Mass Air Flow 101
#1
Mass Air Flow 101
There has been a lot of misinformation spread around about the modifications we make to the non-DBW Stage 3 ECUs and why they are done. Here are the facts:
The ECU's microprocessor is fed the analog signal from the mass air flow sensor. The microprocessor's A/D convertor is ONLY capable of measuring inputs in the 0-5V range. Readings above 5V will be seen as 5V. The mass air flow sensor on the pre-2000 non-DWB A4 1.8Ts will output voltages in excess of 5V when pushing more than approximately 245-250hp worth of air.
So how is this extra airflow measured?
There is only ONE way to measure this extra airflow utilizing the stock MAF sensor- adjusting the gain on the input circuitry. The standard gain is 1:1. A 3V signal from the MAF sensor will be seen as 3V by the processor. We make adjustments to the gain (via ECU modifications) and rescale the MAF calibration table accordingly.
Can't these modifications be made in the code alone through sophisticated changes?
Ummm, no and it is ridiculous to claim so. Certain things are done in software, some things have to be done in hardware. IT IS IMPOSSIBLE TO CORRECTLY MEASURE MAF READINGS IN EXCESS OF THIS THRESHOLD THROUGH ECU MODIFICATIONS ALONE. For those that are interested, check out the datasheet for the microprocessor used in this application- the Intel 87C196KR.
So what happens if you don't do this?
Well, if you push the MAF sensor over 5V and don't do this, you will not be measuring any of the air in excess of the 5V threshold. This means all of your table lookups will be maxed at their 5V lookups. Any extra fuel or timing changes needed for this extra air will not happen. This is bad...
Does APR use this modification for ALL stage 3 applications?
No, not at all. The preferred method is to use a larger mass air flow sensor housing. Keep in mind that the difference between two different MAF measuring units in many cases is only the housing diameter itself. For instance, the S4 mass air flow unit is pretty much identical to the Golf 1.8T unit. The difference is the diameter of the housings that the sensor is in.
This being the case and since the sensor itself (the part of the mass air flow unit that contains the measuring element and all electronics) is by far the most expensive part of the unit, we can simply change the housing to allow for the measurement of varying levels of intake flow. This is what we have done for all transverse 1.8T applications. We cast a new sensor housing and utilize the stock MAF sensor element. Then the mass air flow lookup tables are modified accordingly to compensate for the changes. This is the ideal method to use. Unfortunately, this method is not practical for the A4/Passat applications.
The 2000+ DBW A4 1.8Ts have different mass air flow sensor than the non-DBW A4 1.8Ts. There is more headroom in these units. Though we are pretty much right at the limit of the stock unit with stage 3, it is not required to recalibrate the MAF sensor for higher flow values because of the extra headroom.
Brett
APR
The ECU's microprocessor is fed the analog signal from the mass air flow sensor. The microprocessor's A/D convertor is ONLY capable of measuring inputs in the 0-5V range. Readings above 5V will be seen as 5V. The mass air flow sensor on the pre-2000 non-DWB A4 1.8Ts will output voltages in excess of 5V when pushing more than approximately 245-250hp worth of air.
So how is this extra airflow measured?
There is only ONE way to measure this extra airflow utilizing the stock MAF sensor- adjusting the gain on the input circuitry. The standard gain is 1:1. A 3V signal from the MAF sensor will be seen as 3V by the processor. We make adjustments to the gain (via ECU modifications) and rescale the MAF calibration table accordingly.
Can't these modifications be made in the code alone through sophisticated changes?
Ummm, no and it is ridiculous to claim so. Certain things are done in software, some things have to be done in hardware. IT IS IMPOSSIBLE TO CORRECTLY MEASURE MAF READINGS IN EXCESS OF THIS THRESHOLD THROUGH ECU MODIFICATIONS ALONE. For those that are interested, check out the datasheet for the microprocessor used in this application- the Intel 87C196KR.
So what happens if you don't do this?
Well, if you push the MAF sensor over 5V and don't do this, you will not be measuring any of the air in excess of the 5V threshold. This means all of your table lookups will be maxed at their 5V lookups. Any extra fuel or timing changes needed for this extra air will not happen. This is bad...
Does APR use this modification for ALL stage 3 applications?
