P0171 Driving me Insane!!!

[XJSteven]

Finally! i know people hate us mechanics but with the time and money you wasted on sensors you could have had it fixed all ready! just my 2 cents.

Maybe so but I have learned a lot over the last few months and now have some spares to carry around with me on my off-road trips, not to mention the handy fault finding tools/techniques. I will definitely save money in the long term because I won't have to run off to the mechanic for every little problem. For example, I had my first 02 sensor replaced by a mechanic at a cost of 5-6 times what it cost me to order one online and put it in myself.

Sorry to have butted in. Mike obviously knows more than I do.

I appreciate all the help I can get so stick around Zuki-Ron! I just wasn't sure what you meant in your post so I needed some more info from you. Thanks.

 

[XJSteven]

I thought I read your post about what the LPG guy said, that the LPG does not use O2 sensor data at all, but uses a feed ratio of LPG to air based on vacuum pressure, with some kind of pressure flow control gadget that is set at idle?

Sorry. I've had numerous conversations with multiple LPG specialists now. This is what I posted after I spoke to the 1st guy:

I asked an LPG specialist today how it works on LPG (when the injectors aren't in use) and he said it uses the Converter at idle and then a Stepper Motor at higher revs using data from the TPS and 02 sensor. I don't know if this bit of info helps with my problem?

I contacted a 2nd LPG guy to get some answers to your questions:

One conclusion I have is that the computer is trying to control LPG fuel flow rate, but is not having any success!

I am not sure how the Chrysler PCM knows how to control it, since it was designed to operate 6 independent petrol fuel injectors, and not a throttle body single injector?

Looking at this data again, just on the LPG, it sounds like the LPG feed rate regulator is not working, like the PCM is not able to control the feed rate?

I spoke to the LPG repairer...and found out the PCM doesn't control the LPG mixture at all. He said "the *gas computer shuts the fuel (gasoline) off" and the LPG mixture is set on the Converter - one bolt is the Idle and the other is the Air/Fuel. *This should have read LPG computer.

The next day I spoke to the 1st LPG specialist again:

I spoke to the 1st LPG installer again today...He said adjusting the Converter won't help me because that's only for idle...and "after that the stepper motor takes over". He said the LPG computer should still try and lean it out if it was running too rich.

Anyway, here's an update:
I rang a 3rd LPG place today and I might finally have some answers! Apparently when it's running on LPG the injectors keep on firing but with a reduced power supply; not enough voltage to operate but enough power to satisfy the PCM and prevent an error code. The PCM doesn't know the gasoline has been disconnected.

In regards to the Fuel Trim figures the PCM believes it is still running on gasoline so can get a bit confused when it first switches over. This surprises me because when I switch from one fuel to the other the LTFT change seems to be instantaneous (from +15% on gasoline to -32% on LPG and vice versa).

He started explaining how some LPG systems use an emulator to give dummy feeds, or something similar, but then he lost me after that.

 

[XJSteven]

The mechanic who had the car all day came to the conclusion that there is no problem with the gasoline side of things. :huh:

He was going to adjust the LPG mixture at the converter but after calling his own LPG guy was told the stepper motor may be faulty. He said they are known to stick which could explain the sudden change from too lean to too rich. I think Ecomike (and McQue?) came to a similar conclusion a few pages back:

What ever controls the LPG flow hardware wise (not the PCM), that is comparable to a fuel injector is what I would look at next.

I will hopefully know for sure once I get it in for further testing of the LPG stepper motor in a few days time. Thanks for all the help from everyone so far.

 

[Ecomike]

Ah,ha, so the LPG system has its own stepper motor!

Sounds like real progress this time!

I would check the petrol system while it is on LPG and see if the injectors are clicking (stethoscope process using a screwdriver to listen for clinking sounds) and test the fuel rail pressure to make sure the fuel pump is not still running, and see if injectors are firing or not on that low voltage trick he described. Also one or more petrol injectors might be leaking if the petrol pump is still running. Just a good idea to check these, not saying there is any indication they are not working the way the LPG says they should. Just a good idea to verify now that we know the LPG games it plays to fool OBD-II.

 

[gradon]

When I was diagnosing my no-start, I had the fuel rail out with injectors still connected and ignition on to prime the rail to 49psi(might have to prime a couple times), then checked for leaks. Takes 30 min and costs 6 new o-rings(have to buy two of the 4-packs). So the system uses the TPS and o2 as parameters and then tells the stepper to adjust? I still find it interesting how the system would interrupt all 6 injector feeds and drop the pcm's 12v power via some resistance to some lower voltage that isn't enough for them to open.

 

[XJSteven]

Ah,ha, so the LPG system has its own stepper motor!

So the system uses the TPS and o2 as parameters and then tells the stepper to adjust?

