RomRaider

v6 updated in first post

I got bit annoyed with previous MAF scaling excel sheets because when using resistor, the top end of the MAF values overshoots quite a bit. Even thought the MAF transfer curve is said to be third order polynomial, I have found that the fourth order curve is much more closer to real MAF transfer curves. So I made my own version using the fourth order polynomial curve estimation in Excel and have used that quite a lot.

So here is my excel to adjusting the MAF transfer curve for different resistors:


Should be easy to use. Paste the MAF map from RomRaider. Choose the resistor value you want to use, the resulting curve should be shown instantly. Then just copy and paste the new map to RomRaider.

Result is not perfect but I have used this with m52, m52tu, m54 and s50 MAF's with 1k to 2.2k resistors and no other MAF map tweaking required. Just apply the MAF map and tune the fuel and ignition.

This is a great thread. Keep up the good work guys.

ba114 thanks for this... was going to buy a new stock MAF will now just get a cheaper nissan one..!

R.E. your knock tables comments, can we use the ms41.2 knock tables for the alloy block? I copied the MS41.0 knock tables accross.

Hello. I would like to use nissan (hitachi) MAF sensor instead stock m52, because my one bad and new one a lot more expensive. I've bought a new original nissan and its time to design a flange. As I can see, on original maf BMW, sensor element in the geometric center of the pipe. Does it necessary to place sensor in the center of tube? Its looks like impossible for 3.5" inch tube without small offset (around 2-3mm).

Theoretically any spot within the circle should have the same airflow, but because of the potential turbulence near the walls I would suggest to put the sensor as close to the center as it possible without any offset.

Very big thanks ba114 for that topic. Good choice when your oem bmw fan worn. My current setup 3.5 inch alu tube with tig welded custom cnc machined bungle and nissan maf. I just edit main fuel map around idle and higher wot enrichment.
But does it necessary to use honeycomb befor maf sensor? If yes, on ebay its 2 different sizes, which could be better: https://www.ebay.com/itm/3-5-OD-x-5-Airflow-Straightener-Screen-187-honeycomb-cell-mass-air-flow/263049773966?hash=item3d3efd038e:g:n-oAAOSwFMdaVTZr or https://www.ebay.com/itm/3-5-OD-x-5-Airflow-Straightener-Screen-25-honeycomb-cell-mass-air-flow/261845173266?hash=item3cf7304412:g:ghwAAOSwn9VaVTbg
And if I would like to rescale my maf setup, which way is better, cause at this forum at least 2 different methods.

I don't know what's wrong with my install of RR (0.5.9 RC 3). But I just can't figure out how to copy the entire 16x16 MAF table into Excel. When I select 'Copy Table' it only copies the first row.

If I select all cells, and 'Ctrl+C', it only copies a few rows. But the number of rows it copies isn't always the same.

Can anyone help? I want to setup a basemap for my 803 MAF with a 2.2Kohm resistor and 42lb injectors on an otherwise stock MS41.2.

Select the cells to copy, table Edit menu - Copy Selection

16x16 is a malformed table display so I suspect Copy Table won't work as the axis are Null.
It should be defined as a 1 x 256 table with a VDC axis.

I can confirm that the copy selection method works.

Yeah, I know the 16x16 table is kind of weird.

When I "Select the cells to copy, table Edit menu - Copy Selection" it only copies some of the rows. Like the 1st 5-7 rows and it varies. I tried to highlight the cells slower and not sure if it had any effect.

I noticed the Alpina MAF scaling (837kg/hr) gets a perfect Rsquared value of 1 using a 4th order polynomial trendline in Excel. This is probably what pazi88 meant when he suggested using it.

However, the stock MAF scaling (725kg/hr) needs a 5th order polynomial to get a perfect Rsquared value of 1.

So, I agree with pazi88 that we should be using a 4th order poly at least. But maybe we should use a 5th order poly for even more perfect results?

Yes the fifth order gets better Rsquared. But the estimated curve beyond 5 volt is totally wrong. Which is much more important to get right than Rsqared value. Using 4th order the estimated curve is close to what the airflow is in real life. When using the fuel map tuning method I use, the AFR's seem to be spot on past the stock 725kg/h airflows usimg that 4th order estimation. 3rd order estimation gives too big airflows resulting too rich AFR after the stock 725kg/h airflow. I may have some pics to show what I mean if I search my laptop.

Maybe your rr_system.log could provide a clue as to why the incomplete copy.

Great info here! Im planning on possibly running a 350z MAF on my M52 Turbo E30. How much horsepower are these MAFs good for in a 3in tube with forced induction?