NefMoto

What do you see at idle?  I always see 3.5-5.5 it seems...  this is the same with bone stock cars.  Regardless of MAF housing diameter or placement.  every car i log here at 6000ft reads this at idle.

As I have been testing various MAF setups on my single car in both draw through and blow through I have been seeing a lot of weird things.

First and foremost, how is it possible that with the same map values I could have a 73mm housing read the same as a 88mm (1.47 times more area)?

I recently switched back to blow through again as i never really got into fine tuning it.  A 73mm is reading LOWER post turbo than a 88mm was pre turbo, same car, same conditions, same boost levels.  At 15 psi and4000rpm it was reading 120gs (4.63 G/S @ idle).  The 88mm reads 160 pre-turbo at the same point (4.63 G/S @ idle).  Stock MLHFM and a KFKHFM of all 1s to remove further variables.  How is it the same at idle yet lower at high flow, yet a smaller housing?  I get that the air is now compressed and heated, but it is still the same volume of air.

Every shred of info i have read about blow through says that the readings will be skewed HIGHER.  This is the opposite of what is happening to my car. 

Next observation is that if I scale the MAF voltage table (MLHFM) at all, even 1%+ it will read 9-17 G/S and be creating 40-50% engine load at idle with AFRs in the 10s.

Today i am going to try two things, first a maf scale that progresses from 0.01% to 15% positive of the stock map across the 512 points and then also a standard upscale of the map with KFKHFM changes to just low loads and idle.

I know very few people are doing anything blow through on the cars, but if I am not somehow screwing things up, this could open up a lot of options.  Imagine a stock housing that you can run more air through than a 90mm and still have resolution.

thoughts?  anyone know the answers to these events already?

I'm guessing that at idle, the MAF is at its minimum value, and you are seeing a fixed additive offset. I recall seeing something like in the FR,  but I dont recall the details. I'll look it up time permitting.

thats what I was thinking to.  like the 4.5ish G/S i see is just locked in as an assumption because the air flow is so low?

I take it back. I was thinking of MLOFS, which is zero for hitachi mafs. There is some other sort of fixed offset hidden somewhere, but I can't find it now.

btw, this is all the same regardless of hitachi or bosch so far in my testing.

I still keep going back to the bosch sensor though.  that is what I am currently on, stock 73mm housing with stretched 3" hoses within my 3" charge piping.

Also, try logging MAF voltage directly. uhfm_w in setzi's logger.

I was under the impression that a blow through setup would read lower?  Not quite up to speed on these setups though.

Either way I can't wait to change over to the HPX sensor

jared - can you step back a second.  did you just say you are running your single-turbo beast on a 73mm MAF setup and it runs fine?  if so, why are we all bothering to do anything to the maf setups?  what did i miss in your comment? (lol...)

Best its ever ran!  Pardon the haggard ass looking piping, I have yet to decide my final configuration as you can tell by all this testing.  Once everything is done the intake will be a one piece CF tube and the charge will all be thin wall tig-ed stainless like my first section there.

post turbo, post BOV, pre inter-cooler.  It seems to have head room to over 800CHP.  I exploded a stock housing (my original stock one from 2000) at 35psi on e85 ;D

fyi, twin turbo guys can run two tdi maf housings with a stock sensor and put a 2.25 ohm resistor in their clips, shared wiring like on a touareg and like that you can be blow through for testing/running bovs, dual intake, whatever... ;)

with all that said, I am obviously still figuring out some of the nuances of this.  blow through MAFs on other platforms obviously do different things.  all their info has to just be taken as info, not facts regarding our ECUs and even sensors themselves so please approach this with caution.  When I first put it to blow through it was going lean at boost onset until I cleaned up other areas of my tune.

