P0172, Can't figure this one out.
03-22-2015, 02:12 AM
#1
Elite Member
Thread Starter
iTrader: (16)
Join Date: Aug 2007
Location: Houston, TX
Posts: 9,292
Total Cats: 475
P0172, Can't figure this one out.
1999 miata, factory ECU/harness, all emissions equipment in place and fully functional. 2001 engine runs great/good compression. I have the 2001 Injectors in the car.
I don't have my 99' injectors. I looked it up, 01's are 265cc/min, 99's are 256cc/min. Injector calculator shows running base pressure at 56 PSI instead of 60 results in these injectors spraying at 256cc/min instead of 265. Even at 60 PSI, these are only 3.5% bigger injectors.
Car drives great, runs great, only code is P0172- Fuel System Too Rich (Bank 1)
The car will pass all emissions monitors, just this one code keeps coming up! If I reset the ecu and drive it, 2-3 drive cycles this code comes up.
I have changed the front O2 sensor, it is new, and still get the same code with it installed as I did with the old sensor. Both sensors pass the O2 sensor test in the emission test done for OBDII diagnostics.
Fuel pressure set at 60 PSI. I have an adjustable FPR and a return style fuel system now. I've tried backing down the fuel pressure to say 56, 54, 52, 50, doesn't help. I believe 60 PSI is correct for a 99' miata. I have a new Marshal gauge to measure pressure.
For fun I figured let's see HOW LOW I'd have to go to get the STFT to zero out. I got to 40 PSI and it got close to zero STFT, then it went back to -15!!!! No idea how this is even possible.
Short Term Fuel Trims (STFT) are anywhere from -2 all the way to -25 at IDLE. I don't think the ECU will go over -25. After 2-3 drive cycles it throws a pending code, then a CEL for P0172. At cruise they are not as severe, and under moderate load they are close to zero, plus or minus a few. Idle seems to be by far the worst for STFT.
LTFT are -16.4
Graphing the front O2 sensor shows a waveform that goes from about 0.1 to 0.7, with a frequency of about 1.5 sec per cycle.
I have cleaned the MAF sensor 3 times today, once yesterday, once 5 days ago. I'm almost out of MAF cleaner now.
I have an OBD to Bluetooth dongle and was using the Torque App on my phone to see OBD II data. I will buy a code reader/better system tomorrow or Monday to get better data if needed.
On my Torque app, it shows the voltage at the OBD II adapter at around 13.1. This is in contrast to everywhere else I've checked. At battery, fuses under dash, fuses under hood, I show 14.95V plus or minus 0.05V. This stands out to me, I plan tomorrow to find the wire going to the ECU and see what voltage it's showing.
My only thought at this point is that if the ECU thinks the car is at 13.0V, and the Injectors are in fact getting 14.95V, then the ECU will be adding in to much Dead Time for the injectors, which would result in it squirting too much fuel. This effect would be worst at low injector pulse widths (like idle) which kinda makes sense. Will check voltage at ECU tomorrow to test this.
My second only other thought is a bad MAF sensor. I need a better code reader to see the g/s that it is measuring. The software I have won't show that, it just shows CFM which is around 28-29 at idle. I need to retest but I think it showed over 600 CFM driving the car WOT around 5-6K. I will retest this and confirm if this is true or not.
Any ideas gentlemen? I've been trying to fix this for a week now! I read every thread I can find that has P0172 in the title and still haven't figured this out yet.
I don't have my 99' injectors. I looked it up, 01's are 265cc/min, 99's are 256cc/min. Injector calculator shows running base pressure at 56 PSI instead of 60 results in these injectors spraying at 256cc/min instead of 265. Even at 60 PSI, these are only 3.5% bigger injectors.
Car drives great, runs great, only code is P0172- Fuel System Too Rich (Bank 1)
The car will pass all emissions monitors, just this one code keeps coming up! If I reset the ecu and drive it, 2-3 drive cycles this code comes up.
