Can this work for an M42 (MAF conversion)

[HaNasich]

taken from 1.6l Hyundai motor to fir a 2.5 M20 BMW motor.
does anyone has info about MAF reqs. of an M42 motor?



thanks


Ron.

[mcfir7]

did u try it ron?

[bmwman91]

Don't bother. There is no simple plug and play way to use a MAF on the M42, or any car that was designed with a VAM sensor for that matter. First, the tuning will only be valid at one air temperature since the MAF and VAM operate on different air flow quantities (kg/hr vs. m^3/hr). Second, particularly on 4 cylinder engines, the lack of valve overlap causes the MAF output to have very large voltage swings that the ECU cannot measure quickly enough to get an accurate reading on the air flow. That's the short of it. And no, you can't just stick an RC filter on the MAF output and expect it to work, you will run so lean that you get fuel cut.

Getting a car to RUN with a MAF is easy. Getting it to run as well as stock is incredibly difficult, if not impossible for anyone that isn't comfortable with designing mixed analog/digital circuitry. A stock M42 stands to gain no power from a MAF anyway, the stock VAM sensor is plenty big and the flap door poses no significant restriction at all.

[Jasonixo]

Are there any VAF to MAF conversions that supply their own IAT (onboard or remote) and modify the conversion rate to compensate for temperature?  I'm not an electronics guy, so I'm sure there is more to it, but why can't a self-contained MAF system simply supply an accurate, temperature-corrected VAF signal to the DME?

Wouldn't this create a somewhat universal system that can be installed on any VAF system by re-pinning the AFM plug?

My motivation to replace my AFM lies in the disconcerting rattle emanating from it.  This M42 driveline in my garage will see new life in my 1976 2002.

[bmwman91]

I am not 100% sure what all Split Second has these days, but I think they have a system that can deal with temperature. That's the easy part. The trickier part is that the ECU still expects an air temperature input signal, and just about every IAT thermistor will be different between models. The M10, M20B25, M20B27 and M30 AFMs all have different IAT thermistors (I measured their resistance curves). A universal one is tricky in that sense.

Then there is the fact that every AFM and MAF has a different transfer function (air flow vs voltage relationship). That is tough to guess at, and you will really want to dial a universal type unit in on a dyno, which is $$$.

Finally, there is the issue of intake resonances and back-flow. A MAF can measure those things, but the ECU just cannot handle a sinusoidally pulsing voltage input since it is designed for the AFM which damps those things out. Look up Nyquist's Theorem and Aliasing. At least on the M42, you can't just plug in a MAF. The converter I built involved years of experimenting, and ultimately a solid 2000+ lines of microcontroller code and careful digital+analog filtering to get right. In short there is literally, honest to god, no way to just plug a MAF into the M42 and expect it to work anywhere as well as the stock AFM.

Just find a used AFM on the forum for $30. No big expense, and it swaps right in without any nonsense. Unless you are going to try to make 200+ HP with your M42, the stock AFM isn't causing any loss of power.

[dimiras]

I am not 100% sure what all Split Second has these days, but I think they have a system that can deal with temperature. That's the easy part. The trickier part is that the ECU still expects an air temperature input signal, and just about every IAT thermistor will be different between models. The M10, M20B25, M20B27 and M30 AFMs all have different IAT thermistors (I measured their resistance curves). A universal one is tricky in that sense.

Then there is the fact that every AFM and MAF has a different transfer function (air flow vs voltage relationship). That is tough to guess at, and you will really want to dial a universal type unit in on a dyno, which is $$$.

Finally, there is the issue of intake resonances and back-flow. A MAF can measure those things, but the ECU just cannot handle a sinusoidally pulsing voltage input since it is designed for the AFM which damps those things out. Look up Nyquist's Theorem and Aliasing. At least on the M42, you can't just plug in a MAF. The converter I built involved years of experimenting, and ultimately a solid 2000+ lines of microcontroller code and careful digital+analog filtering to get right. In short there is literally, honest to god, no way to just plug a MAF into the M42 and expect it to work anywhere as well as the stock AFM.

Just find a used AFM on the forum for $30. No big expense, and it swaps right in without any nonsense. Unless you are going to try to make 200+ HP with your M42, the stock AFM isn't causing any loss of power.

I am trying to run m42 (In a euro M43 1.8 Z3) with itbs and for that I am running a motronic 1.7.3 and will need to run without the AFM.
So I want to use the AFM Limp Mode table (Alpha N), but I want to keep the IAT corrections.
Is there any IAT sensor with a similar curve  to the thermistor in the AFM that I could use? (the one from the m44 airbox  for example)
Or could you help me figure out where is the transfer function in the ecu code? so I could change it to the new sensor

Thanks