Has anyone ever tried to take a MAF sensor and cut the plastic away, placing the existing electronics into another (Larger) tube, and then re-assemble? Then, re-program to Accommodate for the extra Volume that you have added.

It should work, and I am thinking of trying it. If someone can tell me that it does not work no-way/no-how then I will re-think it. BUt I cannot forsee any problems, other than the car will run lean until you can reprogram... How lean it would runs willd depend on how large of a tube I have gone to.

It may be possible to roughly calculate the extra volume and be close enough that I should not throw a code...

For example the Dia of the OEM MAF is 2.5" (a guess here) and I go to a 3" tube. I can figure that the area of the original MAF was 4.908 sq inches, and the area of a 3" tube would be 7.068 sq inches. THerefore if I recalculated all of my MAF tables to 69% then it should be close enough to get off the ground and program from there...

Any thoughts or am I going frankenstein on it?


John

 

The sensing elements are somewhat dependent on the air velocity across them. You just can't take a sensor that's designed for reading something as small as 7 gm/sec, and then expect to be able to read 5 gm/sec (just using your math here) accurately. Not to mention all those pesky terms like laminar, and boundry flow, which may pose problems. IMO, if you really want to stick with a MAF, then use one of the later ones, that has the cross sectional area, that you think you need, and figure out how to use that. FWIW, while GM did get the early ones to work, OK, they sure revised the code in the later stuff to get them to be *close*.

You might just mock up a little test rig. ie a scope, and gather up some MAFs, and see which one will report low air flows, best. You can improvise, using *channeling* tract, and then just blow thur a 1/4" hose, or something to restrict you lung capacity, and see what kind of results you get. Kind of a homemade Spirometer.... <G>

WOT, isn't that hard to get right, it's making a car that's fun to drive, that can take some time, and talent. If you're running a loose converter, and only going a 1/4m atta time, that makes life much easier.....

One other way to do things would be using one of the designs that used a *sampling tube*.

 

I thought Granatelli made a thirdgen MAF sensor, but I cant find it on their website or at Jegs anymore. What gives, did they discontinue it?

 

Just how fast is your car that you need to do this? A stock MAF with some attention to the PE tables at high flows will be fine down into the 12s. Making your 8 bit MAF accomodate high flows is going to make the resolution worthless.

 

Not into the 12's thats for sure...

Iguess I was hoping for another .1 in the quarter by enlarging the MAF... It was worth a shot.

Thanks

John

 

http://www.jegs.com/cgi-bin/ncommerc...73&prmenbr=361

there it is.

i have one, not sure how to use it in turn with tuning my chip.
on the stock setting now, which flows better than stock w/out having to remove screens.

 

Do it man, modify that bugger! Post some pics too.

I've got one of the local shops here that is setting up a MAF calibration bench for the LS1 stuff, and I could probably put some of the L98 stuff on there to see how it does in modified trim.

 

www.prosettech.com sells a tpi to ls1 ls6 lt1 maf translator. you have to email them however. i helped do the development on it and its very very adjustable.

 

How's development going on the original source code project?.

Why wouldn't you just edit the appropriate tables?.
It just seems like correcting the calibration would be the simpliest way of doing that. And doing that gives you complete control of adjusting it's calibration, at every step, for the best possible resolution.

 

Youd probably have to do both. Adjust to get the desired output range, and then recalibrate on the ecm side.

 

The range is a given. I don't see why there'd need to be an adjustment, for range, if the tables could be edited. Unless one's just going for the *close enough* strategy. I've just never had a car run as well, when just settling for close enough, correct has always been more then worthwhile, least for me.

 

had to park it for a while. got a living to earn. get back to it closer to winter.

ahh the translator that i helped develop has 10 PC calibration points for its transfer funtion. so you can have a ton of low flow resolution and then ramp it up later in the flow curve to WOT fueling where the error % doesnt work against you so much.It was a feature that i made sure when the device was being developed was input into the device.


device feature i dont know why there not on the website
rs232 serial communications
flash calibrating
10 point transfer function both the FQ in and the voltage out points are adjustable.along the scale so they can be moved
Fq in V out
FQ in and FQ out
V in and FQ out
V in and V out.

and by the way the tpi computers (1k pull up instead of 10k) require alot of drive current on the MAF end just check the schematic.

 

Here's another one. Windows interface looks kinda familiar...

http://www.maftpro.com/