I have changed the front O2 sensor, it is new, and still get the same code with it installed as I did with the old sensor. Both sensors pass the O2 sensor test in the emission test done for OBDII diagnostics.
Fuel pressure set at 60 PSI. I have an adjustable FPR and a return style fuel system now. I've tried backing down the fuel pressure to say 56, 54, 52, 50, doesn't help. I believe 60 PSI is correct for a 99' miata. I have a new Marshal gauge to measure pressure.
For fun I figured let's see HOW LOW I'd have to go to get the STFT to zero out. I got to 40 PSI and it got close to zero STFT, then it went back to -15!!!! No idea how this is even possible.
Short Term Fuel Trims (STFT) are anywhere from -2 all the way to -25 at IDLE. I don't think the ECU will go over -25. After 2-3 drive cycles it throws a pending code, then a CEL for P0172. At cruise they are not as severe, and under moderate load they are close to zero, plus or minus a few. Idle seems to be by far the worst for STFT.
LTFT are -16.4
Graphing the front O2 sensor shows a waveform that goes from about 0.1 to 0.7, with a frequency of about 1.5 sec per cycle.
I have cleaned the MAF sensor 3 times today, once yesterday, once 5 days ago. I'm almost out of MAF cleaner now.
I have an OBD to Bluetooth dongle and was using the Torque App on my phone to see OBD II data. I will buy a code reader/better system tomorrow or Monday to get better data if needed.
On my Torque app, it shows the voltage at the OBD II adapter at around 13.1. This is in contrast to everywhere else I've checked. At battery, fuses under dash, fuses under hood, I show 14.95V plus or minus 0.05V. This stands out to me, I plan tomorrow to find the wire going to the ECU and see what voltage it's showing.
My only thought at this point is that if the ECU thinks the car is at 13.0V, and the Injectors are in fact getting 14.95V, then the ECU will be adding in to much Dead Time for the injectors, which would result in it squirting too much fuel. This effect would be worst at low injector pulse widths (like idle) which kinda makes sense. Will check voltage at ECU tomorrow to test this.
My second only other thought is a bad MAF sensor. I need a better code reader to see the g/s that it is measuring. The software I have won't show that, it just shows CFM which is around 28-29 at idle. I need to retest but I think it showed over 600 CFM driving the car WOT around 5-6K. I will retest this and confirm if this is true or not.
Any ideas gentlemen? I've been trying to fix this for a week now! I read every thread I can find that has P0172 in the title and still haven't figured this out yet.
Reply
1
1
03-22-2015, 02:29 AM
#3
Elite Member
Thread Starter
iTrader: (16)
Join Date: Aug 2007
Location: Houston, TX
Posts: 9,292
Total Cats: 475
Will test voltage tomorrow, car is too loud to start right now this late, I gotta have some respect for my neighbors.
Injectors I have no idea, they are from a 2004 miata with 48K miles, but that's still kinda old so who knows. It holds fuel pressure fairly well when I shut it down. If I set pressure at 60 PSI, after say 10 minutes it will still be a bit above 50, after several hours it will be 20-30. I have a Aeromotive FPR and Walbro 255HP.
Injectors I have no idea, they are from a 2004 miata with 48K miles, but that's still kinda old so who knows. It holds fuel pressure fairly well when I shut it down. If I set pressure at 60 PSI, after say 10 minutes it will still be a bit above 50, after several hours it will be 20-30. I have a Aeromotive FPR and Walbro 255HP.
Reply
1
1
03-22-2015, 04:10 PM
#6
Senior Member
iTrader: (2)
Join Date: Oct 2013
Location: Goleta, Southern California
Posts: 520
Total Cats: 27
Exhaust leak upstream from O2 would cause the engine to run rich but the ecu wouldn't see it that way. It would give a lean code.
Edit: I believe you're on the right track with MAF if injectors are not leaking/sticking. I don't know that the O2 has the ability to make fuel trims go into the 20s. However I do know that MAF can and will.