Here's some more info I found from a quick online search:

The system starts the engine on gasoline using the injectors and then switches over to LPG switching off the injectors. Then using its own computer connected through the car's PCM, it computes the correct LPG/air ratio using information from the sensors on the engine. To actually change the mixture the LPG system uses a stepper motor on the LPG line just before the mixer.

Stepper motors have been know to jam which closes all LPG to the engine and if the valve happens to fail it allows the Converter to breathe as normal causing it to run richer than normal but function quite OK.


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


Normally placed midway between the converter and the mixer you will find a Stepper Motor (closed loop system). This stepper motor opens and closes the vapor hose passage, and is electronically controlled by a fuel processor PCM.

Fuel processor PCM's obtain signals from various sensors on the vehicle and then adjust the fuel volume to achieve the optimum air fuel ratio or mixture. The signals that are typically used are:

• Engine revolutions per minute (RPM)
• Throttle position
• Exhaust oxygen content.

Closed Loop - Closed loop control utilizes feedback from the vehicles oxygen sensor which is positioned in the exhaust system. As the name suggests the oxygen sensor measures the amount of oxygen in the exhaust. The amount of oxygen in the exhaust gas is a direct reflection of the mixture entering the engine. A lean mixture (insufficient fuel) will result in excessive oxygen in the exhaust and a rich mixture (excessive fuel) will result in very low levels of oxygen in the exhaust. A closed loop system therefore uses this signal from the oxygen sensor and adjusts the amount of fuel to constantly maintain the correct air fuel ratio or mixture.

 

[XJSteven]

I would check the petrol system while it is on LPG and see if the injectors are clicking (stethoscope process using a screwdriver to listen for clinking sounds) and test the fuel rail pressure to make sure the fuel pump is not still running, and see if injectors are firing or not on that low voltage trick he described. Also one or more petrol injectors might be leaking if the petrol pump is still running. Just a good idea to check these, not saying there is any indication they are not working the way the LPG says they should. Just a good idea to verify now that we know the LPG games it plays to fool OBD-II.

When I was diagnosing my no-start, I had the fuel rail out with injectors still connected and ignition on to prime the rail to 49psi(might have to prime a couple times), then checked for leaks. Takes 30 min and costs 6 new o-rings(have to buy two of the 4-packs).

Thanks, I will keep these ideas in mind. It's going in for repair tomorrow so the first thing I will do when I get it back is recheck the fuel trims and 02 sensor data and see if there has been any change.

 

[XJSteven]

Well I spent $400 at the mechanic today...LPG stepper motor replaced, converter cleaned/adjusted, and a smoke test found a vacuum leak at the LPG mixer/air intake which has now been sealed.

I got home and hooked up the OBDII software only to see the same fuel trim figures and a fresh P0171 fault code.

I guess I'll know for sure if anything has changed next time I fill up and calculate the fuel usage.

 

[Ecomike]

Well you found one of the vacuum leaks! LOL

Time to sniff out the ECU wiring on that O2 sensor?

Any way to find out if the LPG or petrol mode is throwing the code?

 

[four_shot]

The best part about spending money on a tech, is that if they "fix" a problem, and it comes back, they (at least a reputable establishment) Should delve further at no cost to the customer due to a misdiagnosis. The key here is if you mentioned the Check engine light in the original complaint. @ $400 Au. I'd be throwing a fit. (i've seen customers throw fits for less @ my shop.)

 

[XJSteven]

Well you found one of the vacuum leaks! LOL

True - it's good to know there are no vacuum leaks now but I'm disappointed the vacuum gauge I bought didn't pick up a problems when I tested for vacuum leaks a few weeks ago.

Any way to find out if the LPG or petrol mode is throwing the code?

I've had the P0171 code come back 4 times since I bought the OBDII software for my laptop but each time it has only shown itself the next time I've started the EasyOBD2 software, so I haven't been able to capture the data using freeze-frame. Maybe the lean condition causing the error code is only present on engine start up when the EasyOBD2 software is still running through it's comm port checks etc and establishing a connection to the PCM (which usually takes 2-3 mins)? If this is the case it might be hard to tell which fuel source is causing the code because the LPG system uses petrol for the first few seconds on start up.

After the next tank of LPG I might drive around on gasoline for a while to see what happens to the P0171 code.

Time to sniff out the ECU wiring on that O2 sensor?

Any suggestions on sniffing this out? I have already tested the wire from the 02 sensor to the ECU connector with a multimeter and to answer your earlier question (below) the reading probe to probe was 0.7 so the 0.9 reading I reported was actually 0.2 ohms which seems OK.

If the reading probe to probe is say .8 and the reading on the wire is .9, then the wire is .1 ohms which is OK. 1 ohm on the O2 sensor wire might be getting a little high.

