Stalling and poorly running Duratec

[frizille]

Hi all, just a quick update. I've been able to figure out how to replicate the problem and when it triggers. The car starts and idles fine until 60C when the ECU listens to the lambda sensor and goes into closed loop mode. Before 60C it's running open loop as the lambda timer hasn't counted down and the engine isn't up to temp. The moment 60C is reached and the lambda activates the motor starts surging and will stall a minute or so later. I've swapped the TPS sensor and lambda sensor for new units as well as verify all the grounding cables.

 

Question for the group - what happens at 60C other than the lambda sensor being activated?

[JohnCh]

There should be a coolant table that adjusts ignition and fueling based on coolant temp, but I'm not sure it's possible for the table values to change on their own. Are you able to log the AFR once you hit 60C? If so, what does that show? Are the values showing extremely rich or lean, or jumping all over the place? Does your ECU have a native driver for the WBO2, or are you running a standalone system like an Innovate? It's possible that the controller is fried, or it could simply be a wiring issue. Have you checked resistance of the WBO2 wires to confirm there isn't a partial break inside the insulation, and also confirm the voltage reading as Tom suggested? Lastly, when you replaced the WBO2, did you calibrate it?

 

-John

[Joe R]

Hey Dan, your o2 sensor (lambda) and your ECU are not in sync. Check with the manufacturer to get the procedure for "free air calibration" of the new sensor first. Next step confirm ECU procedure for getting the o2 synced up with the ECU.

[frizille]

John & Joe -

 

There is no AFR logging unfortunately. I'm using the Easimap 6 software from SBD Motorsports and stock Caterham ECU, the MBE 992. The lambda/O2 sensor is a Bosch part, 0258003229 which is labeled as a Rover part MHK10006 (which is the same as found on the Elise S1 for a bit of trivia). The US equivalent is 0258005727 (or part num 15727) which I got by calling Bosch directly and talking to their technical support folks. It is a narrow band consumer-grade O2 sensor rather than a more feature rich wide band sensor that would generally be used as part of a tuned engine. There isn't a calibration procedure that I can find anywhere - it should just be a plug and play sensor supplying a voltage to the ECU. I've got the ECU unplugged now to reset it in case that does anything. According to the ECU, the lambda/O2 sensor voltage stays pretty steady until 60C. The battery also starts to fluctuate wildly at the same time. It's like something is starting at 60C and I have no idea what it might be. Graph below (upper line is battery voltage ranging from 14.1 with dips to 13.0, lower is oxygen voltage going from about 1.1V to .22V)

 

My next idea is to do a leak down test this afternoon and see if that yields any further hints, but the motor seems fine at idle right until 60C.

[DaveStruve]

My initial thought from looking at this is that the battery and o2 are a result of the issue and not the cause. The engine hits 60 degrees and it does its stalling thing, alternator starts making less power as the cylinders shut off and theres a bunch of fuel being dumped into the system. Maybe disconnect the O2 and see what happens, just to try and eliminate it from the equation.

[frizille]

Does the water pump run all the time or is it ECU controlled in some way? My understanding is that the pump functions as an impeller and would spin whenever the motor is spinning, i.e. powered by the serpentine belt. I just want to make sure that is accurate.

[frizille]

My initial thought from looking at this is that the battery and o2 are a result of the issue and not the cause. The engine hits 60 degrees and it does its stalling thing, alternator starts making less power as the cylinders shut off and theres a bunch of fuel being dumped into the system. Maybe disconnect the O2 and see what happens, just to try and eliminate it from the equation.

 

Could the alternator be going bad?

[DaveStruve]

If its pulley driven then yes. The thermostat opens and changes the flow direction through the motor instead of around it.

 

As for the alternator your hitting 14v so it generating.

[frizille]

I just tried running it with the O2 sensor unplugged and it ran longer, but ultimately still did its stall thingy. I unplugged the ECU for ~5 min to reset it, unplugged O2, then plugged the ECU back in. It ran up to about 69C then started to pulse, I could see the RPM going from 800-1500 and the ignition timing jumping from about -8 to 14 degrees. It made it past the normal 60C mark by about 10 degrees, but ultimately started pulsing. The battery did not dip like it did before - it stayed between 13.9 and 14.3 the entire time. Graph below.

