MAF Dropout

[TurboGeezer]

I am probably WAY oversimplifying things here, but my understanding is that the ECM "sees" a voltage, that it corresponds to 255 g/sec, for a length of time that caused the ECM to reset itself. Does this mean that you could simply change either the MAF table scalars or corresponding g/sec values, or something as simple as that, in order to prevent MAF dropout?? (prevent it from ever really seeing a true 255?) Otherwise, I suppose you could keep replacing the MAF value, etc. for anything above a preset point (ie 245) with a "locked" value of 255........sort of like the BLM lock.

Has anyone successfully "fooled" the ECM with a homegrown "brew"?? It just seems that keeping the ECM from "resetting" might just be simpler than locking the BLM.

Any thoughts??

TG

 

[TurboJim]

Chuck:

As I understand it, MAF dropout occurs when the ECM "overflows" says this cant be right, then resets the MAF counter.

ONly person I know fixed this is Bob, who did it by rewriting code. I cant say what he does. I'll let him spill the beans if he wants.

 

[bruce]

Originally posted by TurboGeezer
I am probably WAY oversimplifying things here, but my understanding is that the ECM "sees" a voltage, that it corresponds to 255 g/sec, for a length of time that caused the ECM to reset itself. Does this mean that you could simply change either the MAF table scalars or corresponding g/sec values, or something as simple as that, in order to prevent MAF dropout?? (prevent it from ever really seeing a true 255?) Otherwise, I suppose you could keep replacing the MAF value, etc. for anything above a preset point (ie 245) with a "locked" value of 255........sort of like the BLM lock.

Has anyone successfully "fooled" the ECM with a homegrown "brew"?? It just seems that keeping the ECM from "resetting" might just be simpler than locking the BLM.

There are several causes for MAF Dropout.
One being in the actual code, and several mechanical/electrical. If the voltage regulator IN the MAF is going bad that can cause it, as well as the sensor's element dieing. An out of range voltage/frequency are those causes.

Best answer I've seen is the Translators and Extender chip.

 

[TurboGeezer]

High Boost Drop out.......

What exactly is happening then when MAF dropout is ONLY happening at high boost levels (ie 22+ psi) ??? It seems that this condition would be code related (long periods of high flow). What is really strange is that I have seen dropout occur about once every second or so during high boost conditions.........that seems fairly rapid.

TG

 

[ijames]

First, when you see repeated dropout every second or so, what is most likely happening is a dropout which causes the engine to miss from going way lean, so the boost drops because for a little while the engine isn't running, then the maf drops until it is read correctly and the engine fueled properly, then the engine fires and the boost rises, the maf drops out, and the cycle repeats.

Second, the maf outputs a variable frequency that is proportional to airflow. Assuming nothing is actually broken, as Bruce mentioned, some maf's can output a frequency that is higher than the ecm can measure. When that happens, the ecm "measures" something like the true maf minus 255, so when it should have measured 280 it gets 25 and fuels accordingly - see why it went way lean?

 

[TurboGeezer]

Soooooo........

So..........we must then either limit the MAF output that the ECM sees or fool the ECM into thinking it is receiving a particular value for any reading above a particular set limit...................again, would it make sense then to create an algorithm to replace the MAF "reading" with a set quantity for anything above the set limit?? In other words, once the MAF hits say "250" we just keep plugging in a value of say "250", every cycle, no matter what the real MAF reading is.........similar to a BLM lock?? Plus we would want to keep rechecking the true MAF value so that when it drops below 250, we resume normal operation..............

Is this kinda the thought process and where in the code should I look to do this sort of "pull and replace" ??? Thanks.

TG

 

[ijames]

Yup, you've hit on the two approaches people use. One of the very, very nice things about the translators is that they include a limit switch to keep maf overflow from happening. Also, some chipmakers (such as Bob in the Extender chips and Tom in the Thrashers, and others, I'm sure) use a code patch to lock the maf at 255 under circumstances where they think it should be 255. I like the way you say "keep checking the true MAF ...". If we could do that, we'd just use it, wouldn't we? No need to lock to any arbitrary value . (Oh, if you any DSP background, the best way to understand the maf overflow is that we've exceeded the Nyquist limit of our counting system.) As for where in the code to do this, I have to say I've never bothered to write a maf lock since I have a translator and all the other cars I make chips for have them as well.

 

[TurboGeezer]

Hmmmm........

Oooops, Carl, you're right. If we knew the "true" MAF value, then we wouldn't need to jump through all of the hoops.

Unfortunately, I am not working with a translator here...........too bad, I would then have access to a whole bunch of neat features!!!

Anyway, I guess you could lock the MAF value for TPS above 4.5V and MAF above 250..............then drop back to the "read" value when the TPS drops below 4.5V again.......................or something like that. Anyone have any experience creating such a thing??

BTW, I am a Mechanical Engineer with Electrical experience, in an electronics world (a necessary evil)............My usual experience with DSP is with packaging them so that they don't burn up or emit nasty "waves" into the air............plus I get to let the smoke out once in a while.

