I've been researching prom burning for a couple months now,
I recently purchaced a laptop and plan on puchacing the Moates autoprom shortly


but my question is,

at a local forum I belong to the idea of running dual Mafs was brought up by somebody doing a TPI swap into his pickup,

I've done a search and didn't really find anything

would it be possible to run the MAF's together in paralel,
and if I understand MAF scaler right, cut them in half, and then tweak the maf tables to smooth everything out?

what do you guys think?
am I way in left feild on how it works?
is it plauseable to run dual mafs and get a driveable tune out of it?

 

or would you have to double the tables cause you'd be running super lean,

twice the airflow at a given voltage,

which would mean you would have to double the scalers which wouldn't work in the higher tables

 

Simplicity is the key to design.......
If you're easily pegging the MAF, for the work you're considering you could repin it, and just go to the MAP ecm.

Could you?, well maybe. But be prepared for ALOT of tuning, and lots of time with a WB to get it right.

If you want to slightly increase the MAFs range, divide it by say 30%, and drop your injecotr constant down about the same amount.....

But, like I said, you're gonna be dancing around for a while to get the dual MAF working, correctly.

 

thanks for the input

 

One other idea, would be to use one of the big Ford MAFs, and a MAP GM ecm. The Ford MAF is a straight voltage output, and input that to what would normally be the MAP input. It probably would take all sorts of fudging in the tune to get it, but if you're really wanting a MAF, that's make as much sense as anything.

 

I don't get it, why not use a MAF ECM? AFAIK 165's in TPI cars used the analog MAF, right? And you're meaning the Ford MAF is analog, by the looks of what you wrote. I actually was thinking the Ford MAF was a digital output (frequency vs flow), but I haven't researched it at all. A frequency to voltage converter could cure that (among other options). Either way, it'll take a lot of testing to get the MAF tables back in line, and you'll still have to lie to a 165 to get past the 255 g/s limit - using smaller numbers (maf Flow vs maf Volts) and a smaller injector constant than actual as you mentioned. Just as long as your MAF is a 0-5v, and the voltage response is such that the MAF is flowing quite a lot more than stock when it approaches 5V.

 

yeah sounds like a total nightmare IMO not just on the tuning end, but the blue prints of how to get it to flow the best without tunblance and what not. Sounds like a total hack job, good idea but theres an easy way round a not so big problem.

 

IMO, the 165, and early codes just aren't up to the task. If you're just interested in WOT, that's one thing, but if you want something that's fun to drive all the time, the 730 will get you there.

The 3" Ford MAF has alot of flow potential, I forget the exact numbers on it, but it's healthy. I guess if he really wanted to get nuts, then going to a ZO6 MAF would be an option, but then things get really sticky unless you use the whole OBII ecm.....

************

The Ford MAF use a sampling tube, so they're not as fussy about turbulence.

 

I don't see how that would work. If the MAF is flowing max airflow such that the amount of air causes ~5v at the output, and then the V_out is reduced to 3.5 Volts then you really gained nothing.
Yeah, no true 5.0 volts but it goes non-linear near the extremes and the slope of V_out is much less. Much like a soft limiter (e.g. diode).

I think I must have missed something here. By reducing by 30%, did you mean a voltage divider?

 

The major problem isnt as much the MAF, but the code. The resolution capabilities of the stock code is limited, and wouldnt even take full advantage of the A/D converters capabilities. In order to properly do it, youd have to redo the code to use better 16 bit tables and other routines to store the larger terms without adding additional error.

 

I was talking about the MAF table scalars, and injector constant. While there is a lose of resolution at the extreme top end, you can fudge that with the TPS or RPM enrichments.

It's been done.
Again, it's just a way to make it work, and by no means the best way, IMO.

 

Yep, gotta splain that to Mark Jackson, and his TR that ran mid 9s with the stock MAF.
He's since stepped up to a MAP, and is running low 9's at last report.

With a 255 limit that gives, about one step per K/Pa with a 3 bar MAP, or 2+ counts Gm/sec for a MAF. At the higher airflows it all gets *close*. That's why you need to add a few corrections to compensate for the *lose* (for lack of a better term), for either system.

Not to mention that once you get to that extreme of fueling, the electronics of the ignition get to be a headache.

BTW, the new Hot Rod has an article about the new AL Buick V6 block and a KD engine. The details are interesting, ie MOTEC, and GenIII LS CNP.

 

I see what you are saying now.
Overall, using a MAF that will max. out at a certain grams/sec is like using a MAP that will max. out at 100kpa in a boosted app.
Once you max out the unit then it is strictly Alpha-N and only the fuel/rpm tables will keep it right (but the MAF or MAP algorithm doesn't know it). The measuring device is at its extreme and can't provide useful inputs to the algorithm.

I don't see how 16 bit would help at all. The MAF is set to heat the wire to temp. X and at some point produces a 5v value with Y grams/sec. If you exceed that value then it still produces 5v. Just like a 100kpa MAP in boosted apps.....the algorithm is blind to more CFM or KPA. At that point, it is all about the fuel table.

