Check Engine Light with code P0170 and P173 on Montero/Pajero

Question: Could these error codes be caused by dirty fuel injectors that disturb the spray pattern. Or could it be caused by an intermittently sticking automatic hydraulic tappet (it clatters from time to time even when hot).

I recently purchased a Mitsubishi Pajero in New Zealand. It had been imported from Japan in 2010 and seems to be exactly the same as a full Montero which I also own in the USA (I flit back and forth between NZ and USA). It has the label Excel GDI which stands for Gasoline Direct Injection. A V6 3.5 liter petrol engine. It has done 165,000 Km or about 100,000 miles.

The check engine light goes on and off randomly at intervals of about 100 miles. The engine runs great. Fantastic even under heavy loads. Fuel consumption is good at 6.8 - 7 liters per 100 Km (34 MPG ?). I have had it checked by two workshops, one of which specializes in Mitsi. They suggested an injector and injector pump overhaul. They did the following:

Smoke tested intake manifold -no leaks.
Checked fuel pressure OK
O2 sensors OK
Ignition OK
Checked EGR operation - OK.
Replaced spark plugs.
Flushed and changed oil and oil filter twice within 600 miles.
Replaced air filter.
Recommend customer have fuel injectors removed, cleaned tested and rebuilt.

However I have also noticed the tappet noise. Not very often but it may be enough to trigger the error occasionally if a valve failing to open fully causes an altered mixture that cannot be compensated by the computer.

Does the injector theory sound likely, Should I consider any other causes?

I do not have an OBD2 code scanner and am thinking of buying one in the US so that I can check and reset the codes. Not too expensive. Any suggestions?

The on-board computer and radio have all the buttons in Japanese and the manual is Japanese, but that is another problem!

Thanks!

 

How did you check your 02 sensors evan, Plz pm me.

 

I bought OBD software for my iPhone and MacBook Pro to test my Mitsi and it was cheap at $30 and comprehensive. Excellent. I will create a separate thread about it as it may be of general interest to others.

This uses iPhone app software called AutoDoctor ($US10 from iTunes) and a WiFi transmitter ($US20 from eBay) that plugs straight into the standard OBD2 socket. These sockets are standard on all vehicles as a requirement for emissions testing. Unplug the WiFi transmitter when you are not using it.

The app tells me the error code is always P0173, and it allows me to clear all errors on the "on board Diagnostic computer or OBD". The error only comes back on if I let it idle for at least 5 minutes. After running about 100 miles on the open road the check engine light will go off again although the error code is still recorded by the onboard computer. The iPhone app also gives the comment that the oxygen sensors and catalytic converter may be faulty. But it does not give me a specific error code for that and does not mention which bank so I don't know how it can tell. I thought a faulty oxygen sensor would cause a specific error code. The app does include a database of error codes.

The app gives a lot of data about what the computer was recording when the error occurred (a snapshot or freeze frame). It shows the voltage on the oxygen sensor and the adjustment the computer is making to the long-term and short-term mixture adjustments for each bank (it is a V6 so there are two banks of three cylinders). Bank 1 shows that the short term fuel trim is about -5% which means that the computer had to decrease the fuel injected per stroke by 5% to get the oxygen sensor to give the required signal.

But on bank 2 (and I don't know whether that is the left bank of the river Seine or the right), we find about -20% to -25% short-term mixture adjustment. That means the computer thinks the mixture is too rich and is trying to make it leaner, but when it hits about -25% for a long enough period of time it decides that something is wrong and triggers the check engine light (CEL) to come on and sets error code P0173 "Fuel Trim bank 2".

The long-term fuel trim is rock steady at -10% for bank 2. Freeze frame at the time of the error shows fuel systems one and two are both in "closed loop control" modes. RPM=691 at 0 km/hr and 23% engine load, temp 91C. But after running at higher revs for a while it finds the mixture is OK and cancels the CEL.

Closed loop means it is continuously adjusting the mixture depending on the oxygen reading in a negative feedback loop. Apparently at very high engine speeds or loads the computers prefer to switch to open loop control which I assume ignores the oxygen sensors.

On the display showing -25% fuel trim you can click on it while the engine is idling and it will show a graph showing how the fuel trim varies over time. In fact you can graph any of the parameters including the voltage coming from the oxygen sensor.

