l2r99gst

iTrader: (2)

I know some of the newer GM vehicles use a 'cold wire' maf type setup, but I know very little about them. The hot-wire mafs I am referring to are not cold-wire inductance mafs. They operator as I explained above, with the current directly proportional to the mass airflow.

But, like I said, it was probably 6+ years ago since I did all of the research and read about all of the research of the GM Mafs and translators on the DSMLink boards.

Many people, including myself, ran them and the DSMLink creators even were writing code to directly run the GM MAF from the ECU. Everything that I can remember agrees with what I have been saying. I could very well be off in some regards, but I think my general though is correct and why MAF translators even work on our ECUs.

l2r99gst

iTrader: (2)

OK, I did a bit more research myself, using my trusty old DSMLink boards and everything that I have said was verified there by the DSMLink creators. So, Ted, I know that you think what I said was wrong, but it just isn't. Not in respect to how the GM MAF operates and in reference to how it calculates the mass airflow when using it with our ECU. It does measure mass airflow directly, without the need for any input from a temp or pressure sensor.

I say just let's agree to disagree here. I repect your opinion on many issues, but on this one, I don't think we are going to convince each other either way. Or, at the very least, let's take it to PM, so keep this thread from being even more cluttered with our posts.

Ted B

iTrader: (6)

Here is how the GM sensors function. Feel free to pass it along to the trusty old DSM link forums:

"The (GM)AC MAF sensor has a cylindrical housing through which intake air is passed. The air first passes over a ceramic resistor which measures the air's temperature. The incoming air is then directed over a heated film. This film is heated to and kept at a temperature of 75o C above the temperature of the incoming air as measured by the ceramic resistor. As air flow is increased or colder air is introduced into the sensor, more power will be required to keep the film heated to 75o C above the incoming air temperature."

"Inside the computer there are lookup tables which have pre-programmed information. The computer compares the signal from the MAF sensor with these tables to find the air flow rate in grams per second. The computer uses this reading plus engine RPM and air temperature to correctly adjust the air/fuel ratio."
http://www.tomco-inc.com/Tech_Tips/ttt15.pdf

Here's more:

"...the sensor (MAF) uses a heated element and thermistor to measure air mass. ... As the air passes into the air induction system, it cools the heated element, the amount of cooling is directly proportional to the volume of air entering the engine and inversely proportional to the temperature of the air entering the engine. This information is then processed by the computer and used to calculate the proper injector on-time and ignition timing."

http://books.google.com/books?id=9wd...um=9&ct=result

So what happens when the ECU cannot adjust for the incoming air temperature, you may get the incorrect air/fuel ratio and possibly ignition timing. Sounds familiar? Ask JohnBradley.

---------------------------------------------------------------------


More:

"While air temperature also affects density, cold air is denser than warm air; a similar amount of air can enter the engine at different temperatures. So an Intake Air Temperature (IAT) sensor is incorporated into the MAF to measure the incoming air temperature and that is taken into account by the computer. "
http://autotechrepair.suite101.com/a...r_flow_sensors

"The mass-air meter also includes a temperature sensor that provides a correction for intake air temperature . . . "
http://www.carcraft.com/techarticles...ion/index.html

"IAT is sometimes referred to as MAT (Mass Air Temp) "
http://www.diy-efi.org/gmecm/compone...o/sensors.html

l2r99gst

iTrader: (2)

Ted,

Can you take this to PM if you want to argue.

I can also put plenty of quotes that argue my point as well. Just to prove my point, here is one:

http://en.wikipedia.org/wiki/Mass_flow_sensor


But, can we please just bring this to PM if you want to talk about it. I already have everything I need to know. I don't need to argue anymore.

And when I said the 'trusty DSMLink boards', I meant the people who have completely disassembled the DSM ECU (like this forum is doing for the EVO ECU) and have not only incorporated direct GM MAF control, but also speed density on the stock ECU. It's not just Joe Schmoe I'm talking about. Take all of the knowledge about the ECU that you have seen in this forum and multiply it a few times, and you have what has been done with the DSM ECU by the DSMLink creators. I don't know why you had to be condescending there, just because you don't know the source. I have even seen you quote the disassemblers in this forum as 'rocket scientists'. So, why the condescending tone now...just because we are disagreeing?

All I am doing is passing on what I have read and experienced with the GM MAF on my own car as well as what I have seen from other's experiences as well as very well respected members in the community. Quoting hot-film sensors, or cold-wire sensors, or Ford sensors doesn't help much in this discussion. All I was trying to say from the start is that the GM MAF hotwire sensor directly measures mass airflow. Plain and simple. No temp or pressure sensors are needed for that calculation. For VE corrections, sure. For setting the initial temp wire, sure. For the mass airflow calc...no. The mass is directly measured and directly proportional to the current needed to maintain the temperature of the wire. Higher mass airflow due to more pressure or lower temps will need a higher current. Seems pretty simple to me.

jcsbanks

To add a datapoint here, the Subaru mass airflow meters on the GC8 models I worked with were hot wire or hot film. The output was non-linear, but once it had gone through the table to linearize it, the value was proportional to mass airflow. The only temperature sensor on that car was the coolant temperature sensor, the STI models added an IAT at the airbox, although it was not used in the ECU code. There was a MAP sensor used for boost control, and periodically this sensor was switched to atmosphere to sample BAP. There was no significant AFR drift in varying environmental conditions. More importantly, when I converted this ECU to speed density, a drift was introduced related to temperature, as you would expect, although it was less than predicted based on gas laws.

jcsbanks

Regarding 2 byte airflow limits, the frequency measurement performed by the Evo ECU uses timers, counters and interrupts. This is the value that is then divided by the crank angle sensor interval to give a value that becomes the MAF Hz we log.

