john banks

Using a microcontroller (with onboard ADCs say 90S4433) to take the MAP signal, along with a K-type thermocouple after the intercooler interfaced through an AD595AQ, you could simulate a MAF sensor signal. The JECS maps already use RPM so this might not be required as an input.

Ideas/comments welcome.

Would be good to remove the MAF, esp for those with FMIC/ind kits.

[Edited by john banks - 6/23/2002 2:18:31 PM]

David_Wallis

Would be very interested in this... Hmm no airfilter... just turbo...

(David thinks john may have killed his maf...! )
Let me know!!!

dowser

What are the real benefits of using the map rather than maf? #1 has to be service life, but what else?

Richard

Andy.F

Pro's and Cons really.

A MAF based system will self-adapt better to any tuning mods such as change of exhaust. This is due to the increased air flow being detected and the ECU adding the appropriate fuel. It does however have a flow limitation imposed upon it by the MAF meter.

MAP based will not self-adapt. It requires changes to the fuel/ignition mapping. Fitting a free flow exhaust for example will cause it to run weak as the MAP pressure will not increase, it may actually decrease due to the better exhaust flow.

No dig intended but MAP base is good for tuners as any change to the car and it requires a remap.

Most big power installations run MAP/TPS/RPM based fuel and ignition maps or speed density as it is generally known.

I think the intake temp post intercooler is an important one to include in the mapping. Some aftermarket ECU's don't compensate for charge temp. This is fine if there is not much rise, such as low boost FMIC installations.
On a Subaru, running 20+ psi boost on a TMIC, charge temps can vary from ambient to +60 deg C. There is a difference in the amount of timing advance that can be run at 20 v 60 deg C.
I can only assume that 'non temp compensating' ECU's are mapped safely for the higher temperature. If that is the case then at cooler temps they are giving away BHP due to non optimum advance.

Would you still be able to retain the closed loop idle on the JECS without the MAF input ? Or are you compiling an algorithm from MAP/TPS/Temp to input to the MAF tables ??

Andy

john banks

"Or are you compiling an algorithm from MAP/TPS/Temp to input to the MAF tables"

The was the thought. Another (easier) idea was to replace the Subaru MAF with a known robust MAF sensor which puts out a DC voltage and then adjust the MAF compensation map to suit. No electronics then required, just a change of sensor and a small area of the map.

Andy.F

Just re-read your initial post and that was what it said right enough - Doh

I think the 'better MAF' is going to be a MUCH easier option to integrate.

Although the JECS uses RPM already, you would also need to integrate it again in order to calculate flow from pressure.
There would need to be some sort of sq root function as flow is proportional to the sq root of differential pressure (read boost) which was then multiplied by the RPM (1 bar at 2000rpm is not the same flow as 1 bar at 2100rpm) taking account of volumetric efficiency variations throughout the range and finally temperature. It obviously can be done but........ probably best left to a MAP based ECU IMHO.

Interesting brain teaser though

Any MAF meters in mind ?

Pavlo

MAF remapping should be relatively easy to do. MAF scale has 64 16bit values on 96 ecu map, provided you can get the response curve should be a doddle.

Any paticular MAF sensor in mind?

Would also allow a larger bore maf sensor. 2 in parallel would work but only be accurate if both were flowing the same, as the voltage is highly non-linear, this could be overcome with voltage-freq-voltage or similar to linearize the response.

Moving to a MAP based system on a jecs is probably fundamentally flawed, in that the ecu code thinks in terms of MAF. Anything you make to provide a signal wont be measuring MAF as it depends on speed and VE, so I suspect that the response of the fuel control loops might go out the window too.

Rambling now, must stop.

Paul

[Edited by Pavlo - 6/25/2002 11:57:24 PM]

Mark Pitchford

How about using one of the older style hot-wire MAFs?

Then, do a compensation for the different voltages at the output.

Just an idea...

Edited for Aarron Spelling

[Edited by Mark Pitchford - 6/26/2002 10:05:56 AM]

john banks

Some of the older MAFs saturate at good airflow levels apparently. Any suggestions as to an alternative MAF sensor, as this seems the more sensible route.

I have access to the calibration table so that can be adjusted no problem.

[Edited by john banks - 6/26/2002 10:37:30 AM]

AJbaseBloke

There was a company that used Infinity Q45 snesors, and some sort of black box to translate that to "Subarese" for B4s (same sensor as all the ones that die).

I would love to tell you more about it, but the company that was doing them closed down before I could get a close look at their kit. AFAIK the Nissan part bolts straight up, but the signal does need "adjustment".

Hope that helps a bit...

MorayMackenzie

There are at least a couple of products already out on the market that translate from speed density input to mass-air metering outputs. I looked into such a system for my M3 in order to remove the fairly massive restriction of the mass airflow meter.

John: Drop me a line if you want me to have a look for the URLs i found.

Moray
bbs.22b.com

john banks

Thanks Moray. I am leaning towards looking at better MAF sensors that are maybe robust enough to work without the restrictive grids rather than going to speed-density.

Certainly there is a Link product to do the job, but it does not appear to use temperature.

Removing inlet path restrictions would be nice. I can see the appeal of removing the resonator and using a large bore intake pipe (BPM sounds as though it fits and doesn't suck closed), along with no mesh around a sturdier MAF sensor.

[Edited by john banks - 6/26/2002 2:52:08 PM]

carlos_hiraoka

Hey guys don't get mad for what I will be suggesting, but I think that the MAF sensor of the Mitsubishi Lancer Evolution (6 or 7) should work fine, this car comes with as much as 1.4 peak boost in standard form.

