Miller Gen 3 MAF analysis

[ba114]

I agree with all said, if I get this figured out I'll release my calibration changes, part numbers etc to do the swap.

Ideally I'd actually like to learn to fully disassemble the code and write a maf function to replace the AFM logic completely. I've freed up a significant amount of space in the code by removing several unused fuel and timing tables so there is plenty of space to fit a standard 256 byte maf transfer.

[nando]

Nice work! I don't understand how these guys can still sell these half-baked MAF kits after all the years of horrible service and no customer support. The only thing really going for it is you can "tune" it yourself, but the software barely works so even that is pointless. It doesn't surprise me at all that the calibration is barely changed from stock (that is par for the course in the BMW tuning scene).

really, it's long past the time where as a community we should just figure out and share how the E30 DMEs work (much is already known - if not everything). it isn't rocket science and keeping secrets is not helping the community. There have been fits and starts, but the last few years have been rough on the forums - maybe we can make a comeback though.

I'll say realistically swapping from an AFM to a MAF is not going to be a huge difference in performance. But it is an option at least to replace aging mechanical sensors with an electronic one. The real key is to understand how the DME measures air mass / load, and then really any input can be made to work.

[ba114]

I have found this map in the DME.



From the logic and position in the binary, it looks like it is a fuel additive table based on RPM and Intake temp.

In the table on the left, at 13.5C intake temps there would be no modifier. At lower intake temps it adds fuel to adjust for the denser air and at higher IATs it removes fuel as the air is less dense. This would be because the AFM doesn't measure air mass, it measures air flow and requires the IAT values to calculate air density.

The table on the right contains the values from the Miller MAF file. This is one of only a few tables that is changed from a stock file for their MAF conversion. I don't really understand why they have done it this way.

To convert to MAF (which calculates mass already) I would set this table to all ONES so there there is never any compensation performed for changes in air temperature.

[ba114]

In case anyone cares, there is a possibly a more suitable maf, cheap, and in a 2.75"tube that would fit the stock intake boot that you can get a decent match to the curve with:

[ba114]

I thought the motoric's logic does not computer maf input. I was duped?

Yah people get it to "work"

https://sssquid.com/v3/e30-maf-conversion/

You can re-write the transfer functions to match a maf curve (with difficulty) in addition to removing the AFM compensations based on IAT (because MAF flow is already affected by air temp)

[moatilliatta]

I thought the motoric's logic does not computer maf input. I was duped?

Yah people get it to "work"

https://sssquid.com/v3/e30-maf-conversion/

[2mAn]

Keep at it! Id love to see more MAF conversions replacing old AFMs

[ba114]

Thought some people might find this interesting.
I reverse "engineered" the miller encryption on the W.A.R chip files for the miller gen 3 MAF (3" tube) for the M20 and found something interesting with the AFM/MAF scaling.

The chart below shows three curves:
1. Stock M20 AFM
2. Miller Gen 3 MAF
3. Ford Lightning MAF in 3" tube (same part used for the miller conversion as per https://www.m5board.com/threads/s38-...-heavy.322098/



As you can see there are a couple of issues:
- The scaling used for the Miller G3 substantially differs from the Ford Lightning MAF (1L3F-12B79-AB). Now i do believe they still use this MAF but happy to be proven wrong which would make this issue irrelevant.
- The Miller G3 values used in the transfer functions result in significant dips (5-10%) at at regular intervals. Maybe not an issue for most builds but a drop in 10% at 4.3v ~530 kg/hr could cause a lean spike for forced induction setups during spool or for N/A builds at peak RPM.

Anyway, thought people might like to see what goes on with this setup as im currently looking to design a MAF setup for these cars and was interested in what Miller had done and where it can be improved.
And, yes, i have found the tables that make corrections to the airflow measurement based on IAT and have zero'd those out because a MAF already takes into account the IAT in the measured airflow.