Other than the limits of the sensor itself which I can't comment on, the only limit electronic limit I've come across so far is the 380 load limit which I have posted up about and how to alter it already. This would of course be likely to appear at peak torque, not peak power.

I think we'd all do a lot better with speed density conversion. It avoids all the issues of lift off problems near big turbos, and allows freedom of inlet size and placement.

Ted B

iTrader: (6)

We discussed this via PM, and after reading a few resources, you aren't likely to find IAT being considered in the ECU code, because it is already taken into account internally at the sensor. Hot wire and hot film sensors constantly measure ambient air temp via an IAT (internal or external), and their sensors are maintained at 100C and 75c (respectively) above ambient air temp.

MAFT ignores the IAT reading with a preset value, so that would compromise the accuracy of the MAF transfer function to some degree, which will impact fuel and ignition timing (both of which are affected by load). As to exactly what happens and to what degree it happens is subject to speculation at this point. However, it may be to blame for MAFT-related issues.

jcsbanks

Do you (or anyone) have an info (pressure drops, power gains on removal) on how restrictive the Evo's MAF sensor is at various power or airflow levels? Unscientific I know, but just looking through it, it looks restrictive. I wasn't clear from Aaron's info whether any testing had been done on Evo sensors, or had reached a clear conclusion.

Ted B

iTrader: (6)

I don't know, but I've been curious about the same thing. I have one here that I might just have flow tested.

Mellon Racing

iTrader: (38)

I tried multiple MAF-T boxes and GM MAF of various sizes and found that there were wild AFR changes that coincided with ambient temperature changes. This testing was done on my 91 TT Stealth which shares a similar ECU/MAS. Even though there was a much bigger temp shift while testing the stock MAS it clearly allowed consistent AFR from pull to pull. The GM MAF failed miserably every time no matter what model I tried. It doesn't appear that the GM MAF at least in this configuration was doing anything to compensate for temperature changes. It was sending a fixed IAT temperature to the stock ECU so it couldn't do anything about it either.


maf-t 43* vs 61* (18* shift)
http://img.photobucket.com/albums/v4...verlayLM-1.jpg

stock mas 33* vs 59* (26* shift)
http://img.photobucket.com/albums/v4...-12-04LM-1.jpg

I finally gave up on the MAF-T "solution" and began refocusing on how to go fast on the stock ECU/MAS so that I could get back to enjoying my daily driver.

l2r99gst

iTrader: (2)

John,

Your experiences and comments on the Subaru sensors above seem to coincide with my experiences and others I have known with the GM hotwire MAF. I used it for several years here in Chicago from winter to summer with no issues that I can remember. However, I rarely tracked the car, so perhaps I just didn't notice the AFR swings...but I did have a PLX wideband display mounted right on the steering column, so I should have noticed any significant deviation.

However, many people, like Mellon above, have experienced deviations using the MAFT. I just think there are many questions still unanswered.

As Ted mentioned, we discussed this thoroughly via PM and I think we came to 2 basic conclusions.

1. The IAT isn't needed to measure mass airflow. The measured current required to keep the hotwire at a specific temp is directly proportional to a mass airflow.

2. However, the hotwire MAF is calibrated with the hotwire being (usually) 100C over the incoming IAT. In something like the MAFT box, that doesn't sample an IAT, when the actual IAT changes compared to the hotwire is greater than or less than this 100C delta T, then the calibration of the measured mass airflow will be off. The question here is by how much? It doesn't seem to agree with the ideal gas law from some quick calculations that I did and without data regarding this, we can only guess.

I'm wondering if what you experienced with the Subura was because the actual IAT varied little from changing seasons or the wire was kept at a temperature where the temp variance had little effect to the calibration? These are just more questions that I have.

Anyway, time for speed density, so we don't have to worry about these things.


Eric

jcsbanks

UK temperatures are fairly flat, but in the -3 to +27C range or thereabouts (daily driver over the whole year, continuous wideband), the expected 1+ AFR difference didn't arise. There was a circuit board hidden under some epoxy or similar in these sensors, presumably that looked after the current source/temp comp etc, nothing was handled outside the sensor except linearization - just give it power and it gives something very close to mass air flow.

I think the differences in observations must be due to different sensor arrangements, perhaps some drift more with temp changes than others, perhaps some need the ECU to control aspects of the compensation?

I am planning to disassemble the full route from MAF pulse to injector pulse width and describe it.

Ted B

iTrader: (6)

I mentioned to l2r99gst via PM that it seems most MAFs have an internal IAC, and some of those may regulate the temp of the hotwire internally. If they do, that could give the outward appearance of no IAC, but somehow, IAC must be considered.

Until or unless someone can clarify the exact basis of operation of a MAFT, we can only speculate the result of ignoring the IAC, but it does not make for an ideal situation at best.