By running dual MAFs you are creating an air leak (theoretically cutting the CFM that the MAF sees in half). In doing this, you increse the dynamic range of the MAF. In that case, going 16 bit would be benefical. Is that what you guys are saying?

 

I wonder how well a dual MAF system would work if you just use ONE MAF, and the other is a "dummy" MAF that just attemps to balance the flow from one side to the other. All that you should need to do in the calibration is halve the MAF Scalars, realign the MAF tables, redo the spark vs LV8, and maybe double the AE vs LV8. Probably a few other things, but basically treat it as half the engine. Disclaimer: I may be saying something opposite of what you'd want to do, I haven't thought about it much.

I'd guess that wouldn't be such a bad setup, as long as the inlet ducts are pretty much equal resistance.

 

Yes,
yes (adders),

Not really. The MAF doesn't read gm/sec. it's though the ecm's calculations, that the output is converted to gm/sec.. If you want to say, x output is equal to something else, and change the injector PW to match a different gm/sec, you can do that. You just gotta watch all the other air flow based variables. ie the LV8 goes to ****.

It still winds up easier going to MAP, and not having to fudge everything to get the car to run.

 

That is what I was describing. Theoretically cutting the airflow in half will give a 2x in dynamic range of the MAF as seen by the ECM. That is, a MAF that is limited to X CFM will now be able to do 2X CFM according to what the ECM sees. That is all theoretical, you would need to figure out the balance or fudge factor between the MAF and the pipe/MAF that is in parrallel.
If you went with two MAFs then you could read each MAF and add the A/D counts in the code to get the total CFM. 16 bit would help in that case.

This is all acedemic though. Like Grumpy said, it is more simple to go MAP and have a better more reliable system (less parts, not the whole MAFs are unreliable thing). Both MAF systems and MAF systems are good. They just need to be used in the proper application. I think that if you need two MAFs then you have exceeded the useful limit of the MAF system and should go to a MAP system (or getting deep into assembly coding and flowbenching)
A used 730 ECM and brand new MAP is less than $100. JYard MAPs are a dime a dozen. 730 ECMs are too. A used setup would probably $75 or less.

 

Yes, by running anything that registers over 255 gms/sec, or any high flow application, changing over to 16 bit values will help because the error introduced by an 8 bit term thats used to store the flow calibration (IOW, the MAF tables) becomes a significant portion of the overall error. You want all the calculations and tables in the ecm to have an overall error thats much less then the error from the A/D. This way the reported airflow error is as small as possible.

When I did this with my ECM, I had to be careful how I executed the fueluing calculations so I didnt create additional error. Basically I wanted the term to start large and get smaller as the calculations progressed. Finally, the injector pulsewidth term I arrived at was usually smaller then any of the previous terms in the calculations, preserving resolution.

The later TBI/CPI PCMs are capable of reading in the actual MAF frequency, giving just a few hundreths of a percent error verses 4 or 5% error with the A/D, but theyre so damn complicated that theres not a snowballs chance in hell that Ill figure out how to get it to work any time soon

 

LS1 truck MAF. Got one for $25 dollars. Reads to some rediculous airflow value, like 450 gms/sec. **** poor choice for an 8 bit system, though. The signal vs. flowrate is much lower then the old MAFs. With signal scaling or some other fancy hardware it would probably work ok.

 

ok a couple of things here are Incorrect. MAF on 730 ??? ahh you'll wind up with a pretty screwy looking Ve and spark table. with a MAF your only going to use the values crossing most directly across the table along a diagnoal axis. Stick with a 165 there is no advatage to attempting this.

the for 3" MAF has serious Flow potential problems. First is the fact that in full non ported housing trim it can only flow around 560cfm of air before it starts choking.Secondly the input impedence is around 10 mega ohms on the MAF input on the ford ecm and the input impedence on the 165 is around 1k ohms. Even with the ford maf at its choke point you end up with a MAF that can only produce 2.75 volts if you've got a good one. So youd lose half of your resolution but youd gain a ton of overhead.

The ford Sample tube is a massive turbulence inducing night mare. I should know i worked on the problem alot at Pro-M racing when i use to work there on a tier 1 program. the answer was to stop trying to port/recalibrate the sensor and just build a different sensor housing. The new Z06 maf for the 06 Z06 was having trouble with sample tubes. Ducting creates serious issues.

i wouldn't use a z06 maf on a z06. its got airflow turbulence issues. the hot ticket is the 2002 and up 5.3 and 6.0 lsx maf they have a screen/Honeycomb/laminar flow straightener. I have swapped out the z06 maf on a zo6 and picked up 5-6 hp contiually becuase the fueling is so much more stable it also solves idle suring issues with large camshafts.

So these are the facts as observed in the field and tested rigorusly.

 

Just one thing...
The ZO6 MAF (at the least the one I have) doesn't even use a sampling tube.

 

I have done the ford maf convwet. It is easy beside find the resistor in the ecm and removing it. I can get a copy of the transfer tables. I run the 90mm maf.