Is it possible to run 500n/a with a 355 and still run a MAF setup? Please i need help with this on what to do...motor gets out of machine shop in less than 1 week

 

Is it possible? Probably. But, the ecm will only register up to 255 grams/sec of airflow since its 8 bit so youll be flying blind most of the time. Any good motor will easily overwhelm the MAF. A speed density setup would be better if you want to do things right.

 

Sure, there are Buick GNs running in the 9's with a basically stock ecm, and MAF. Just be prepared to learn alot, and possibly breaking a few parts doing it.

A MAP system in a bit more intuitive, and easier to learn. Plus has more resolution so that you can better fine tune it.

You have a ton of reading to do, to begin to grasp what you're going to need to do in the next week.

 

is it possible to convert to 16-bit for MAF? or is this a dumb question? What are my alternatives? Is the flaw in the MAF too or just the computer? do they make a programmable ecu to work with a MAF setup? Thanks for any help


:lala:

 

What I'd like to know is if somebody has taken a stock MAF and spaced the wires out further. This would increase your spread but decrease your resolution (keeping 8-bit). The tricky part is getting the spread which would require very fine adjustments and then measuring.
If you're going to get a MAF flow tested you might as well give it a shot on an older maf you've got hanging around.
Just my

 

Dont know much about maf. One thing I wondered, though, is if maybe a larger digital FM MAF couldnt be used. Just speculation on my part but maybe by using one of the 16 bit counters and the f. div. routine it would allow a 16 bit MAF reading to be taken in. I dont know if it would help or not but maybe some of the loss of resolution could be avoided.

MAF is cool. Although its not much use if only an 8 bit reading can be used. Sucky resolution just kills things. With speed density, it can be used on anything since the fueling is based off of the VE + flowrate of the injectors using ideal gas laws, but those same things are also its weak point. I really hate having problems with speed density since things that effect the map can cause a feedback loop with the fueling causing more fuel to be steadily added as teh map rises from the AFRs dropping like a brick. Ever have an egr thats stuck open with speed density system?

Oh well, there are pros and cons to everything.

 

You're over my head on that first part but as for the 8 bit readings, it isn't all that bad for a MAF. Having 255 different load settings for an engine is rather good, anymore and it's just going to do nothing unless the code is interpolating. Although when you flow bench a maf you usually don't do all 255 loads so the MAF scaler table is what needs resolution more than anything else. That's just my opinion which could definatly be wrong .
As for the speed density, the good ecm's use ideal gas law, our TBI computers don't . They use a really fumble f*** way of getting this close enough.

 

You **can ** peg the MAF and still use the WOT % PE AFR correction, to still get you close. It's just a matter of how close it is you'll settle for. Obviously, close is close enough to run in the 9s. But, it takes alot of work, and blown headgaskets to figure out which way to close is close enough.

If you want to build any serious HP then you want to look for the code that will allow for getting as close as possible with winding up with a Headgasket/roc bearing collection. With a DBC, you'll beat the corners of the pistons, and wipe out the rod bearings from detaontion, before popping a Headgasket thou.

The 58 code allows for good timing and fuel based on MATs, which will save anyone alot of time and broken parts since that's an entire element of the tune that you don't have to worry about, and is more or less already pretty close.

Can you make 500 HP with a MAF, yepper, but doing it with a MAP means less foolin around, and using a system that's ALOT more intuitive.

 

Doubling the flow taht can be registered and halving the resolution would be a solution, but to me it seems like the loss of resolution would be noticable if one really cared about low load performance when just driving around. Who knows, it might work.

As for the tbi ecms, just about everything in them seems like a half-assed approximation

 

But you're still totally ignoring the spark being solely related to LV8.

Toss in reversion, and then it all falls to what's close enough.

Ifinyawant to use the MAF for accurate airflow say at WOT, and then set up some tables for properly doing that, that's one thing, but for all around idle to WOT, and high HP, you need alot more tables, and compensatations, to do a MAF only correctly.

 

I was talking about using the MAF only for fuel. MAP would be used for the spark . Who said stock code? The question was about MAF and supporting 500hp which I think it could do with some good code.
I really don't think the 8 bit maf is an issue even for 500+hp. The load points are interpolated imbetween with straight line interpolation.... not good if you only have a few points between WOT and cruising. That's why I think the code just needs to use the maf better, not a better maf with the same code .

 

Id be open to trying something like that if there was a digital (or maybe analog?) MAF that could actually read above 255 grams a second. Im sure something like that has to exist.

 

LT1 & LS1 MAFs can. . .

RBob.

 

How much can tehy read? I heard like 500 grams/sec or something.

Seems like operating it and getting a reading from the sensor would be a lot more work then hooking up some power and stabbing the signal wire into the MAF/VATS frequency input on my ecm.