Edit: I believe you're on the right track with MAF if injectors are not leaking/sticking. I don't know that the O2 has the ability to make fuel trims go into the 20s. However I do know that MAF can and will.
Reply
0
0
03-23-2015, 02:09 AM
#7
Elite Member
Thread Starter
iTrader: (16)
Join Date: Aug 2007
Location: Houston, TX
Posts: 9,292
Total Cats: 475
Ok I did a little testing tonight. But first let me answer the questions asked. Also sorry it's long but did a lot of testing.
Misfire- None that I can detect. I idled the engine down to 400 RPM as it's easier to detect a miss at lower RPM. No miss detected, I'm 99% sure there is no misfire happening at all. Engine feels smooth, exhaust has no change in pitch even at 400 RPM.
Exhaust Leak- I don't think so. At worst there could be a pinhole or two that I can't find/hear. From what I understand an exhaust leak would let air/O2 into exhaust and cause a lean reading.
Back to testing!
Voltage at the OBD II port is a bit lower than everything else. I tested this with the Key On Engine Off.
Battery Voltage 12.10
Voltage at Front of car, inside main fuse panel 12.09
Voltage at fuses at main fuse panel 12.09
Voltage at fuses at Fuse panel under Dashboard 12.08
Voltage at OBD II connector 11.51
Voltage reported at OBD II connector with OBD II to Bluetooth Device: 10.3
I stopped and looked at wiring diagrams for a 99', but I could not find the OBD II port listed anywhere. Was wanting to trace where the OBD II port gets it's power from. My first guess is those wires go to the ECU harness/ECU.
So then I stopped and changed gears, went and got a better OBD II scanner. This one isn't perfect, but it's better. I'll have a 500 dollar scanner tomorrow that will do whatever a 500 dollar OBD II scanner does. Today I had a 180 dollar one.
I watched the O2 sensor response vs Short Term Fuel Trim, graphically. STFT follows O2 exactly. So if O2 says rich, it STFT pulls fuel until O2 says lean, then STFT adds fuel, etc. This cycle averages 3 sec/cycle, over 20 cycles, but it varies a bit from cycle to cycle. Based on this data, it seems the O2 sensor can very well cause large changes in STFT. Also I'm pretty sure it should be way faster than 3s/cycle. Anyone know for sure?
To test this, I introduced a big air leak into the engine at the plenum, so the MAF would have no idea that extra air was going into engine. O2 went lean, STFT went all teh way to +25 to catch it and bring it back to spec. So confirmed O2 can cause large swings in STFT.
I also thought, what if an exhaust leak is letting exhaust gasses build up around O2 sensor. It compares oxygen in exhaust relative to the air around it. Put a squirrel cage blower fan blowing on O2 sensor, and another large 4,600 ft^3/min fan blowing over engine bay at ~20mph. No change in sensor response.
Also tried blocking each vacuum line, one at a time. No difference.
Onto MAF Sensor...
But by this point I was really starting to notice something. Hood open, STFT's get closer to zero AND MAF grams/sec come down. When I shut the hood, as things heat soak, the MAF Grams/sec starts coming up, and STFT's start going negative. The longer I let it idle/heat soak, the worse it gets. This is also true driving, STFT's are better/normal highway speeds, but at idle worse, and extended idle it throws a code.
If I open the hood, it slowly reverses. If I open the hood and put a huge fan blowing on the engine, it quickly reverses. Confirmed heat is causing the fluctuations in some form. Also Confirmed MAF sensor shows higher grams/sec when heatsoaked vs cooled by ambient airflow.
If I have a fan on engine bay, MAF shows around 3.2 g/s, STFTs avg -3 to -4.
If i let it heat soak really bad, MAF shows around 3.6 g/s, STFTs average -14.
I can't find a spec from Mazda, but forums say MAF should show 3.0 g/s. Mine shows, at the lowest, 3.2 g/s with an avg of -3 to -4 STFT. 3.2/3 = 6.66% more air, so if ecu commanded that much fuel, it would have to pull about 6% to get it back to stoich, and STFTs are close to -4.