Carlos H.

john banks

I was thinking of that seeing a lot of HKS filters on them without issue. What is the size of the fixings? If I was David Wallis or Andy F it would not be an issue I would just make one out of ordinary household items

David_Wallis

Cheers John...

I like the idea of not even having a maf.. but I think im dreaming...

I run a HKS Super racing suction kit on my 94 with no probs... possibly because its an early one though?

David

carlos_hiraoka

John, since Subaru and Nissan also share a lot of parts (like the rear R180 diff for example) it could be possible to use a Skyline GTR MAF sensor and maybe it could fit with less hassle.

Carlos H.

David_Wallis

Price???

David

carlos_hiraoka

MAF sensor for EVO 5 or 6 ,from the Coordsport website ( www.coordsport.com ):

" ..... Sensor, air cleaner air flow E5 6 £269.50 ....."

Carlos H.



[Edited by carlos_hiraoka - 6/26/2002 7:10:24 PM]

Pavlo

John,

With any replacement MAF you use, you should really retain the mesh, as it's there to help prevent turbulence in the air flowing throught maf. Too much turbulence upsets the reading, as is the case of putting an induction kit on or similar, its the change in flow charactristic that may make the engine run lean/rich.

On the turbulence front there is an argument to putting in a small pipe between filter and maf body to help create staighter airflow through it.

However, contrary to what I wrote before, if a box of tricks does create a MAF signal, I can't see why the ECU wouldn't work with it provided is was representing MAF in the end.

paul

WREXY

Pavlo,

Have you seen the APS cold air induction kit? If so do you think that there is too much turbulence with the APS cold air induction kit, making the MAF read lower? My lamda readings are 0.86 Volts with Delta Dash and the MAF has been replaced. I originally thought I had a faulty MAF and after replacing it, got the same readings. So with the help of others on here, we are suspecting the induction kit. I will remove it to test sometime in July with the help of a friend, when he comes over, but untill then and since reading your post above I thought I'd ask these questions. Would this extra pipe you mention work with the APS? What exactly is it and where is it available from and how does it fit?

Cheers,

Wrexy.

Pavlo

Wrexy,

I have no idea! The MAFs tend to calibrated with the factory airbox in place, so the resonator will resonate at some speeds and influence the results for starters. My main point above relates to fitting things like inlet trumpets right on the MAF, as you may get turbulence, or some other flow phenomenon that will give greater/less flow in a local region around the sensor wires in the MAF sensor.

So putting in something like an APS filter may give lean readings due to change in resonances*, or change in flow.

Having said that, if there is a way to put something like a 2-3 inch long, smooth straight pipe between the filter and the MAF, it may be worth a try.

I think the easiest way for users is for Ecutek (or equivalent) equiped tuners/mappers to begin to scientifically compensate for filter changes my recalibrating the MAF. I would imaging Power Engineering would be well placed to do this as they have a good flowbench. If they could characterise common filter/MAF combos, it should mean easier mapping for them in the future.

Paul

*The resonanator work in a simialr way as throttle body inlet trumpets, the length of which infulences flow at different speeds, generally, shorter is better for higher speeds, although the volume of the box will also add another variable.

Pavlo

I don't know if an extra pipe would work with the aps, but it's worth trying. What you are trying to do is get the filter a little further away from the body of the MAF sensor. Without looking at I can't say exactly how you would fit such a pipe.

I suspect it would need to be a pipe about 50mm long, same internal diameter as the MAF sensor, and with a flange on each end, one to attach to the MAF sensor flange, and one to attache to the filter bell.

BUT

if it's this filter
http://www.airpowersystems.com.au/wrx/airfilter.htm then it may be due to having removed turbulence that was present with the factory induction. As this kit looks as smooth and straight as it can be pre MAF sensor.

If the above is true, you are probably looking at something like a remap/maf recalibration/electronically alter MAF signal.

Paul

WREXY

Paul,

Thanks very much for replying. I think your info sums it up. It is the APS kit making the car run leaner. Great info! It helped me understand the whole problem better. Ah, so less turbulence with the APS.

I think your advice in the last paragraph is the way to go cos the induction kit in the link you provided is the one I have and it is smooth and straight before the MAF, so less turbulence.

Cheers,

Wrexy.

Andy.F

"If I was David Wallis or Andy F it would not be an issue I would just make one out of ordinary household items "

Hehe, One seems to have forgotten ones "Boost controller for 50p" humble beginnings

john banks

You overvalue me sir! The cheapest one was 6p. I didn't dare do the free one which was unplug the feed to the MAP sensor (and it would not be very mysterious and I would have had a hard time working out what to do with the loose hose and would have had to ask you or David to weld up a hook for it ). Electronics bits are a lot easier than welding fings together

[Edited by john banks - 6/27/2002 3:11:49 PM]

David_Wallis

weldings easy... just like soldering

now tig welding.... wheres that book I ordered from amazon.

David

Andy.F

"weldings easy... just like soldering "

Note to self - Don't travel in a car that David welded the floor on

Pavlo

Tig welding is quite easy if you're used to gas welding, just replace gas flame tip with spark (there is more I know, but this sums it up).

Practice makes perfect.

Paul

David_Wallis

Andy I meant weldings easy.... and soldering is easy...

Mig welding is easy as ****.. setting the welder up is hard... and making a strong neat weld is another thing...

Last time I tried gas welding I was ****... I was about 12 though..

10 years and Im probably no better...

David

john banks

Andy, I can just manage to assemble these headers on my floor, but on the car may be a different matter - fancy a fiddle again when the VF23 arrives?

[Edited by john banks - 6/27/2002 5:34:15 PM]