Using non Ferrari Spec MAFs! - Application to stoogeness!!!

This document can be used by anyone with a Motronic 2.5 (Mondial t/348) or 2.7 (Mondial t, 348, 355) to:
1) replace a Ferrari MAF with one of the same series but different MAP
2) survive a broken MAF until a replacement is found
3) get home

One of my MAFs blew up - full of water! - what a MAFtastrophe!

There is a great post by No doubt describing a pretext to this post http://www.ferrarichat.com/forum/showthread.php?t=191675&highlight=280+212+018

Ferrari MAFs (Bosch 0 280 212 018) are a bunch of cash and difficult to get. Saab, BMW and Volvo's use MAFs similar to the Ferrari ones, ending in -016, 017 and 019 but have different MAF curves (these curves map the airflow to a voltage sent to the ECU) stored in their logic, evidenced by the electronic boards have different codes, while the MAF bodies have the same codes. I found a -017 MAF from a local scrap yard for $30. When plugged in it did not even start the bank. Dialing the CO screw to the highest OHM resistance, just barely improved the running to a very rough idle. Its curve was very different, producing a signal of less than 1.8v at idle, when the Ferrari MAF produces at least 2.8v (resulting in a VERY lean mixture that will not even fire)

1) Took the lid off the MAF body (see the aforementioned post) and rolled back the plug boots for both MAFs
2) Unsoldered/cut the leftmost and middle pins (these connect the circuit board to the wire) meaning the MAF will not be able to read the airflow, producing a ZERO reference voltage on its signal wire (pin 3 on the connector with the ECU), but the remaining pins, power, ground, reference and CO screw returned the right signals. The picture is somewhat unclear, but the leftmost pin is disconnected from the lower body sensor pins, the next 2 pins are joined into one, then the next one is also disconnected from the lower body pins, then the next two are connected
3) connected Pin 3 between the working MAF on bank 1-4 and the replacement one on bank 5-8, ensuring that the 5-8 bank ECU got all the right signals from the 5-8 MAF, but the airflow curve of the 1-4 MAF (See picture with pinout description)
4) tidied everything up
5) re-tuned the 5-8 MAF to the CO screw resistance according to part 4 in the main body of http://www.ferrarichat.com/forum/showthread.php?t=202434
6) reset the ECU's
7) started her up and back to normal we went!

While I am sure this will work as well as with two MAFs, I have ordered a -018 which I will refit. Just to cover myself in the case that my good MAF also has water within it...

Enclosed are pictures of the damaged MAF (compare to the ones of ND's post), the disconnected pins inside the MAF body and the pins to connect between the MAFs

I hope this gets you out of trouble if you drop a MAF for whatever reason.

I'm here to help if you need it - PM me!

Marco
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Incidentally, the 0 280 212 017 from the Saab is an HLM 2/4.7 120° degrees Celsius sensor. Meaning it uses a reference temperature of 120 degrees celsius on the resistor circuit. This temperature varies proportionally to airflow, diving a variable resistance on a reference voltage at a rate determined by the curves stored in the AD/DA converter circuitry in the MAF body.

The reference voltage is always constant for the family of sensors (the 0 280 212 XXX bit in the p/n) as is the logic in the hybrid AD/DA circuit board, but the operating temperature of the sensor and the curve mapping change for the individual MAF part number (the X XXX XXX 017/18/19/20 bit in the p/n).

The a lower reference temperature for the same map, will make the MAF resolve airflows in a lower range (smaller engines etc.), higher references do engines with bigger flows (eg. turbo or larger volume, such as the 1.7 litre per half bank of the 348).

My mod effectively bypassed the sensor circuitry, disconnecting the ping inside the MAF made it generate a 0 volt output, so there was no possibility of of generating a higher voltage then the other MAF (the good one) and making both ECU's believe there was more air flowing than was real (a richer mixture).

I found that the Bosch MAFs starting with 0280 212 018/9/22 all have the same operating volumes and flow range temperatures "HLM 2/4.7 155°. The difference between them is the absolute resistance of the trimmable potentiometer (CO screw). You can de-solder one from the 348's - 018 MAF into a -019 MAF and you have the same!!! Alternatively, if you know your MAF CO value you can buy a normal resistor with a 10 watt rating and solder it in the same place for the same, but no longer variable, effect.

The resistors need to be fitted, as removing them drops the output on the CO pin to 0 volts and kicks off a CEL error as well as going to limp-home mode, but with O2's fitted a mid range value resistor of around 400 ohms will work fine.

If you want to tune the CO screw to fit your car perfectly as opposed to the factory default (which I have never been able to confirm with Ferrari and is speculated in this message board), warm the car up for at least 10 minutes, go to open loop (disconnect sensors), record your current MAF CO screw setting, set both CO screws to around 250 ohms (rich), then while monitoring the voltage on your O2 sensor, increase the CO screw resistance until you see a SUDDEN big drop from around 8/900 mVolts to 50/100 mVolts, then move back to just when it pops back over to 8/900 mV.

This will put you just on the rich side of stoichiometry, ensuring you don't risk an over lean situation if you go open loop. Compare your new MAF values and post them below... There is a risk that setting 383 ohm, in open loop may be quite off on the lean/rich side. For me the 1-4 MAF was at 420, the 5-8 at 530. This made the 383 very rich and I could tell how the idling to 3000 RPM smoothed out on the new MAF settings (My car was a stock non cat/open loop car).

The Maintenance manual does not refer to a CO screw value, but to set them so to deliver a CO of .7% and an HC of around a few hundred PPM (don't have the manual here). This is emissions driven and not set to an empirical value. Using the O2 sensor, is saving you from having to buy expensive gas analysis tools, as at stoichiometry 465mV the typical CO output is around .6% ;-)

Marco