MAF 0-5v signal wire not responding

Tiny
TRAVIS DUNCAN
  • MEMBER
  • 2003 MAZDA PROTEGE
  • 2.0L
  • 4 CYL
  • FWD
  • MANUAL
  • 143,000 MILES
My MAF signal wire is not getting voltage 0-5 volts.
•I have good ground to MAF
•i have good 12 volts power source to MAF •iv replaced the MAF just to see if it would change anything.
•i have traced every wire back to ECU, more than twenty times to check for frayed wires, but for some reason the signal wire does not fluctuate from 0-5 volts when rpm's are raised. Could this be a problem in my ECU? Everything else works just fine.
Do you
have the same problem?
Yes
No
Sunday, October 23rd, 2016 AT 9:07 AM

8 Replies

Tiny
CARADIODOC
  • EXPERT
What kind of running problem are you trying to solve? All of the diagnostic procedures call for viewing signal information on a scanner, and reading weight of the incoming air, not signal voltage. There are two important details when doing that. First, all sensor voltages are only accurate when taken when they are plugged in, so you will have to back-probe through the back of the connector. Second, the fresh air tube must be connected between the MAF sensor and the throttle body, with no leaks, so all incoming air has to go through the sensor.

Rather than trying to read voltages at the sensor, read them on the scanner instead. If a signal wire is cut or a connector terminal is stretched, the reading at the sensor could be okay while the reading at the Engine Computer is quite different. Have you checked the diagnostic fault codes?
Was this
answer
helpful?
Yes
No
Sunday, October 23rd, 2016 AT 1:29 PM
Tiny
TRAVIS DUNCAN
  • MEMBER
The problem I am having is, when I try to punch the gas it will bog because it is not able to read how much air is coming in the intake and cannot adjust the injectors to send the correct amount of fuel to the engine.
I am getting a MAF signal code (P0102) and I know it is not working properly because the signal from signal wire is supposed to change from 0-5 volts when the rpm's increased and yes of coarse I back probed it, I even cut some of the insulation of of wire to make sure I was getting good connection while reading it with my multi-meter.
And yes all of the piping/MAF components are connected properly.

What type of scanner would I need to read the signal like your saying? All I have is a 0Bll scanner and it only will give me codes. I have checked the connection pins on ECU Also, and everything checks out okay.
Was this
answer
helpful?
Yes
No
Sunday, October 23rd, 2016 AT 1:48 PM
Tiny
CARADIODOC
  • EXPERT
Please do not say, "of course I did... " I do not have any idea what you did except for what you tell me.

I never allowed my students to rip insulation from wires to take a reading, or worse yet, poke a hard-to-see hole in it. I know it can be a problem knowing if the meter's probe made contact with the terminal when you back-probe through the weather-pack seal, but if you get any voltage reading other than 0.0, you know youare in there. I had a dozen cars with prepared problems, and damaged insulation on wires provided an unfair clue as to where the previous people were looking, and it leads to corroded wires and more problems in the future. They were required to replace wires they damaged. Black tape is also unacceptable as it will unravel into a gooey mess on a hot day.

What you are using is not a scanner. You have a code reader. Professionals use scanners that cost between about $3000.00 and $10,000.00. You might find one at an auto parts store that rents or borrows tools. You can also find some inexpensive models on eBay that will do what you need for under $100.00. Be aware that those I have tried do provide sensor data, but the screens only have three or four lines of data, and they take as much as five seconds to update. Rapidly-occurring glitches and changes will be missed. These also do not have a record function which is useful when watching for intermittent problems on a test-drive.

If you are going to be using it a lot, you might ask around at some local independent repair shops if they have some equipment they like to sell when they upgrade every year. A lot of shops bought a Chrysler DRB3 in the 1990's because with an extra plug-in card, they will do emissions-related stuff on any brand of car sold in the U.S. Starting with 1996 models. I have one for all of my personal vehicles. All of these professional scanners update their displays many times per second. Also, true scanners are "bidirectional", meaning you can talk back to the computers and command them to do things like turn on relays and run the idle speed up and down.