Math shows the percent change in g/s vs change in STFT's are very closely related/proportional.
I believe the MAF sensor should not change it's readings in g/s whether hot or cold, right? My understanding is it has it's own temp sensor, and corrects for temp automatically. So hot/cold shouldn't matter. Idle speed never changed during all testing, 800 RPM, so grams/sec should stay more or less constant.
I also changed the AIT sensor as I have a spare, no difference.
It's also worth noting that the MAF sensor readings are noisy. As in, if it's showing 3.2 g/s, in any 2 sec period it varied from 3.08 to 3.32, and withing a 10 sec period it can vary from 3.02 to 3.42. It seems to me this number should be very steady, not bouncing around.
So from all this info, I'm not sure about what the voltage differences mean, or if that is significant or has anything to do with this problem, but based on what I'm seeing with heatsoak = higher MAF g/s = increasingly negative STFT's, I'm lead to believe the MAF sensor is malfunctioning.
Thoughts? Sorry this is long but I did a lot of testing tonight varying 1 thing at a time trying to pinpoint what affected what.
Misfire- None that I can detect. I idled the engine down to 400 RPM as it's easier to detect a miss at lower RPM. No miss detected, I'm 99% sure there is no misfire happening at all. Engine feels smooth, exhaust has no change in pitch even at 400 RPM.
Exhaust Leak- I don't think so. At worst there could be a pinhole or two that I can't find/hear. From what I understand an exhaust leak would let air/O2 into exhaust and cause a lean reading.
Back to testing!
Voltage at the OBD II port is a bit lower than everything else. I tested this with the Key On Engine Off.
Battery Voltage 12.10
Voltage at Front of car, inside main fuse panel 12.09
Voltage at fuses at main fuse panel 12.09
Voltage at fuses at Fuse panel under Dashboard 12.08
Voltage at OBD II connector 11.51
Voltage reported at OBD II connector with OBD II to Bluetooth Device: 10.3
I stopped and looked at wiring diagrams for a 99', but I could not find the OBD II port listed anywhere. Was wanting to trace where the OBD II port gets it's power from. My first guess is those wires go to the ECU harness/ECU.
So then I stopped and changed gears, went and got a better OBD II scanner. This one isn't perfect, but it's better. I'll have a 500 dollar scanner tomorrow that will do whatever a 500 dollar OBD II scanner does. Today I had a 180 dollar one.
I watched the O2 sensor response vs Short Term Fuel Trim, graphically. STFT follows O2 exactly. So if O2 says rich, it STFT pulls fuel until O2 says lean, then STFT adds fuel, etc. This cycle averages 3 sec/cycle, over 20 cycles, but it varies a bit from cycle to cycle. Based on this data, it seems the O2 sensor can very well cause large changes in STFT. Also I'm pretty sure it should be way faster than 3s/cycle. Anyone know for sure?
To test this, I introduced a big air leak into the engine at the plenum, so the MAF would have no idea that extra air was going into engine. O2 went lean, STFT went all teh way to +25 to catch it and bring it back to spec. So confirmed O2 can cause large swings in STFT.
I also thought, what if an exhaust leak is letting exhaust gasses build up around O2 sensor. It compares oxygen in exhaust relative to the air around it. Put a squirrel cage blower fan blowing on O2 sensor, and another large 4,600 ft^3/min fan blowing over engine bay at ~20mph. No change in sensor response.
Also tried blocking each vacuum line, one at a time. No difference.
Onto MAF Sensor...
But by this point I was really starting to notice something. Hood open, STFT's get closer to zero AND MAF grams/sec come down. When I shut the hood, as things heat soak, the MAF Grams/sec starts coming up, and STFT's start going negative. The longer I let it idle/heat soak, the worse it gets. This is also true driving, STFT's are better/normal highway speeds, but at idle worse, and extended idle it throws a code.