TG

 

[DavzGN]

You can't lock maf when tps > 4.5 and maf > 250 and prevent Maf dropout. When you have maf dropout, your maf would drop below 250 and stop your maf lock. Using TPS above ?.? and rpm > ???? might work.

David
87GN

 

[TurboGeezer]

Precisely......

David,

You're right. I meant that the MAF value could be used to set the condition and the TPS (or TPS & RPM) could be used as a latch. The first time the MAF value was greater than 250 and the TPS was greater than 4.5V, the lock would set. Then if the TPS ever dropped below 4.5 (regardless of what the MAF was really doing), the lock would "unlock".

 

[bruce]

And don't forget there are two GN injector worlds.
Those that just run em static, and in which case all you need is a TPS vs RPM for full on, vs actually having a fuel curve.

But in order to have a meaningful fuel curve means having a refernce to boost. Since the VE from N/A to Boost is so far different.

Unless, all you care about is Drag Racing. Then you can slide alot on drivibility.

I'd be interesting just to do an RPM biased lock. From what I've seen pretty much everyone pegs or maxs their airflow at 3,200-3,600 rpm.

 

[ijames]

And don't forget there are two GN injector worlds.
Those that just run em static, and in which case all you need is a TPS vs RPM for full on, vs actually having a fuel curve.

It's true that an 8 year old ATR chip I bought basically set the injectors to static in PE mode, but no chip from any of the major, name chipmakers that I've looked at in the last several years has done this (okay, with one exception in a race chip). Yes, the fuel curve might touch static at one rpm, or above some rpm, but that's a different strategy than simply setting the injectors static at all rpms.

 

[TurboGeezer]

Injectors and MAF

Many of the "street" chips that I have taken a look at, either run the injectors static (if they are using "small" injectors) above a particular RPM or vary the PE vs RPM table in relation to the fuel needs of the particular car (ie from DS data).

Hmmmmm, a MAF lock triggered by just RPM.......that is a thought. That may actually not be a bad idea considering that most of these vehicles max out the MAF, long before their RPM max. In fact, a conservative approach might be to moderately throttle up to say 5000 RPM and watch the "MAF values" and see at what RPM the MAF goes above say 240 or so, under a "conservative" amount of boost.

Now I just need to figure it all out..........the RPM trigger should not be too bad, but where to "stick in" that pesky MAF value...........??

TG

 

[bruce]

Originally posted by ijames
It's true that an 8 year old ATR chip I bought basically set the injectors to static in PE mode, but no chip from any of the major, name chipmakers that I've looked at in the last several years has done this (okay, with one exception in a race chip). Yes, the fuel curve might touch static at one rpm, or above some rpm, but that's a different strategy than simply setting the injectors static at all rpms.

Ugh, don't forget while the DC might be just be 95% DC, the injector is static, or should I say erratic at that point?.
If you use the design rules set out by some folks, you can consider 90% as the lower limit of stable injector behavior.
Lowest I've been able to get a one to chatter, er, go erratic, was at 92%. At 92% DC, lots if not all the smaller injector chips are just running them static. 95% hardly changes the issue, IMO.

 

[ITSAV6]

extender chip

What does the extender chip do?

 

[bruce]

Re: extender chip

Originally posted by ITSAV6
What does the extender chip do?

It corrects a code error, that allows one form of Dropout. When used with a Translator or Translator Plus, it eliminates any form of MAF Dropout errors.

Amongst other none Dropout related items.

 

[TurboGeezer]

Hardware Solution........

If the "dropout" problem is a matter of the ECM seeing a frequency that it does not like, would it be possible to create some kind of LRC filter to plug into the MAF and therefore filter out the errant signals??

It just seems that for the stock MAF, perhaps a "code" solution does not cover all of the bases.............??

TG

 

[bruce]

Re: Hardware Solution........

Originally posted by TurboGeezer
If the "dropout" problem is a matter of the ECM seeing a frequency that it does not like, would it be possible to create some kind of LRC filter to plug into the MAF and therefore filter out the errant signals??
It just seems that for the stock MAF, perhaps a "code" solution does not cover all of the bases.............??

At that point thou, you wind up reinventing the wheel.
For the time and effort you could easily have a T or T+.

And the newer sensor brings it's own gifts to the party. ie, it seemed to me to be alot better in hot to cold weather. Not to mention availability, and price.

 

[TurboGeezer]

MAF Locations

Does anyone have a list or can anyone list all of the places in the stock code where a "locked" MAF value would have to be "replaced" ?? I seem to remember (from past posts, etc.) that there are several places where the MAF value would have to be "plugged in" and I don't want to miss any. Thanks in advance!

TG

 

[bruce]

I'm a little over my head in repling, but as I see it you'd want to ID the RAM location that holds the MAF calc., and then just edit/ find thru the document for that location.

And if that still hangs, then there might be mention of it in the ROM code.

And you might want to be real surea bout code changes before driving them on a public highway.
And ecm bench is priceless in developement work like your mentioning.