Be aware there are some mass air flow sensors that put out a signal frequency, not a voltage. You must have a scanner to read those, but you are actually reading what the Engine Computer is seeing. That has been a problem for Ford owners. Your MAF sensor should read a voltage like on GMs and other imports.

The first thing I would suspect on yours is the power supply terminal is stretched in the connector body. That will cause the signal voltage to be always below the threshold at which it sets code 102. If your engine uses a MAP sensor too, it may be running in back-up mode off that sensor.

There is something else you can do to help me help you. Unplug the MAF sensor, turn the ignition switch on, then read the voltage on the signal wire. That is going to be either 0.2 volts or 5.0 volts. I will explain later what I am after when I am done fighting with this miserable computer!
Was this
answer
helpful?
Yes
No
Sunday, October 23rd, 2016 AT 3:04 PM
Tiny
TRAVIS DUNCAN
  • MEMBER
I was not sure if I was getting good connection and could not get anything to fit in the connector to back probe it the first time I tried. So I just removed some insulation just to get a good reading, with plans to just cut, solder and shrink wrap. I Later found something to probe it with

the connector and wires/prong inside the connector looks like it is in pretty good shape. What do you mean by stretched? As in the prong inside connector does not have good contact with the signal wire? Yes I do also have a MAP sensor. And if that is supposed to take over than why do I still have bogging issues?
I just checked the connector like you said and it is is getting a reading of 0.05 unplugged with ignition on.
Was this
answer
helpful?
Yes
No
Sunday, October 23rd, 2016 AT 3:42 PM
Tiny
CARADIODOC
  • EXPERT
Yes, "stretched" means the female part is too large in diameter. That comes from cleaning off excessive corrosion or from stuffing a meter probe in too far that is larger in diameter than the male part of the terminal. Less-obvious is a pushed-out terminal. Service manuals always list that as the first thing to look for, especially when there might be a plug that was recently disconnected for something else. In over thirty years of doing auto electrical work, I only ran into that once myself, and a friend ran into it once a few months ago. It is not that common, but do not overlook the possibility.

When there is a break in the signal wire circuit, whether it be a cut wire, or corroded terminal, since the wire inside the computer is tied to a lot of other circuitry, the voltage inside the computer on that line can "float" to some random value. For explanation purposes, the acceptable range of signal voltage from a sensor is 0.5 to 4.5 volts. Anything outside that range is what triggers a fault code. When the signal voltage is allowed to float, if it happens to stay within that acceptable range, the computer will regard that as a proper voltage, and it will try to run the engine on that. To prevent that, all signal circuits have an internal "pull-up" or "pull-down" resistor tied to 5.0 volts or to ground. Those resistors are so big electrically that they have no effect what-so-ever on a properly-working circuit, but when the circuit is open, (cut wire), they will put 0.0 or 5.0 volts there to force it to go to a bad state and set a fault code. Since you found 0.5 volts, that tells us you have a pull-down resistor, and since you don't have 0.0 volts, the wire can be assumed to be not open or grounded.

When a code is set related to the MAF sensor, the computer knows it cannot rely on the signal voltage, so it has two back-up strategies. One is to run on the MAP sensor. Chrysler is the only manufacturer that has consistently been able to make their engines run right with only a MAP sensor. All other manufacturers use a MAF sensor on all or the majority of their engines. MAP sensors are so sensitive that they could be used to measure engine speed because they can detect the tiny little extra vacuum pulse each time a piston takes a gulp of air. That is why a lot of engines can run on them when the MAF sensor has failed. The clue is they wont run right under all conditions because the computers are not designed or programmed for that to be the primary fuel metering calculation component, and that may be what you have.