(http://www.vlmspec.com/assets/images/maf.jpg)

(http://www.vlmspec.com/assets/images/maf2.jpg)

(http://www.vlmspec.com/assets/images/maf3.jpg)

ok, so forgive my speculation here, but what you've said here would directly imply we do not require larger maf's nor maf scaling...  there's been a ton of discussion about this and frankly it seems like a lot of hot-air and wasted time...     nyet: any thoughts?

i've compered 2 different diameter maf housings one from bosch 70mm maf (1.8t 180ps) and the other from 80mm bosch maf (1.8t 225ps). It seams that the scaling is not linear. In lower voltages the bigger housing reads less than the smaler about 8% and in higher it keeps readind more and more increasing up to +25%. I have attached some pictures. The first one is the % difference between the 2 and the second the 2 different callibrations. These are oem values taken from original files. Also i think there is an offset MLOFS (200kg/h) in the damos that i have.

Ran a test for everyone, went stock for stock and took a stock m-box file, modified as few maps as possible to make it even run on my car.  Results are the same.  A housing reads lower post turbo than pre, significantly lower.  Also I can confirm it reads 60-70 G/S lower across the range than a un-scaled 90mm that is pre turbo.  I wanted to make sure that no other areas of my highly changed files were causing this.

I couldnt test this at anything above 18 psi though as with the minimal maps changed I was going crazy lean.  At 18 psi it pegged readings at 137 G/S and my AFRs were in the 16-17 range according to my wideband guage.  I obviously had to shut down.  I didnt get any really clean logs of the behavior but I feel I have backed it up now on my car atleast buy taking a totally different file and tuning strategy and getting the same results.

hey - so have you tried the stock/stock pull-through as-in basically bone stock?

"During idle conditions, a selection is made between the measured air mass flow and the maximum possible air mass flow at this operating point, mldmx_w (taken at a height of -500 m and a temperature of -40°C) weighted by the multiplication factor FKMSHFM".

Doug

You should also check KFMLDMN and FWMLHFMMN.

some very weird behavior...

now it seems I am discovering what i overlooked before.  I am still working with this near stock file to try and lock down blow through tuning.  It never requests lambda of less than 1?  Why?

If I unplug the maf it behaves like it should in regards to fueling.

In these pulls i didnt lift, regardless of risk of engine failure as I was running e85 for protection against detonation.  I was trying to get longer pulls but as the wideband indicated 19.5:1 (lol) it just cut out. 

Am i missing something?

Are you using BTS? Log calc egt and compare with your TABGBTS

yes, like I said this is a new behavior thats isolated to the blow through.  in draw through it doesnt happen, in speed density (maf unplugged) it doesnt happen.

calc egt is well over tabgbts

Mayby the ecu doesn't read enough load so it thinks that you are in low load condition and doesn't enable the high load fuel enrich mode. I think you should try to rescale the maf properly before you proceed any further.

Agreed. Load is RIDICULOUSLY low. What AFR should you be seeing according to your BTS table?

that is actually already scaled, that is a stock 73mm housing scaled up 15% across the board.  issue is that anything more and it becomes so rich at idle that you literally cant remove enough fuel.

also, if I scale too much to get the desired maf readings at wot and in boost then the idle becomes the issue as I was talking about in the first post. huge g/s readings and engine load at idle with severe rich conditions to the point of practically needing a negative krkte value...

so far no one has told me what they see at idle in regards to g/s either.  do some peoples cars idle with that high of air readings?  cause none do that I ever see.

I guess the point of all of this is that i know how to make if work for me at this point on my car.  I am trying to understand how and why the maf works this way in blow through.  effectively opposite of what every other blow through setup seems to do.

now i am talking to myself...  but another thought that i should point out.

my issue with insanely low krkte values is this, maybe you get it in check with the maf plugged in, but as soon as its taken out of the mix (unplugged or fails) the car is running off a crazy low value and is so lean it cant even run.  the crazy scaling of the maf is needed to be back in the correct range again in regards to fueling calculations.

remember i base everything off my obsessively calibrated wideband. 

i even went back to stock values on like 9.5:1 for bts.  my old maps that i had working on the car great with draw through were 11.5-11.6 at full load.

this has to be a maf thing, it has to be scale related, i just can figure out how to achieve the right balance without modding other maps, i want to know that air is being properly read.