Donno, might be interesting to bum a function generator off of someone and simulate it on the bench.

 

On a sidenote I think we totally hijacked this thread Oh well, its all in the interest of learning

 

Yes, a 3rd gen MAF car can make 500HP (that is engine HP).

As Grumpy pointed out, you will max out the MAF registered flow at 255 gm/sec, but you have the PE vs RPM table to get you the extra fuel that you need at flows above 255 gms/sec, and, furthermore to maintain 12.8:1 AFR. You will need a wide band O2 sensor to tune this in right.

Case in point: my '99 LS1 Camaro SS makes 400RWHP using 310gms/sec. If you allow me to use a simple ratio, 255 gms/sec will make 329RWHP and at 511gms/sec (the max of the LS1 MAF) you could make 659RWHP (in an LS1 car).

A screenless 3rd gen MAF should flow about 700CFM, this should be enough to make 500HP at the flywheel given that everything else is going good, matched well, and correct. It takes about 1.3CFM to make 1HP in our 3rd gen cars. Therefore, 700 CFM should make about 538HP tops.

I have seen published flow values for 3rd gen MAFs (screenless) from 658 to 719 CFM,,, so supposively, the range of engine HP is from 506HP to 553HP.

BTW: 658CFM is equates to 372 gms/sec and 719CFM equates to 406 gms/sec,,, if my conversion is correct. So we are definitely about the 255 max reading of our MAFs.

 

Just dont' tell certain Vette owners...

 

99?, well you're in another league with that, since it'll read to about 500 Gm/Sec.. And the code is designed to work at those RPM, and HP levels, where as the pre 89 stuff was at like not even 1/2 of what a LS1 makes.

I'm not for sure on how the spark is handled on the OBDII stuff, but I'll wager the timing tables are ALOT better in concept and execution then the 3rd Gen code even thought about.

Once you start gutting the MAF, then you just hoping that you don't have troubles, with the Gm/Sec metered vs actual are even close. While a Flow Bench is an *indicator*, remember, the intake air tract isn't very stable, flow wise, there are all sorts of possible errors, and while GM has taken to running screenless, there's no one I know on this board with guite the resources they have.

 

Grumpy, LS1Edit will allow a max value of 511 gms/sec in the MAF table, which as you know is frequency vs flow. I installed a so called calibrated GMS MAF, but it was not close to being "calibrated", in fact, it was off a good 10 to 20%. I used the LTFT (Long Term Fuel Trim) values over the 16 part throotle and the one WOT cell to calibrate the MAF correctly. I corrected values in the MAF table until the LTFTs were all slightly under 0% (3rd gen BLM of 128).

Yep, my '99 Camaro SS is a pretty nice car and very fast. With initial simple mods like K&N air filter, a descreened MAF and MAF translator (this was before I acquired the LS1Edit tuner) the car pulled 319RWHP. Not bad for a nearly stock car.

 

Im at least somewhat interested now. Who makes the LS1 maf? Delphi? Bosch? Is there any tech data on it? Probably more apropriate to ask the 4th gen guys...

If I could find some data on it maybe I could simulate it on the bench to see how it might behave in the car, or at least get a feel for how the calcs. might be done.

 

I can only assume that it is made by Bosch. However, it is frequency based so some kind of a "translator" would be necessary to put the output into 3rd gen terms.

Besides, if you could get the LS1 MAF to work on a 3rd Gen car, why not move to the 85mm aftermarket MAFs from the get go. Then there would be no arguments about sufficient flow. The stock MAFs are 75mm diameter.

Edit: we have hi-jacked the thread again. This has been discussed before. We should start a new thread and do a search.

 

Its desirable for me to use one of the unused freq. counters in my ecm since I have no free inputs if Im still going to use my map as well. Not to mention that the A/D converters are 8 bit.

Your right... I guess Ill search instead.

 

LT1 MAF's have a fequency range of 2-12kHz, three wire. Made by delphi.

LS1 MAF's also have similar frequency range, 5 wire, Made by delphi, the two extra wires are a built in MAT sensor.

 

damn? Up to 12 kHz? Guess the idea of using an LS1 MAF and a freq. input on the ecm goes clear out the window!

 

Dunno, the ECM counters run at 64KHz. . .

RBob.

 

the BMW bosch mafs will work in the 3rd gen camaros with the 1227165 ecm. just have to totally rescale. again big issue is resolution on the bottom end. adding more tables would help. problem is the granularity of doing this.

 

is anyone here able to do the conversion and programming with the BMW maf?

 

If your going to go crazy then why not just do it right?.

Map for timing, and transistional work, and then MAF for WOT/PE, like the OBDII stuff.

BTW, have you noticed the 749 has the input for a low frequency MAF?. Or change one cap, and it becomes a high frequency MAF input. Oh wait, then there's public domained 58 source code.