The second strategy is most Engine Computers have the ability to "inject" an approximate value for a sensor that has been detected as having a defect, and run on that. For example, you know if an engine has been running for twenty minutes, the coolant temperature cannot really be minus forty degrees, but that is what would be read if the sensor was unplugged. The computer also knows that if the engine is running at 1,000 rpm's, there has to be air flowing into the engine, and there has to be fuel commanded from the injectors to go with it. There is one clinker I have to add for people researching this topic. There was a software problem with some GM scanners where they would display that injected sensor value instead of the actual value coming from the sensor. That made it look like the sensor was reading correctly, and there was no reason for the fault code to be setting. Normally we do not have customers' time and money to waste checking voltages, but in a case like this, when the scanner is displaying a value that does not agree with the fault code or the engine's operating condition, that is the time to take a reading right at the sensor, and compare it to the scanner's reading. Normally the computer will be running the engine on the injected value, but it will be telling the scanner the actual sensor value. The first time you run into this, you will be scratching your head until you are bald!

Going further without a scanner is going to be difficult. Also, I do not know if your signal voltage is supposed to go up toward 4.5 volts or down toward 0.5 volts when engine speed increases. Most MAP sensors' voltages go down as vacuum goes up, but that is not relevant to this sad story. On top of that, the service manual only says the scanner will list MAF as grams per second, and that it should go up with increased engine speed. I know what a volt is. I do not know what a gram looks like. Chrysler's DRB3 lists sensors with their signal voltages and with the value that it is computed to represent. In other words, the coolant temperature sensor signal voltage might be 3.5 volts, but it will also be listed as 182 degrees, for example. Almost all of the more expensive scanners work the same way.

Normally professionals do these electrical tests before condemning a sensor, like you did, then they order one if appropriate, like you did. This is one of those frustrating times when proper procedure didn't yield the expected results. Assuming we can trust the sensor is good, you might consider unplugging it, then using four jumper wires to connect the wires to the sensor. If you still get 0.5 volts on the signal wire, pop that signal jumper wire off, then measure right on the sensor's terminal. If the voltage comes up to normal while the engine is running, suspect that signal wire is grounded. The 0.5 volts you found can be explained by the meter picking up magnetic interference. You can prove that to yourself too by just setting the meter on the running engine without the probes connected to anything.

If you should find the signal voltage is correct and responding normally, but you still have a hesitation problem, fuel pressure is the next thing to look at. From driving around with a fuel pressure gauge attached to my old 1988 Grand Caravan for a year, I know that normal pressure is around 50 pounds, and the engine runs fine down to twenty pounds. That is unusual. A lot of engines develop hesitations and / or failure to start if fuel pressure is only four or five pounds too low.

There is one more parting comment about digital voltmeters. In my forty years as a television repairman and thirty years as a mechanic and instructor, I have accumulated over a dozen digital meters, but I refuse to own one with the "auto-ranging" feature. I have used them, and often went down the wrong path from failing to notice that the meter changed ranges on me. I have found "15 volts", which is good, when in reality, I had 0.15 volts, which in effect, is 0 volts, and that is not good. If you are using an auto-ranging meter, you wont be the first person to be fooled. Somewhat related to that, if you are on a scale that is too high. Say the "200-volt" scale, there is very little difference between 0.5 and 4.5 volts. All meters have a tolerance, meaning the amount they can be off, and that typically includes a percentage of the voltage being read, and one to three steps in the last digit. That means a reading of 0.5 on the 200-volt scale could actually be 2.8 volts. One percent of 200 volts is two volts, and three numbers higher on the last digit comes to 0.8. A meter could read only 0.5 volts when measuring 2.8 volts, and still be considered to be in calibration. To prevent that error, always use the lowest scale possible that does not send the meter into an over-range condition.
Was this
answer
helpful?
Yes
No
Sunday, October 23rd, 2016 AT 5:21 PM
Tiny
RENEE
  • ADMIN
Great information. Thank you CARADIODOC!
Was this
answer
helpful?
Yes
No
Monday, October 24th, 2016 AT 12:20 PM
Tiny
TRAVIS DUNCAN
  • MEMBER
I am not going to keep bothering you but I do really appreciate you taking your time explaining this to me. It is hard to find someone so knowledgeable to help and respond so fast.
But I have pretty much ruled out the terminal connection problem by disconnecting the ecu and using a used nine volt battery (only about three volts left on it) to put power to the harness and using a multi-meter to check the unplugged connector On the other end by MAF. So the connection seems to be okay.