Another think to consider is that maf is a sensor designed to suck air from the outer enviroment. If you are using it as blow through the air mesured is far more hotter than the air it should read. So there has to be a map to correct this. Sorry but i cant express correctly in english. For example if you read 4V in suck setup with a stantar enviromental temperature the 4V reading in blow through setup won't be the same because the air temp is now +20 degrees (especially in wot with 1+bar of boost). I think some mafs have an intergraded circuit for calculating the air loss from higher temps. So this circuit affects the maf readings. So this is another reason you can't use just a scaling of 15% across the board. I also wrote before that the scaling for bigger maf is not fixed across the board but varies as the airflow increaces. Sorry everyody for the big topic. I hope that i helped you.

Yea, but what is your BTS at 60% load? Stock is pretty much lambda 1 until 122%

hmmm... true... I will have to look.  which then brings us back to the maf not accurately indicating load.

Which should have showed up in your fuel trims for part/idle, and in wideband vs req AFR WOT deviation.

Also, modern MAFs should be properly temp corrected, even with +20F variations.

If in doubt, check the sensor datasheet.

I meant +20C not Fareneit. Mayby there is some limit until witch temp can be corrected. In a wot run with no correct afr and ignition as in this case, it can reach 100+C max IAT. I think that a sensor designed for reading air from the enviroment won't correct in such temperatures.

Excellent point. On my K04 car, I rarely see over 150F IATs (65C ish), but I can certainly see it is possible for a blowthrough setup to get air temps that are beyond the spec of MAFs intended for ambient.

The temperature correction of the MAF voltage is done by the sensor itself if I recall. I don't think the ECU has any involvement in temperature correcting the MAF voltage. I agree that it is possible the high intake temperature could be an issue with the stock MAF sensor in a blow through setup.

In case it hasn't been mentioned before be sure to check KFMLDMN, FWMLHFMMN, and MLOFS. They may give you some hints about your MAF readings at idle.

I am going to try a rs4 style scale of the bosch maf with a large scale on positive voltage and a minor scale on negative...

also, I dont have FWMLHFMMN and have never seen it in my stuff, FR says "Weighting factor fu ¨ r minimum threshold HFM-air masses in, reliance d.Ho · He"


???

Why would you do that?

if your problems are temp related, it wont help to dick with the MAF curve regardless.

if they aren't temp related, why not start with a stock curve, and start logging stfts at various operating points and tweak kfkhfm, NOT MLHFM

FWMLHFMMN is the scale factor for KFMLDMN

who said the issues are temp related Nye?

I dont think they are at all. atleast not at these boost levels and ambient temps.

I previously used kfkhfm to fix all this, I am trying to get to the root of why blow through reacts this way and know how to make it work without a bunch of correction map intervention.

Could it be flow related? Perhaps the flow is not well developed, depending on speed.

Do you have a flow straightener post compressor outlet?

yes, the maf is an oem housing with straightener and screen.  its about 18 inches post turbo

i will try and get a new log or two tonight

I am starting with the biggest issue, the low maf readings.  I can still get data even with idle on the 10s afr wise.  id rather waste fuel then melt down.

So just a quick summary, you moved your MAF from after the air filter to after the turbo. And the problems you now have are, at idle your MAF readings are too high and under high load your MAF readings are too low?

Looking at the pictures you posted earlier, it looks like you have a coupler hose going from a smaller diameter to a larger diameter immediately before the MAF sensor. I am wondering if the change in diameter could cause air turbulence dependent on air velocity which is skewing your MAF readings.

Your summary is correct.  I agree on the coupler thing.  I need to do post ic again with no transition.

I kinda picture the smaller id pointing air through the maf housing.

As far as I know, changing diameter or direction introduces turbulence. And if I recall correctly, turbulence will increase the cooling effect on the MAF hot wire, which will cause the reading to be higher.

