifold.
AFM: will output a low signal as there is very little air flow into the motor.
MAF: will output a low signal as there is very little air flow into the motor.

1000RPM at cruise
See above basically.

5000RPM at cruise
MAP: will output a low signal as there is a vacuum in the manifold.
AFM: will output a moderate signal as it will be sucking a fair amount of air into the motor
MAF: will output a moderate signal as it will be sucking a fair amount of air into the motor

1000RPM at WOT
MAP: will read high as there is no/very little vacuum in the manifold
AFM: will read a moderate/low signal as there is very little airflow through the motor
MAF: will read a moderate/low signal as there is very little airflow through the motor

5000RPM at WOT
MAP: will read high as there is no/very little vacuum in the manifold
AFM: will read high as there is a high airflow into the motor.
MAF: will read high as there is a high airflow into the motor.

The difference in a AFM and a MAF is that for a given volume of air the AFM will always register the same value. But a MAF will go up and down according to if it is a hot temperature or a cold temperature as this is how it takes into account air density. This is also why the MX-5 AFM has an air temp sender in it as it needs to take into account air temp.

So I hope from this you can see why you need to take into account the RPM in correcting the signal going into the EFI ECU as there is no direct correlation of AFM values to MAP value. AFM measure the exact quantity of air the motor is using, a MAP sensor allows the computer to calculate how much air the motor is using based on air temp,RPM, and manifold pressure.

[Boags]

I understand that completely Pupat, and now i see what you mean about RPM.

Boags

[Fatty]

ok thank for the detailed info.

but the curve can be corrected, thats not a problem. the biggest problem, as i see it, is this :
from what you are saying, the map sensor MAY output the same reading at , say for example, 1000rpm and 5000rpm, depending on throttle position.

if this is the case, i'm screwed

[The Pupat]

ok thank for the detailed info.

but the curve can be corrected, thats not a problem. the biggest problem, as i see it, is this :
from what you are saying, the map sensor MAY output the same reading at , say for example, 1000rpm and 5000rpm, depending on throttle position.

if this is the case, i'm screwed

Not may will. And yes unless you can get an RPM correction you are screwed.

[Boags]

MAF then eh?? LOL

That or RX7 AFM...

Boags

[Fatty]

well well well...
so there is a company in NZ that builds a piggyback called the MAP-ECU which fits the bill perfectly by replacing the stock afm with a map sensor

problem is , it's $400 US so for that money you might as well look at getting a full standalone for a few hundred more, like the adaptronic or something like that. or diy a megasquirt for even less.

anyway, so it can be done. i just wonder if there is a diy version of something similar to the MAP-ECU, without all the bells and whistles.

so the search continues

more info here, but in short \"The output data is computed from a 374 Zone table at over 100 times per second based on Manifold Absolute Pressure (-10 to +30 PSI) and RPM (0-8000).\"