Cheap MAF option?

[EddySpaghetti]

Hi,

Have been thinking about MAF kits and all seem very expensive.

Physically, are the MAF kits sold by the specialists basically the same as the type of units that are now mass market available for any fuel injected car? If they are then it would seem worthwhile buying a 2nd hand MAF of approximately the right dimensions and then working out what signal conditioning you would need to create to give the same result to the ECU as the barn door does currently.

I'm sure its a lot more complicated that I've made out to get the signal conditioning bit correct, but in principle someone could pick a widely available (now - as in from scrappies / ebay etc.) MAF, figure out the signal conditinoing once for that specific make/model of MAF and then create a much cheaper kit for the likes of us. I would have thought that once you had the conditioning figured out then the unit cost for people to buy the right components/circuit board and MAF would be pretty low - certainly a lot less than the hundreds/thousands that the kit costs from the specialists?

The rig could be pretty simple - basically join the barn door and the MAF together in series with the right size of hose, blow air through and condition the output of the MAF to match the output of the barndoor sensor. - could do it on the car too to create the right range of 'suck'

Just a thought - keen to know what you guys might think to such an idea,

Eddy

 

[Peter Empson]

The Scivisin MAF I have is a Bosch part commonly found on BMW's (which is good as they're quite cheap and common). As you suspect the tricky bit is generating a suitable signal. This has been done on several threads on Rennlist, so would suggest starting there as it could save a lot of time.

 

[944 man]

I suspect that the real problem is that the ECU disregards the AFM after a certain (relatively low) RPM, meaning that any simple conversion benefit will be limited.

 

[zcacogp]

Eddy,

I've considered something similar on occasions, but while it sounds simple in theory I don't know enough to understand how difficult it would be in practice. Linking the two together in series with the same hose would be a good way of measuring it (although for completeness you'd want to repeat the test with them connected both ways 'round - one before the other and then one after the other).

However, that is quite a lot of hassle for producing one set-up. You'd want to do it well, and then sell the chips to allow others to do the same.

Just remind me, what are the advantages of running MAF over and AFM?

ORIGINAL: 944 man
I suspect that the real problem is that the ECU disregards the AFM after a certain (relatively low) RPM, meaning that any simple conversion benefit will be limited.

Simon, sorry to be thick but what does this mean? []


Oli.

 

[barks944]

Eddy, if you have a Lux put a KH chip it and you won't want a MAF any more . On a side note I have made a home made MAF converter for about £40. I have never got round to mapping it properly but I have had the car running and revving up. Its not that complex and the main component is a USB programmable dev board called arduino. Coupled with an analogue output chip and a few resistors/capacitors and you are away. All I needed to do to finish it was to match up the MAF signal to the AFM signal.

 

[sawood12]

ORIGINAL: zcacogp

Eddy,

I've considered something similar on occasions, but while it sounds simple in theory I don't know enough to understand how difficult it would be in practice. Linking the two together in series with the same hose would be a good way of measuring it (although for completeness you'd want to repeat the test with them connected both ways 'round - one before the other and then one after the other).

However, that is quite a lot of hassle for producing one set-up. You'd want to do it well, and then sell the chips to allow others to do the same.

Just remind me, what are the advantages of running MAF over and AFM?

ORIGINAL: 944 man
I suspect that the real problem is that the ECU disregards the AFM after a certain (relatively low) RPM, meaning that any simple conversion benefit will be limited.

Simon, sorry to be thick but what does this mean? []


Oli.

The benefit of a proper MAF solution are two fold. Firstly a MAF presents less of a restriction in the intake of the engine, therefore allowing it to breath better. Also by not having any barn door or other mechanism in the system there is less turbulence induced in the intake therefore less losses etc.

Secondly a MAF will fuel much more accurately across the whole rev range - it is a much much more precise measurement device. The AFM will typically only be of use upto about 4krpm under acceleration, after than it is wide open and therefore surplus to requirements. So you can think of an AFM as largely being an open circuit control system and a MAF is a real time closed circuit control system. Also because a MAF measures temperature in real time (the AFM doesn't) it allows the ECU to carry out real time mass airflow calculations which enables much more accurate fuelling. Also the sensitivity of the MAF is much much greater than the AFM - small changes in airflow might not be measured by the AFM as the barn door might stick and not react to small changes in airflow - the MAF will notice tiny changes in airflow.

I'm sure these benefits are only the tip of the iceburg. The stock ECU is not capable of working with a MAF - so needs a piggy back computer to provide the extra umph required - the piggy back effectively short circuits the core of the ECU and directly commands the injectors. Any MAF solution that doesn't have this piggy back computer will only be using the MAF as a gloryfied AFM - you'll be getting the mechanical benefits of a MAF (i.e. free'er flowing intake) but you wont be getting any of the engine management benefits of a true MAF solution.

A MAF doesn't necessarily give you any more power on its own - it will however unlock more power out of other mods and allow much more precise fuel mapping and therefore safer mapping meaning you don't have to build in as much conservatism into the fuelling maps.

You can cobble together a cheap set up that will get a MAF working with your car, but you wont necessarily have true MAF.

It is true that a decent MAF kit is knocking on the door of a standalone ECU, but the initial purchase of a standalone ECU is only the tip of the iceberg of the total costs of going standalone and getting it working well, reliably and safely so in the long run they are much cheaper and will get you 99% of what a standalone will get you (negating any of the extra gizmo's a standalone can give you).

As Tom has said, you're far better off with a good AFM setup than a cheap MAF setup.

 

[barks944]

If you are good enough at coding you could do what Vitesse did and re-code the AFM section of the DME code. A guy called rogue_ant is doing it over on rennlist. The 944's DME uses a transfer function to relate the AFM voltage to airflow. The transfer function is such that it cant describe a MAF curve properly as its too curvey . You need to re-write the code so that you can accurately map the MAF signal. I would do it using a map and interpolation personally, I think this is the approach the guy from rennlist is taking, I would be surprised if this isnt what Vitesse has done too.