I didn't read all this but are you aware of the fact that the Bosch and Hitachi MAF sensors are only speced to see a max temp of 120°C? At 6000ft with a AMP of ~ 810hPa and 18PSI boost you are at a absolute PR of ~ 2.7 which will easily exceed this temperature no matter how efficient the compressor is and how low the ambient temps are.
Stock cars are limited to a temp post compressor of ~150°C just to give you an idea.
I have never tried this but I can't imagine it's a good idea to place a MAF sensor in such a harsh enviroment.   

:o You moved your MAF from after the air filter to after the turbo?  :o

Sorry, I didn't appreciate that just by scanning your original post. I'm not surprised that you're having difficulties on grounds of temperature, pulsations, return air flow on overrun, air density, etc. I have never come across such a thing. Motronic ME7.x seems to have quite enough compensations for the usual pre-turbo MAF setup that I would expect putting it after the turbo would introduce so many more that it would be too difficult to deal with. I admire you greatly for trying it.

Not particularly helpful comments though. I apologize.

Doug

I think I solved it for the most part.

I am now getting proper readings, I will need to drive and log some more then I will post up results.  For starters, you have to massively scale mlhfm.

Bob, thanks for the info, that is why I have always intended to go to an aftermarket maf that is designed for this style operation.

progress... ;D  need to do some big kfkhfm tuning.  at least its following BTS again. 

I am also thinking about putting a temp probe post compressor.  fwiw I never see IAT post IC higher than 110 F or so.

Looking decent! You could still use quite a bit more load/MAF... see if you can get it near 190 at peak torque and g/sec maybe around 300 at redline.. (and pull back KRKTE accordingly). maybe another 10-12% MLHFM and you should be gtg.

I am contemplating blow thru as well with an HPX ford style MAF. I am very interested in the data here.

Do you intend to acquire a MAF designed for this operation?

yes, I actually have a second hand pro-m but its garbage, its calibrated for a totally different setup within the actual sensor so no amount of tuning on my part will fix it.  I intend to buy a new one and have it configured for this application

I have been meaning to buy a proper setup but decided to revisit stock sensors before doing so.

What sensor do you plan on running?

I am looking at this in 3" tubing.

http://www.vmptuning.com/store/index.php?p=product&id=193&parent=30

Please tell me more! I would guess the primary difference in a MAF designed for blowthrough is different operating temp compensation characteristics... anything else?

I honestly dont consider myself knowledgeable with MAF sensors in the grand scheme of things.  Mark at ASP turned me onto them because they calibrate the specific sensor.  You can tell them what your peak intended airflow is, your injector size and how you plan to plumb it.  They then set one up for you, they calibrate the actual sensor on a infinitely variable orifice flow bench and turn out a voltage transfer data sheet for your needs.  People who have ones that are properly setup rant and rave about it being the best thing ever.

They make all sorts of maf solutions.  I bought a 3" boost tube that was off a 3000gt running similar injectors, no dice.  also the previous owner didnt pony up for the detailed voltage data.

Interesting. Can you have them flowbench your own setup? Or only theirs?

They say they can re-calibrate any sensor.

Interesting. I would be tempted to send in a stock hitachi maf in an 85mm housing...

but what about intake tract? It makes a HUGE diff... airbox vs open element.

Do you send them the whole fucking airbox? :)

Let me know if you think this is feasible; I'm willing to partially fund it.

ASP's site is dead.

I honestly would rather just buy one of their standard offerings.  If you look at their boost tubes (blow through) they are all 1 foot of tube and they say thats so that the air collects right.

their pre turbo stuff seems to be specific also, they say that if you run a cone filter to send that in with your maf.  I bet if you sent them an airbox they could do it up right.

oh and btw http://www.promracing.com/

http://www.autospeed.us/

they changed it a while back, mark slazak suggested pro-m.  it seems mark knows jeff atwood (very skilled motronic tuner) who I am assuming is the son of bob atwood, who started pro-m.    Jeff has made flex fuel and blow through stuff work good on the bluewater R32 and other cars.  all those fueling kits spec out pro-m maf setups.

something tells me that we have over looked a great tuning resource here...

the x-axis for mlhfm is voltage, does anyone know what the y-axis is with the units?

I would assume the mappack is kg/hr. Have you checked it?

yes, thanks for the heads up.  the measurement is in kg/hr...