What I found it that the oxygen sensor voltage is oscillating up and down like a sine wave from 0.1 to 0.85 volts with a cycle length of about 2 to 5 seconds. This looks quite erratic to me and I suspect this behavior may be the cause of my problem. However, it surprises me to find that the oxygen sensors in both banks are doing the same thing (oscillating) but the error is occurring only in bank 2. The short term fuel trim on bank 2 (O2S) only fluctuates from -21 to -23% most of the time. This is relatively steady. There is a similar variation from -7% to -8.6% on bank 1 (O2S) short term fuel trim.

At one time the service department reported both P0173 (bank 2 short term fuel trim error) and P0170 (bank 1 short tern fuel trim) errors. This made me think it must be something affecting both banks, most likely air flow sensor (MAF mass air flow). But the app tells me that the air flow is rock steady at about 5 grams per second at idling speed and varying appropriately with revs when accelerating. So I think it is behaving normally. Furthermore I have never seen the P0170 error again.

So my thinking is that the oxygen sensors may need replacing.
Has anyone else with a Mitsubishi observed the oxygen sensor voltage fluctuating like this?
Is it normal?
Has someone installed the wrong kind of O2 sensors?
Or are they just old and dirty.
Or is this normal behavior?

Another possibility is that a sticking automatic tappet may be preventing one of the valves from opening fully and this could affect the mixture. It is a direct gasoline injection engine which means the petrol is injected directly into the cylinder and the quantity injected is calculated by the computer. If an inlet valve failing to open fully causes insufficient air to enter the cylinder it would result in the mixture being too rich. I do have at least one tappet that is very noisy. It sounds as though it is coming from the left bank cylinder head (is this one or two?). It occurs normally when you first start the engine and the oil pressure has not yet built up. But in my case it continues when the engine is hot and idling and is less obvious at higher RPM. It can stop and start randomly which is consistent with valve lifter problems which are caused by a sticky non-return valve in the hydraulic system that automatically adjusts the tappets. I have had several oil changes with special additives that are supposed to fix this problem but it hasn't worked (One ProMod? And another from Repco). I have also used two injector cleaner additives in the fuel.

I imagine the hydraulic system that controls the end-play in the cam shafts would not make such a systematic classic tappet noise, although I have read that this is a common cause for valve train noise in these vehicles. The cure for that problem is to rev the engine at 3000 RPM for 15 seconds, drop suddenly to idling for 15 seconds and repeat up to 30 times. This helps to fill the hydraulic cylinder with oil after an engine has been unused for a while.

I took the car to Wellington Mitsubishi in New Zealand and they said it would be $60 per valve for parts and there are 4 valves in each of the 6 cylinders so 24 valve systems to replace. With labour it would be $3,000 to overhaul the valves and I decided against it. But a non-Mitsubishi mechanic Phil Berry in Paeroa told me he could get OEM parts much cheaper including oxygen sensors.

This raises the question as to what brand of oxygen sensor should I use? I understand there are various types including one by Bosch. Wellington Mitsubishi checked with their parts importer for New Zealand and they were told there are no sensors in New Zealand and it would take 30 days to get them from Japan. So that idea was cancelled. They did cut me two spare keys and I was a bit shocked at the $550 price tag! But they did paint my tyres black for free! I plan to go back to Phil in Paeroa, and unless I get alternative advice from the forum I will probably replace the oxygen sensors first.

Any advice on how to proceed would be most welcome!

 

This link says that these regular swings in sensor voltage from 0.1 to 1 volt are normal and it shows example graphs:

Help with o2 sensor voltage - S-10 Forum

There is a very detailed article about O2 sensor technology in Wikipedia but it does not discuss the voltage fluctuations seen in these graphs. However it does discuss the reasons for O2 sensor failure. Normal life is 100,000 miles (160,000 km) but it can drop to 15,000 miles. See:

https://en.wikipedia.org/wiki/Oxygen_sensor

 

The following link shows graphs of voltage v. time with clear sine waves similar to those I am seeing. Apparently this is the feedback loop going through normal cycles of slightly rich to slightly lean and this does not indicate faulty O2 sensors. They say:

First of all, engine needs to be warm. Second, RPM should be revved up to about 2500 to get a nice exhaust stream over the sensors.

Voltage should constantly fluctuate between .125v and .875v. This up and down should happen a few times every second. Average should be about .450v.

o2 sensor readings - LS1TECH

 

This link gives general advise about the mixture error codes. It says these errors are rarely caused by the oxygen sensors. It mentions they can be caused by cam problems. I assume this may include sticking hydraulic tappets: I may have to get the tappets fixed.

Welcome to Leagend: Easier OBD II Diagnostics, Smarter Gauge