Automotive MAF sensor vs temperature

[Ryoko]

I'm working on a DIY EFI system for automotive use. I have a question about the mass air flow sensor operation. I know they measure the volume of air passing thru them. But the question is whether that measurement needs to be corrected for air temperature or if the MAF sensor does that by default. My guess is that it does it by default since temperature changes would cause the air density passing thru the sensor to change and thus affect the cooling rate on the hot wire. But I'm not positive about this.

(This questions applies to modern hot wire based sensors typically found in GM cars. I think they actually come from either Delco or Bosch.)

 

[jrmichler]

Start with the manufacturer's data sheet for your mass air flow sensor. All of that information will be in there.

 

[Ryoko]

Unfortunately, data sheets don't seem to be available -- at least none that I could find via google.

 

[MRFMengineer]

Some info here: http://autoditex.com/page/mass-air-flow-sensor-maf-19-1.html

 

[cosmik debris]

I'm working on a DIY EFI system for automotive use. I have a question about the mass air flow sensor operation. I know they measure the volume of air passing thru them. But the question is whether that measurement needs to be corrected for air temperature or if the MAF sensor does that by default. My guess is that it does it by default since temperature changes would cause the air density passing thru the sensor to change and thus affect the cooling rate on the hot wire. But I'm not positive about this.

(This questions applies to modern hot wire based sensors typically found in GM cars. I think they actually come from either Delco or Bosch.)

Yes, the volume of air must be corrected for temperature but in my 6 vehicles the temperature is measured separately by a temperature sensor. The ECU does the correction not the MAF sensor itself. This may not be universally true.

Cheers

 

[Ryoko]

Yes, the volume of air must be corrected for temperature but in my 6 vehicles the temperature is measured separately by a temperature sensor. The ECU does the correction not the MAF sensor itself. This may not be universally true.

Cheers

I'm finding mixed info on whether it's necessary to correct for temperature for a MAF. Some data I've read says that they are temperature compensated. But I'm seeing info elsewhere that suggests a correction is needed. I guess it depends on capability of the MAF unit being used.

As I understand it, the OAT sensor is used by the ECU to help with engine idle in cold weather and as a fallback to allow a MAF-based system to run in a limp mode as a speed-density configuration. (A MAP sensor is needed as well.)

 

[Mech_Engineer]

It's my understanding modern MAF sensors have both a flow sensing heated wire as well as a temperature sensing wire. Both measurements are needed to calculate the mass flow rate of air passing the sensor, but the calculation is done on the ECU side using the two input signals.

See here: http://www.denso-am.com/products/au...-systems/mass-air-flow-sensors/how-they-work/

Currently the most common MAF sensor is the plug-in hot wire type which is located inside the intake air duct between the air filter and the throttle body. This consists of a heating resistor, intake air temperature measurement resistor (for compensating intake air temperature), intake air temperature sensor, and control circuit (printed circuit board).

 

[Ketch22]

Is this a kit type DIY ECU or are you creating it? You must also understand that the ECU operates with multiple feedback loops. This is one of the simpler ones if any of them are. The MAF determines the amount of air entering the engine. The throttle position sensor establishes what power level the driver is requesting compared to the actual level of operation. The MAF sensor helps to relate how hard the engine is working. The programed fuel map determines how these items affect the actual fuel injection rate compares to the air flow. As a final check up the Oxygen Sensor provides feedback as to too rich or too lean for the current operating state. The ECU can adjust the injection rate to provide optimum performance.
All one needs to do is change the MAF for a MAP and the whole thing changes, it is still corrected by several feedback loops.
My question would be what is the reasoning of your question? Are you trying to create a fuel map structure? Are you just trying to figure out how a hot wire calculates air flow in varying conditions?