 

[XJSteven]

The best part about spending money on a tech, is that if they "fix" a problem, and it comes back, they (at least a reputable establishment) Should delve further at no cost to the customer due to a misdiagnosis. The key here is if you mentioned the Check engine light in the original complaint. @ $400 Au. I'd be throwing a fit. (i've seen customers throw fits for less @ my shop.)

Well it hasn't been setting the Check Engine Light and I didn't specifically mention the P0171 code but it is the only code visible when you hook up a scanner so I thought it was obvious. I wish I'd told them that now! The description I gave them of the problem was poor fuel economy (using 25% more LPG) and I also provided details of the STFT +32% and LTFT +32%, except for in PARK under 2500rpm when both the short/long trims show -32%.

 

[Ecomike]

The best part about spending money on a tech, is that if they "fix" a problem, and it comes back, they (at least a reputable establishment) Should delve further at no cost to the customer due to a misdiagnosis. The key here is if you mentioned the Check engine light in the original complaint. @ $400 Au. I'd be throwing a fit. (i've seen customers throw fits for less @ my shop.)

"reputable establishment", In my opinion those are only legends!!!!!:laugh:

I have yet to find one in my lifetime!

I gave up on using others long ago. They charge way too much, do way too little, and frequently break more than they fix, and they take way too long to break things, tying up the vehicle in the process.

 

[Ecomike]

You might take it to a muffler shop and ask them to use their trained ear and steth-scope to listen for an exhaust leak on the E manifold, or crack.

Vac gauge is good for finding major hardware problems, like bad valves, large vac leaks, blocked exhaust, blocked air filter, etc. Not much good at finding small leaks like a cracked E manifold that is confusing an O2 sensor at certain engine speeds.

You might try clearing the code and then do a restart to test the lean at start up theory?

Perhaps it is detecting a problem during the fuel switching process at start up? Too lean during the switch over?

.02 ohms on the O2 wire to the ECU is OK.

Any word on finding a manual for the LPG system?

Have you tried comparing live OBD-II O2 sensor data (from the PCM) to a volt meter on the O2 sensor when it reads .8 V steady at the O2 sensor on petrol at 2500 rpm?

I still find it odd that you have a steady rich signal for the O2 sensor at one point but the PCM is throwing a lean code, when the O2 sensor seems to be working and the PCM seems to have a good electrical connection to the PCM, and the sensor and PCM seem to be working at idle and 1500 rpm on petrol. Has the petrol STFT and LTFT improved at idle and 1500 rpm, now that the LPG vac leak is fixed?

It is my understanding that OBD-II does a fault count on something like a too lean situation, and does not throw the error code and CIL until it has seen the error something like 200 times during a cycle period. Not sure what the cycle period is, but the software should be able to show you pending codes where there is an error count above zero heading for the magic 200 times count.

For instance a cyl miss fire has to reach 200 miss fires in a certain number of firing attempts, or period of time to throw a code and cut off fuel to that cylinder. Until then, only live data like you have can show pending codes, so look for a pending code menu!

 

[XJSteven]

You might try clearing the code and then do a restart to test the lean at start up theory?

I have tried this and did it again today, but it's intermittent so doesn't come back straight away.

Any word on finding a manual for the LPG system?

Still waiting to hear back from the manufacturer. As another Chinese Ebay seller told me earlier this week..."Please wait patience, and maybe you will get it more soon". I've taken his advice but it's not working too well. haha.

Have you tried comparing live OBD-II O2 sensor data (from the PCM) to a volt meter on the O2 sensor when it reads .8 V steady at the O2 sensor on petrol at 2500 rpm?

Results are as follows:

LPG
Idle: 0.940V (PCM) / 0.980V (voltmeter)
1500rpm: 0.920V (PCM) / 0.960V (voltmeter)
2500rpm: Oscillating at both PCM and voltmeter.
* Starts oscillating over 2000rpm.

Gasoline/petrol
Idle: Oscillating at both PCM and voltmeter.
1500rpm: Oscillating at both PCM and voltmeter.
2500rpm: Approx 0.760V (PCM) / approx 0.810V (voltmeter)

Has the petrol STFT and LTFT improved at idle and 1500 rpm, now that the LPG vac leak is fixed?

If you'd asked me this yesterday when I checked I would have said no change (exact same figures) but when I re-checked the data today the petrol LTFT figures were actually much worse!