[Joe R]

Check your connections to your rad fan & temp sensor. Check ground to rad fan carefully.

[frizille]

Ok, a productive afternoon in the garage... kinda.

 

- On a good note, the leak down and compression tests both look ok. The results of the leak down test were all cylinders holding 90 psi without any loss. During compression tests cylinder 1 and 2 were in the 205-210 range while cylinder 3 and 4 were in the 200-205 range.

 

- I did a test with the radiator fan unplugged and it started the normal pulsing on queue.

 

- I ran a test with the instrument panel temp sensor unplugged; same pulsing. I didn't think it would matter, but one more thing to rule out.

 

- I ran a test when it was pulsing where I disconnected each fuel injector and coil pack, each one at a time. When disconnecting any one of the fuel injectors, the pulsing stops but the motor obviously runs lumpy since it is down one cylinder. Disconnecting any one of the coil plugs has the same effect.

 

- Alternator looks ok; as Dave mentioned I'm hitting 14V so it is generating.

 

- I tried another TPS sensor; same pulsing result.

 

- I tried another lambda sensor; same pulsing result.

 

- I tried unplugging the O2 sensor; same pulsing but started 10C degrees later.

 

- I tried resetting the ECU by unplugging it for 5+ minutes (per Caterham's directions).

 

Theories I haven't tried yet:

 

- There's something wrong with the ECU map.

- Cam sensor problem.

- Crank sensor problem.

- Bad temp sensor

 

Anything else I didn't think of?

[papak]

The sensors are a relatively cheap fix. On the other hand, they are just magnets and don't wear out. If the crankshaft sensor is loose, you will need the plastic alignment guide that comes with it to reset it.

[frizille]

Crank sensor was tight and aligned with the alignment tool. I think the crank and cam sensors are ok - if they weren't I wouldn't expect the motor to idle or run at all since the ECU wouldn't know where any of the internals were at in the cycle. The "bad missing tooth count" in the ECU is not increasing when the motor is on, which per SBD is as it should be (http://www.sbdev.co.uk/FAQs/General_running.html).

 

That leaves the ECU and temp sensor. The temp sensor seems to be providing good data back to the ECU and CAN BUS, I see that in the Easimap software, and it matches pretty close with what I'm seeing on the temp gauge in the cluster (they are fed through different signals and sensors).

 

That just leaves the ECU and don't really have a way to test it. I may need to pull it and send it back to Caterham in the UK for them to look at.

 

Dan

[coffee break]

In my observations from a distance, the temp sensor input seems to be a common theme. If you could control the temp sensor environment independent of the motor it may help isolate the issue. For example, remove the sensor and plug its location. Start the motor and heat the sensor with a heat gun/blow dryer and look at the response.

Off the wall can't hurt.:banghead:

Good luck!

[frizille]

Coffee break - good suggestion. I don't have a heat gun and Google says a hair dryer only gets to about 130F or 54C. That's below my 60-70C threshold.

 

So... I ran up to the Ford dealer and picked up a replacement coolant temp sensor this afternoon. Those things are expensive! Online generic replacements were like $20, the "official Ford part" was $58. Note, I got the "official Ford part" cause I'm not introducing another variable in this mess if I can help it. If it doesn't resolve the issue at least I've got a spare. The question is, if I pull out the sensor is coolant going to come surging out...

 

I also spent a bunch of time staring at the ITB setup from SBD. Comes with a replacement 9A4 ECU. One take away from this process, other than learning a whole bunch about engines in general, is that I don't much like not having access to see what's going on in the ECU. In my Jeep that's fine - it's a Jeep - but for the Caterham where my nearest support is 5,000 miles away I really want to be able to look at what the maps actually mean and what they're telling the motor to do.