No, not at all. The preferred method is to use a larger mass air flow sensor housing. Keep in mind that the difference between two different MAF measuring units in many cases is only the housing diameter itself. For instance, the S4 mass air flow unit is pretty much identical to the Golf 1.8T unit. The difference is the diameter of the housings that the sensor is in.
This being the case and since the sensor itself (the part of the mass air flow unit that contains the measuring element and all electronics) is by far the most expensive part of the unit, we can simply change the housing to allow for the measurement of varying levels of intake flow. This is what we have done for all transverse 1.8T applications. We cast a new sensor housing and utilize the stock MAF sensor element. Then the mass air flow lookup tables are modified accordingly to compensate for the changes. This is the ideal method to use. Unfortunately, this method is not practical for the A4/Passat applications.
The 2000+ DBW A4 1.8Ts have different mass air flow sensor than the non-DBW A4 1.8Ts. There is more headroom in these units. Though we are pretty much right at the limit of the stock unit with stage 3, it is not required to recalibrate the MAF sensor for higher flow values because of the extra headroom.
Brett
APR
#2
Thanks Brett, one question...
...by "scaling down" the voltage values of the MAF in Pre-DBW applications, is there any issue with the resolution of the voltage? Can the A/D converter measure the "smaller" increments in voltage correctly (i.e. 0.1V change in MAF output used to be X change in CFMs, now 0.1V change in MAF output is X*1.5 or whatever in CFMs)?
Thanks..
Ash
98.5 1.8TQMS
APRv3, Cat-back LLtek/Brullen (custom downpipe)
Thanks..
Ash
98.5 1.8TQMS
APRv3, Cat-back LLtek/Brullen (custom downpipe)
#3
AudiWorld Expert
Brett, one question....
Why is "this method is not practical for the A4/Passat applications." In other words, why wouldn't you cast a new housing for the A4/Passat? Is it a space limited issue? Actually, not that I think about it, the construction of the tubing where the MAF sensor is, doesn't lend itself to change - is that correct?
This is the first I've read on this topic, but I'm almost done laughing at the thought of someone saying to compensate in software for the shortcomings of the hardware, without any hardware mods. That's just hilarious.
This is the first I've read on this topic, but I'm almost done laughing at the thought of someone saying to compensate in software for the shortcomings of the hardware, without any hardware mods. That's just hilarious.
#4
What would happen if you ran higher boost/power and didn't change the size of the MAF sensor housing
or recalibrate it? What would be the symptoms/side effects to the car? If it matters I am taking about a DBW TT.
My car is running really rich and I can't figure out why. Wondering if this could be the case. It happened when I put a free flowing(no muffler) 3" UUC Velocimax TT dual exhaust on.
Thanks!
My car is running really rich and I can't figure out why. Wondering if this could be the case. It happened when I put a free flowing(no muffler) 3" UUC Velocimax TT dual exhaust on.
Thanks!
#5
it just adds to your error of the air/fuel ration
the resolution is pretty good to begin with but you will have losses from the MAS to the intake.
the effects are negligible as long as you are in the bullpark
the effects are negligible as long as you are in the bullpark
#6
you couldn't maintain a proper fuel/air ratio
Likely you would go lean as you think there is only so many cfm of air and actually there is more. Of course, you could hack things up a little, but since there is no accurate way to measure, you can't actually get it reliably right.
Great info Brett.
One more tiny point is that with this mod on, the stock chip would read the MAF wrong at all airflows. So if you change back to a stock chip you have to remove the voltage divider. I'm just pointing this out, it isn't likely someone would do so, and I'm sure APR warns against it.
Cool picture Jenner.
Great info Brett.
One more tiny point is that with this mod on, the stock chip would read the MAF wrong at all airflows. So if you change back to a stock chip you have to remove the voltage divider. I'm just pointing this out, it isn't likely someone would do so, and I'm sure APR warns against it.
Cool picture Jenner.
Trending Topics
#8
Actually I reread your post. I am running lean at part throttle and the ECU is adding ~10% to the
long term fuel trim, so I am rich at WOT.
Could this be related to a dirty MAF? Could the solution Brett mentioned above fix this?
Thanks!
Could this be related to a dirty MAF? Could the solution Brett mentioned above fix this?
Thanks!