 

[Ryoko]

I under the difference between MAP and MAF based systems. The EFI I'm working on is intended to be a MAF system with a MAP fallback mode. The MAP (along with the TPS) will be used as a load sensor to fudge the mixture based on power demands. And yes, I have circuitry for an O2 sensor and dual knock sensors. (Also water temp, oil pressure, OAT, crank and cam position sensors.) I'm still debating whether to make it throttle-by-wire or staying with a cable throttle body. A TBW does allow a valet mode which would be handy.

 

[jim hardy]

I have interest in such a project. Did you start with a kit ?

"Circuit Cellar" magazine perhaps twenty five years ago had a three part series written by the fellow behind Cadillac's Northstar system.
He explained the reasoning behind his alogrithms. I'd sure like to unearth those articles.

Anyhow Good Luck and i hope you share your learning.

old jim

 

[Ryoko]

No. I started from scratch. I have a design which uses a pair of PIC processors and a lot of glue interfacing to talk to the various sensors and igniters. The idea is an extension of an earlier design which I made to replace the VAFM in my old 4Runner with a modern MAF sensor. It made a big difference. (+10% to 15% power gain.) Unfortunately, I couldn't leave it in because the prototype was not weatherproof or rated for under hood temperatures. It was just a proof of concept effort.

This new one I'm toying with is intended to run a modern LS engine. Yes, I know the LS already has an ECU. But 1) the tuning adapters for LS ECUs are horrifically expensive, 2) aftermarket EFI's aren't any cheaper, and 3) why the hell not? It's a fun project.

 

[jim hardy]

I'd love to do that and inject a metered trickle of windshield washer fluid(basically methanol and water) under load

that should lower combustion chamber temperature making EGR unnecessary
maybe even allow more spark advance ...
and a little bit of steam in the working fluid should go a long way.

so far it's just a daydream

Boring and questionably trustworthy anecdote---
i did run across a fellow who'd plumbed his windshield washer reservoir into the intake manifold of a SEFI small block Chevy. It was primitive, no metering just a restriction. He claimed 25 mpg in a full size pickup truck. I saw the odometer it showed over 400kmiles and the engine was quite smooth.

 

[Tom.G]

"Circuit Cellar" magazine perhaps twenty five years ago had a three part series written by the fellow behind Cadillac's Northstar system.
He explained the reasoning behind his alogrithms. I'd sure like to unearth those articles.

Would you accept 23 years ago?

http://www.diagram.com.ua/english/archive/circuit-cellar-magazine/1995/
About 20% down the page, "Developing An Engine Control System..." Parts 1, 2, 3

Cheers,
Tom

 

[jim hardy]

It's listed there but the page refuses to find it.

clicking them gives

Typical IT horsemanure - nothing works.

 

[jim hardy]

But THANKS - maybe i can find it someplace.

 

[jim hardy]

Found them !

the whole magazine is here
http://www.pestingers.net/pdfs/other-computers/circuit-cellar/1995/circuit-cellar-062.pdf

issues 62, 63 and 64 are now in my documents library

Thanks @Tom.G I've been looking for years !

old jim

 

[Mech_Engineer]

I'd love to do that and inject a metered trickle of windshield washer fluid(basically methanol and water) under load

that should lower combustion chamber temperature making EGR unnecessary
maybe even allow more spark advance ...
and a little bit of steam in the working fluid should go a long way.

so far it's just a daydream

@jim hardy you might be interested to know this is a common aftermarket application that's done by engine tuners. A search for "methanol injection" (also known as Water/Methanol Injection aka WMI) on an aftermarket parts site like Summit Racing yields hundreds of options. Most complete systems include tank for the water/methanol mix, a pump/injector/tubing for installation in your intake manifold, and an electronic control unit for metering the flow rate based on throttle position or some other useful metric like an MAP sensor.

In some more advanced applications you can also utilize a "piggyback control unit" which taps into the engine's signals to modify fuel, timing, and WMI mist in the engine; it all depends on how far you want to go.

 

[jim hardy]

you might be interested to know this is a common aftermarket application

I had no idea ! Thanks for the link .

I'll learn more about this.
I have an old Ford pickup with 240 six that's a candidate because it predates EPA.

Thanks again