 

Dan

Edited September 30, 2016 by frizille

[frizille]

I think I've stumbled onto something! I was drinking a bottle of wine this evening - trying to drown my sorrows no doubt - and searching through Blatchat when I stumbled onto a thread about TPS voltage. The dialog in the thread said you needed to be around .360-.420V - which is labeled as "Throttle Angle 1" in Easimap - to hit "Throttle Site 1" of 0.0. I distinctly remember earlier this week when I was resetting the TPS that I couldn't get it below "Throttle Site 1" of .5 and most of the time fluctuated between 1.0 and 1.5. I also noticed the TPS sensor itself has a lot of play between where it mounts to the throttle body and where the stop actually is on the sensor when you twist it. The voltage reading from the TPS during my tests ("Throttle Angle 1") generally stay around 1.2V which is almost 3x (!!!) as high as it should be. I took the screws that attach the TPS to the throttle body out, then rotated the TPS by hand until Easimap read .42V. The TPS is literally as far as I can turn it at .42V, in fact, I broke one TPS sensor already twisting too hard... thankfully I had a spare on the shelf (they're $26 when you break them by the way, part num 988F-9B989-BB from Motorcraft via Amazon). At .42V the "Throttle Site" now reads 0.0 as it should.

 

So what does that mean to the engine? Well, I can't see the map values since Caterham has the ECU locked, but based on screenshots I found of the different load sites, I'm about 25% higher than where it should be. That might explain why my idle has always been a bit rough and why when I get into closed loop things go off the rails - it thinks the throttle body is open when it's really closed! The ECU is dumping fuel in because the TPS says air is on the way, but its not, and end result is a fuel laden mess in the cylinder where it constantly tries to change ignition timing until it stalls. It's behaving like a vacuum leak, but instead of leaking air out, it just isn't getting air to start with. At least that's my theory... I'll have to test it tomorrow as I try not to run engines after dark so the neighbors don't get buggered.

 

Here's the question though - I can't leave the TPS unscrewed from the engine and I've got it twisted so far away from the mounting screws that there's no way to actually attach it (zip ties maybe? See attached pic). How do you adjust voltage on the TPS sensors? There has to be a way - either that or I've got some deformed throttle body that doesn't line up properly.

 

Dan

Edited October 1, 2016 by frizille

[vstryker]

To adjust voltage you turn the sensor until you see the desired voltage reading. So in this case, you either turn it clockwise or counter clockwise. After adjust it maybe you can screw it on. Check to see if it is pointing in the correct position though as this Caterham is opposite of yours:

 

http://media.caranddriver.com/images/09q4/303952/2008-caterham-7-superlight-r400-duratec-20-liter-inline-4-engine-photo-303960-s-1280x782.jpg

 

If that is the case, turn it around, and readjust. You might have to go into the ECU settings and change the orientation setting: either clockwise or counter. The MBE 9A4 allows full control of the tps sensor including max, min voltages

[frizille]

To adjust voltage you turn the sensor until you see the desired voltage reading. So in this case, you either turn it clockwise or counter clockwise. After adjust it maybe you can screw it on. Check to see if it is pointing in the correct position though as this Caterham is opposite of yours:

 

http://media.caranddriver.com/images/09q4/303952/2008-caterham-7-superlight-r400-duratec-20-liter-inline-4-engine-photo-303960-s-1280x782.jpg

 

If that is the case, turn it around, and readjust. You might have to go into the ECU settings and change the orientation setting: either clockwise or counter. The MBE 9A4 allows full control of the tps sensor including max, min voltages

 

Thanks vstryker, I was afraid of that. The TPS sensor won't go below 1-1.2 V when it is lined up to bolt in. I tried it in both positions, same result. The ECU I'm on is the 992 and it's locked from Caterham so I can't adjust. So... I'm stuck. Guess I'm calling Caterham Monday.

 

 

Sent from my iPad using Tapatalk

[frizille]

No dice. I tried starting it this morning with the TPS unbolted and twisted to show .42V in Easimap and it did idle MUCH better than before. In fact, it's never idled better and have always been chasing a really bad idle so I think that's sort of fixed, assuming I can figure out how to mount the sensor right. I tried it both ways but the same problem persists - the problem is in the geometry of the whole thing. However, once it got to about 70C and I started giving it throttle it was letting out a bunch of smoke from the exhaust - which could just be stuff burning off because it hasn't really run since I had that issue with a lot of fuel coming from the exhaust and it's relatively humid in the garage (comparably) - and then started pulsing. It was pulsing a little at idle too, but not very noticeable.

 

So... still at it.

[Kitcat]

Reading this thread is like reading an Agatha Christie novel: lots of suspects, false clues, red herrings, and occasional small break-throughs. I "tune in" every day to see if the suspect has finally been caught:)!