Nope, never thought of this before...................

 

So what would stop someone from using a frequency to voltage convertor like this one:

http://www.microchip.com/stellent/id...cName=en010486

and using an LS1 MAF with a redone scalar table?

I haven't looked at the pinouts for any of the MAFs, but I thought I'd toss the idea out there.

 

The problem I see is that there will be just 65 counts registered for the lowest frequency reading for the maf and then a mere 5 counts for the highest reading. There will only be 60 counts difference for the entire range of readings. It seems like to me the ability to accuratly resolve the MAF readings would be quite difficult. Throw in the +/- 1 bit precision and the errors can be quite significant at high readings.

I was sort of hoping it would be like 50 Hz to 500 Hz or something. There one would have a difference of ~1200 counts for the lowest and highest readings, assuming the frequency output is realatively smooth and accuratly reflects the maf readings. With the LS1 maf the readings would seem be quite coarse. I donno, what do you think?

AlexHJ,

The problem I see with using a frequency voltage converter or a larger maf is that youd be trading resolution for higher readings. I think Id rather stick with MAP.

 

I didnt know that changing a capacitor can change the sampling frequency. Is it like that for other ecms as well?

 

its probably not a sampling frequency change as much as it is an input filter change, so that the higher frequencies make it to the A/D.

there is an article regarding MAF systems, both voltage and frequency at diy-efi.org, called something like 'Air meter design.....'

shows some good implementation strategies, using a pulse accumulator for better resolution that just strictly using a counter. even with 15usec intervals, you probably could get good accuracy, but code would be required. for a good assembly guy, its free. for anyone else, its alot of work.

probably better off at that point to switch to MAP, or LT1 code using DIS(eDist) if maintaining MAF is that important.

of course, you could just throw fuel at it and monitor the WB, give up a bit of HP due to restriction, and live with it, too.

 

does anyone know a good tuner in the dfw area (dallas - ft worth, Tx ) that could maybe help out a little. Motor goes in in a week and i need some help quick. Thanks

 

It changes the R/C filter, is all I know.
The only schematics I know of with that much detail were at Ludis's site.

 

All the help youll need is here if youll be using a gm ecm. I would seriously consider switching over to a speed sensity system. In the end I think you would be much happier with the results. The MAF is limited and youll have to fudge things a bit to get it to work properly. We can sit here all day and talk about it but I really dont think a good all-around solution could be found for using a MAF with an 8 bit ecm.

 

Oh, so its just the filter then. Im not a hardware guy so...

It seems that the small number of counts per frequency input would still kill the idea. There would be a large difference in frequency for each change in the number of counts registered so the resolution might suck.

I guess I could handle the assembly part of it but if I cant get fairly good resolution it would be a bit of a waste of time.


EDIT: Can the counters also work the other way around? Like instead of free counting between pulses only count the pulses themselves? Guess I should be clear on that before I make myself look like too much of an idiot.

 

Could you please tell me why you think an 8 bit ecm and MAF (8-bit) is a bad thing. I counted 255 load points. That's a very smooth curve. Don't forget that things are interpolated. My next post will contain an image to better explain my point.
I just think 16-bit is only good for emissions, not performance.

 

I dont think its a bad thing, just that if you double the range youll loose resolution. If your only using the MAF for PE like gumpy sugested then it probably wouldnt be a problem. I know that theyre interpolated but I guess it depends on what the transfer function looks like since youll only have a line between any two points.

 

Grumpy will know this one well, on my GN the MAF value is limited to 255 g/s but its a 16 bit value. The MSB is 0-255 with the LSB being the finer resolution. Imagine at idle the airflow is 4-5g/s the extra resolution is used to precisely calculate the fuel PW, otherwise the fuel would jump up and down 20% at a time.

Other lookup tables use only the MSB, 8 bits is enough. So as 8 bit the 255=$FF, in 16 bit its $FFFF. Its still limited to 255 g/s which is a pain but I've 16 bits up until then.

GNs easily max the MAF and the pulse accumulator wraps around and the ECM arrives at a reading like 1,2,3.... and the fuel nearly shuts off as a result causing the car to stumble until the number unwraps again. One guy has crafted a software fix for it and can get to 512 g/s.

 

Thats sort of what I had in mind. Ive thought of this previously before this thread came up. The idea of using the maf to calculate the pulsewidth and using teh MAP for the transitional stuff and the spark sort of appeals to me.

The only thing is how to get a good reading in if your using teh maf entirely for fueling? Is it possible to have the counter measure frequency instead of period using the VATS input? Like 1000 Hz would equal 1000 counts a second, or something similar or is that what it already does?

I think this would enable anyone to use a good maf with their system but I have no clue when it comes to hardware.

 

That's what GM been doing in some appls since 96.