If I open the hood, it slowly reverses. If I open the hood and put a huge fan blowing on the engine, it quickly reverses. Confirmed heat is causing the fluctuations in some form. Also Confirmed MAF sensor shows higher grams/sec when heatsoaked vs cooled by ambient airflow.
If I have a fan on engine bay, MAF shows around 3.2 g/s, STFTs avg -3 to -4.
If i let it heat soak really bad, MAF shows around 3.6 g/s, STFTs average -14.
I can't find a spec from Mazda, but forums say MAF should show 3.0 g/s. Mine shows, at the lowest, 3.2 g/s with an avg of -3 to -4 STFT. 3.2/3 = 6.66% more air, so if ecu commanded that much fuel, it would have to pull about 6% to get it back to stoich, and STFTs are close to -4.
Math shows the percent change in g/s vs change in STFT's are very closely related/proportional.
I believe the MAF sensor should not change it's readings in g/s whether hot or cold, right? My understanding is it has it's own temp sensor, and corrects for temp automatically. So hot/cold shouldn't matter. Idle speed never changed during all testing, 800 RPM, so grams/sec should stay more or less constant.
I also changed the AIT sensor as I have a spare, no difference.
It's also worth noting that the MAF sensor readings are noisy. As in, if it's showing 3.2 g/s, in any 2 sec period it varied from 3.08 to 3.32, and withing a 10 sec period it can vary from 3.02 to 3.42. It seems to me this number should be very steady, not bouncing around.
So from all this info, I'm not sure about what the voltage differences mean, or if that is significant or has anything to do with this problem, but based on what I'm seeing with heatsoak = higher MAF g/s = increasingly negative STFT's, I'm lead to believe the MAF sensor is malfunctioning.
Thoughts? Sorry this is long but I did a lot of testing tonight varying 1 thing at a time trying to pinpoint what affected what.
Last edited by patsmx5; 03-23-2015 at 02:21 AM.
Reply
1
1
03-23-2015, 03:07 AM
#8
Elite Member
Thread Starter
iTrader: (16)
Join Date: Aug 2007
Location: Houston, TX
Posts: 9,292
Total Cats: 475
Small update.
The MAF on my car, that I bought used, is not in fact a mazda OEM sensor, but some aftermarket sensor. Hotwire is different, shielding around/before/after hotwire is different! Looks like all new guts in a modified housing. Highly suspect!
Contacted the vendor I bought it from, hoping I can get a replacement overnighted. I bet this is the problem now.
The MAF on my car, that I bought used, is not in fact a mazda OEM sensor, but some aftermarket sensor. Hotwire is different, shielding around/before/after hotwire is different! Looks like all new guts in a modified housing. Highly suspect!
Contacted the vendor I bought it from, hoping I can get a replacement overnighted. I bet this is the problem now.
Reply
0
0
03-23-2015, 02:02 PM
#11
2 Props,3 Dildos,& 1 Cat
iTrader: (8)
Join Date: Jun 2005
Location: Fake Virginia
Posts: 19,338
Total Cats: 573
Not likely as it's at home in a bin in my shed and I'm at work in a "bin" in my office.
I could ship it overnight tomorrow though, however I question the wisdom of that transaction given you could probably steal one and pay a criminal fine for less.
I could ship it overnight tomorrow though, however I question the wisdom of that transaction given you could probably steal one and pay a criminal fine for less.
Reply
0
0
03-23-2015, 02:06 PM
#12
Elite Member
Thread Starter
iTrader: (16)
Join Date: Aug 2007
Location: Houston, TX
Posts: 9,292
Total Cats: 475
Ok well no problem. The vendor I bought it from is going to send a replacement part, but regular shipping.... But thank you, and if for some reason THIS ONE isn't reading right I'll definitely take you up on that!
Reply
0
0
Thread
Thread Starter
Forum
Replies
Last Post
matthewdesigns
Suspension, Brakes, Drivetrain
1
10-10-2015 02:07 AM