Gasoline/petrol
Idle: STFT +/- 5%, LTFT +22.66%
1500rpm: STFT +/- 5%, LTFT +22.66%
2500rpm: STFT 0%, LTFT +22.66%

LPG
Idle: STFT -33.59%, LTFT -33.59%
1500rpm: STFT -33.59%, LTFT -33.59%
2500rpm: STFT +32%, LTFT +32%
* Fuel trim turns positive above 2000rpm

It is my understanding that OBD-II does a fault count on something like a too lean situation, and does not throw the error code and CIL until it has seen the error something like 200 times during a cycle period. Not sure what the cycle period is, but the software should be able to show you pending codes where there is an error count above zero heading for the magic 200 times count.

With mine I have found it will throw the P0171 "pending code" once before confirming it as an actual P0171 fault code the next time it happens. It doesn't set the Check Engine Light off but maybe I just haven't left it long enough before clearing.

 

[Ecomike]

Try this. Disconnect the battery to clear the memory. The start and run it with gasoline only. See if the LTFT is still +22% at start up, and after running at various rpms for a while.

I am thinking the high LTFT for petrol may be an historical issue from switching from LPG, back and forth.

I think the latest data confirms no exhaust leak near the O2 sensor, and no signal loss from the O2 sensor to the PCM since you get the same data at both locations and can find a fuel (pertol) and rpm where the O2 sensor works.

IIRC, the data has changed considerably since the repairs to the LPG control hardware. If the LPG is working as designed then you have a permanent issue, by design to battle, other wise there is still something wrong with LPG control system?

You might try going back to the LPG guys, show them your OBD-II live data, and team up on solving the problem?

Certainly looks like the LPG is now running way too rich, and is out of control at idle and 1500 rpm!

Are these LPG guys using post Cat converter data for setting idle LPG flow rates? Perhaps they should be using your live OBD-II O2 sensor data?

I am still scraratching my head over the rich signal for petrol at 2500 rpm with STFT of zero!

The PCM O2 sensor is saying the LPG is running way to rich at idle and still too rich at 1500 rpm, which sounds like they still don't have the LPG idle flow rate fixed?

 

[gradon]

I might agree with Mike that the +22% adding fuel in gas mode could be a response from how much it yanked out when in LPG mode. I would further that the moment you switch from lpg to gas, the pcm realizes those -33%s aren't cutting it, trigger the 0171 lean, and then basically swing 50%. I wonder if your injectors are really turning off in lpg mode(unplug them while in lpg mode and see if fts change; cel might set off when pcm sees there's no load)? I say it would be a good idea to monitor the trims from a fresh reset first and foremost. Figure out what your data tells you: all is good, still too lean, etc. +22% is a major addition of fuel, so you might have a leak pre o2 sensor, bad o2 sensor/sensor wiring, clogged injectors, the wrong injectors, clogged fuel filter, low fuel pressure. Once you see the trend, disco power and try the lpg from scratch and as Mike suggested, use the obd2 data to tune it. Oh yeah, find an lpg conversion forum so you can get info from people with first hand experience.
p.s. The reason those STFTs are 0%@2500 is cause he already told us those .7-8V values at that rpm so it was in OL(note how LTFTs affect OL), otherwise it would be the oscillating CL(I wish my stroker went to OL@2500 and ignored the 14.7:1CL).

 

[XJSteven]

Thanks for the suggestions, I'll give these ideas a go later in the week. I'll also finish summarizing the last 14 pages and make an updated post here as well as on an LPG forum.

Are these LPG guys using post Cat converter data for setting idle LPG flow rates? Perhaps they should be using your live OBD-II O2 sensor data?

He definitely had some sort of OBD-II scanner connected to the PCM because it was still hooked up when we went for a test drive. I don't know if he was looking at the fuel trims though? He also had some sort of DOS computer which he used to program the LPG ECU, apparently to improve performance & economy (all of which will be reset if I disconnect the battery).

Found an interesting link on LPG systems here:

http://www.cherokeeforum.com/f5/lpg-dual-fuel-systems-65184/

In regards to the P0171 code being set at engine start up this may not be the case. I drove around for 20-30 mins today watching the Live Data with NO FAULT Codes showing and stopped the engine (still with no fault codes showing on the laptop) and left the ignition in the ON position. I then re-connected the OBD-II software because the connection was lost when I turned off the engine. Once the OBD-II software restarted/rescanned the fault suddenly appeared (without starting the motor back up) so it looks like the fault is only detected by the OBD-II software when the connection to the PCM is reset - not when it actually occurs.

 

[gradon]

Have you used someone else's hand-held obd2 scanner yet? Next time you have a cel, go to autozone or advance auto and use one of theirs just to verify it's not your software trippin.

 

[XJSteven]

Have you used someone else's hand-held obd2 scanner yet? Next time you have a cel, go to autozone or advance auto and use one of theirs just to verify it's not your software trippin.

I've only had the laptop software a few weeks but before that I was using my own hand held scanner to detect and clear the same P0171 code.