Im curious if anybody here has flowbenched the hitachi ME7 in any of the popular housings (85mm, 90mm etc) and with various intake setups (stock, rs4, stock+hole, open element) etc.

I'm still puzzling over MAF scaling because I'm getting odd results with my new 85mm MAF and stock airbox.

i would certainly be interested in this data.

i can tell you that Tony's scaling is extremely close to 2 well known top-shelf stg 3 tunes i've seen (setup for 85mm, stock box, paper element, hole in the bottom).  The difference is that they have tweaked a couple of areas in the low flow range using MLHFM rather than utilizing KFKHFM (which is set to all 1s).

I have a few beers in me so if what I'm about to say doesn't make sense, I apologize.

In my opinion in order to properly build a MAF transfer function, you need a flow bench with an infinitely variable diameter orifice.  They are not cheap.  I'll explain more, but bear with me.  

Back in the day, there was a company popular in the mustang community that did immensely stupid things like hack up MAF electronics in an attempt to allow for larger injectors without tuning the ECU.  They also built bigger meters, and provided transfer functions with each meter.  These sheets used to give you only a few data points for voltage by flow, so you could quickly blow up your new motor and feed the spare parts machine.

Flow benches work by measuring pressure drop across an orifice.  Pressure reference is captured with a manometer before and after the orifice (which has a known flow rate), and as a result, flow across the orifice is calculated.  Add a restriction upstream, and the pressure differential changes, and voila, you have measured the flow of whatever you put ahead of the orifice.

But what happens if the orifice is the flow restriction?  Well, whatever the orifice is calibrated to, is the actual flow rate.  Stick a maf in front of an orifice of a known flow rate that's lower than the max flow rate of the maf, measure the voltage, and you have your voltage measurement at that particular flow rate.  So for $75, some schmoe would load your meter on the bench, install an orifice in the flow bench, measure the meter's output, swap a larger orifice, measure the meter, etc.  I think you got 8 or 16 data points.

With an infinitely variable orifice, you can precisely increment the flow across the meter and control the air flow precisely at each flow rate you need a maf reading for.  These machines are expensive.

But what if you could build your own?

My idea requires some plumbing, a ~1000CFM fan, a PWM fan controller with a user-variable output, 2 ECU's, 2 bench harnesses, 2 KKL cables, a STOCK maf, and the other large MAF of your choice.

Stack both MAF's in series, and then duct the fan on the end a few feet away from the sensors, so the configuration is draw-through (the fan creates a vacuum).

On ECU 1, change MLHFM to be linear so that each cell has a unique value you can easily recognize.  .5, 1, 1.5, 2, 2.5, etc.  Connect the large aftermarket MAF.

On ECU 2, connect the STOCK maf.  Use the stock MLFHM.

Start logging both ECU's using Setzi's ME7 logger to dump the values of the MAF to stdout.

Start the fan, and attempt to run the fan where ECU 1 reads the value of the first cell (.5).  Now, look at the value on the ECU 2.  Record that value as the value appropriate for the first cell.  Increment the fan speed and repeat for each cell, or pick a random span of cells and interpolate between them.

Build your new MLHFM, and sanity check it against the stock MAF.  If you're really feeling randy, flip flop the MAFS and sanity check with the configuration reversed.

Obviously you will peg the stock meter at some point.  However, you should be able to interpolate above and beyond with no substantial issues IMO.  I have a few ideas to accurately vary the air flow to calibrate the meter above and beyond the stock meter's capability.  

But realistically I think this would work perfectly as most maf tuning time is spent at low flow rates for stupid things like accordion hose angles, etc.  Obviously you could attach any inlet in front of the mafs that you feel would be affecting the MAF.  Any change/variation could be measured and compensated for with reasonable accuracy I would think.

Just did this:

http://www.myaudis4.com/2015/02/05/hitachi-maf-sensor-values/

Any thoughts on why the recorded airflow values are coming out less than what is predicted based on the stock MLHFM curve?

kg/hr converted to g/s then used 13.4 cubic feet/lb to get to CFM for the predicted values.

MAF Flow Test

I think it was from std day conditions.  I varied the number some and to get the curve to shift enough to match with what has been measured it has to be changed quite a bit.

But interesting enough, idle (no airflow) voltage is right. So the discussion needs to be about why the numbers seem so off. What is the factory system doing differently.  If he has 12volt and 5volt, the correct ground and the sensor is reporting numbers that differ from the factory tables, why.

Are the stock values just so wrong and if so why did they do that. There is enough knowledge on this list that a few folks banging shit off each other should turn something up. You have someone willing to flow test, which in reality will be major help to everyone, how do we get the big brain eyes on this?  Sure maybe it doesn't matter because their tunes work, but think if they new the exact values on every MAF out there, wouldn't that take the tune to the next level?

Tory

i believe prom racing will still flow a sensor for you.   www.promracing.com

ive used them in the past for some of the hitachi slot style sensors i use on skylines.

if not i will search for another company i used back in the day.