BTW all air flow measurements in ME are kg/hr... just a heads up so there isnt g/sec confusion.

i think we need to organize and send a 85mm maf + hitachi/bosch sensor to Pro-M to get exact calibration values...

In the sake of keeping people updated...

At this time I have stopped running blow through, at least with OEM sourced sensors.  I have moved back to a pre turbo maf and just deleted my BOV for now since I dont actually need it.  The only runs the car will get, will be drag or dyno until next spring, it is off the hook for daily driving due to weather.

FWIW it is running very good in this fashion, much better than pre turbo with BOV (obvious loss of metered air) and all and all better then blow through.  I base this off my overall driving impression, stability in the tune, adaptability and perceived safety.

I want to be clear that all 3 styles of running it worked, me7 is amazing in what it is capable of adapting to with the right maps tuned.  Do not be horribly afraid to try blow through, stock sensor or otherwise.  I am confident that if i revisit it with a sensor that is designed for the application that it will be just like tuning any other maf.  For now my car sounds like an old 5 cyl IMSA car when I actually do lift to shift.  :)

So this is not totally relevant to me7 but i am starting to tune a hitachi sensor(ford 0-5v style) on my me3.3.1 M3 turbo this winter.  Ive been reading up a little on these sensors(this post) and the "calibration".  What i am trying to understand is, how does prom "calibrate" these to a specific injector size etc? what are they doing?

my current thoughts are that they calibrate the sensors internally, to match injectors as said in this thread.  They look at the internal calc of the sensor to scale to an injector size to remove the tuning aspect on the actual car's ECU.  just scaling.

how do these sensors work?

I think its a voltage input to the sensor, and it measures how much voltage is needed to satisfy the hot wire to a certain temp.  does a calculation and gives a voltage output.  this calculation that is done is modified by companies like prom to match the factory sensor, or scaled in a way to make a plug and play solution with larger injectors, and working with factory ecu with little modification to the actual ecu tune?

if you have the full option of ecu tuning, like us.  we can use the factory ford transfer function for the specific maf sensor(found in a ford .bin file) and find a way too get a base transfer in our ecu's. 

the only reason to have a prom maf is to have a supplied transfer function chart, which would be awesome to KNOW what that maf is doing in that car with YOUR piping. 

I intend on running blowthrough with my M3, theres a company called TRM that does just this with thier OBD1 tunes, though id prefer to learn this on my own, seems like fun!

any thoughts or comments on this?

Bringing up an old thread here... but I'm in the process of setting up and tuning my single turbo S4 on blow through maf right now...
I'm also seeing the same thing that was mentioned in this thread... with the stock sensor I'm seeing less airflow measured under boost than there should be.
I'm going to scale the table for starters.

Agreed. I think the challenge with blowthrough is getting an accurate transfer function chart, because it is (as far as i know) difficult to reproduce blowthrough-like airflow characteristics on a bench w/o an engine dyno.

Perhaps what you could do is stack a known maf onto your intake pre-turbo along with the blowthrow maf, and use that to calibrate..

I "think" I might have come up with a formula to correctly scale the MAF table for blow through... it's not linear...
I will test it this weekend and report back!

Pretty cool idea.

I'm actually going to do this as well this weekend if I can borrow a Hitachi MAF from someone.
I want to leave my blow through sensor in place so nothing changes.

Is there an additional 0-5V input on the ME7 that I could hijack for testing? It would be cool to be able to log both maf sensors at the exact same time...

If you have a Zeitronix ZT-2 you can hook up one of the MAF's to one of the ZT-2's auxiliary inputs (it's got two), and synchronize them via RPM to construct the linearization curve for the blow-though.

I have an lc-1... so no built in data logging inputs :-(

You might be able to use rear O2 inputs if you have them coded out.

They are coded out.
Will they take a 0-5V signal?
Shit if they do I'll use one for wideband like on dsmlink :-)

Yea i've always wanted to set up my wideband to feed into the rear O2 inputs but never got around to trying it out... let us know :)