But the new MAF I bought while connected and back probing the signal wire reads only 0.8- 0.9 volts while running and raising RPM. It doesn't change. And the Actron scanner I have gives me a low signal code and I have a monitoring feature on the scanner and it reads 0.00(GR/SEC) for MAF.

So I installed the original one I took off that was also giving me trouble. And it gives me a code that signal is too high. Which I tested the same way by back probing and I get a high reading of around (11.34 volts) which is way to high. And I forgot to check the monitoring feature. If that would help you. I can check tomorrow with the old one again. But it is obviously bad if its reading that high of a voltage correct? So maybe the new one I bought is bad also or could have got shorted out somehow.

Oh, and when you asked me to turn ignition on and check the voltage of the unplugged signal connector it read 0.05v and not 0.5 like you said. Or is that the same thing?

And I don't think my MAP Sensor is able to do the job because the car still runs strange. And I'm running pretty rich 7.4 afraid to start up and it goes up to 10.4 and idle when warm but as soon as I rev it, it drops back down to 7.4-7.9. I know this because I have a wide-band 02 sensor

And I also have a adjustable FPR with gauge which is set to spec fuel pressure. So I am sure it is not the fuel system.
Was this
answer
helpful?
Yes
No
Wednesday, October 26th, 2016 AT 8:07 PM
Tiny
CARADIODOC
  • EXPERT
Do not sweat the 0.5 vs. 0.05 volts for this problem. It is too low; that is all that is important.

I probably did not explain the MAP sensor properly as the back-up strategy. On a lot of cars, the engine will not start or run with a failed MAF sensor. On some, the Engine Computer uses MAP readings as a back-up strategy to allow the engine to run well enough so you don't need a tow truck. The engine is not going to run without issues, but it will run.

You have a much better handle on this than most people would, but I would like to point out one thing you should keep in mind for the future. A digital voltmeter has an extremely high internal resistance, (impedance). For all practical purposes, no current flows through it for it to work. When you apply, or have a voltage on one end of a wire, then measure it on the other end, you aren't really testing that wire. Suppose there is a splice buried in the harness that is heavily-corroded and is introducing 1 meg ohms of resistance. The voltmeter will still see the full voltage at the other end of the wire, but you'd ever get enough current to flow through that resistance to do any usable work. You need to cause current to flow through the wire so the results of that resistance would show up. With current flow, voltage will be dropped across that undesirable resistance. In this sad story, that resistance is real high, so the voltage dropped across it will also be real high, as in close to 12 volts. That would leave you near 0 volts measured at the end of the wire. I should make that 3 volts since that's what you were using. 12 volts is what is normally used when talking about car circuits.

Think of a pressure gauge on a compressed air line. No air flows through the gauge for it to work. Suppose you have a hundred pounds of pressure at the start of a long pipe run. Halfway down that line the pipe is 99 percent restricted. You'd still read 100 pounds at the end of the line, until you tried to have some flow to run an air tool. Then the pressure would drop to almost 0 pounds. The only way to test for that restriction is to measure the pressure at the end of the line while trying to run an air tool.

Similarly, the only way to check a wire is to cause current to flow through it. One easy way to do that is to use a test light instead of a voltmeter. Test lights work by the current flowing through them, and you won't get enough current if there is resistance in the wire.

I got confused with the chain of events as to the various voltage readings, but I am getting the feeling there is a high resistance in a wire based on your two extreme voltages. One is maximum low and one is maximum high, and there is no in between. That is the kind of thing we'd see when measuring a wire that is open circuit at the end vs. One we are trying to draw current through an unwanted resistance. I wish I could be there to look over your shoulder. Instead of using a wimpy 3 volts, either apply 12 volts and use a test light, even a head light bulb, or measure the wire's resistance with an ohm meter.
Was this
answer
helpful?
Yes
No
Wednesday, October 26th, 2016 AT 10:02 PM

Please login or register to post a reply.

Recommended Guides