[Red_Z]

I've got a K20Z3 in a road racing car and looking for some help troubleshooting some issues that came up this past weekend. The engine ECU is a Motec M1 and I have full logging so I can see a good bit of data as to what is going on. On track, I noticed the engine go from performing very strong at the beginning of the weekend to being noticeably down on power by the final session. The engine just didn't feel as strong at the top-end, like dragging an anchor or fighting a huge headwind vs. early in the weekend. Other than that, there weren't really any other big symptoms - engine started and idled fine, no weird noises, etc.

In looking through the data, I think I've found the symptoms but looking for help as to what is root cause? See attached snapshot of the logging data showing areas where the VTC Phase tracked pretty well vs. the Aim and then other times where it swings wildly and drops-out completely with gaps. The only correlations that I can see here are:

• Seems to drop-out whenever RPM > ~7,000 rpm
• Oil pressure doesn't look too bad to me in general, but the drop-outs seem to correlate almost exactly with GOOD = >~82 psi and BAD = <80 psi. I wouldn't think it would be that sensitive, but it really does seem to hold true across all of the data. Data from earlier sessions where it performed better and had perfect VTC Phase tracking the oil pressure rarely dropped below 86psi.

Any thoughts as to what would cause this? At first, I figured the data drop-outs had to be a bad sensor or wires. But with the oil pressure correlation, I'm wondering if this is related to the VTC Solenoid/Screen and/or tensioner? I plan to pull the motor and inspect all of these. Would love any thoughts on what to look for and what is the most likely cause?

Thanks in advance.
Dave

K20Z3
AT Power ITBs
Drag Cartel 3.2 intake/2.2 exhaust
12.4:1 CR

 

[KBuilt]

This looks like a good one for @LotusElise can you see what the vtc angles were when the power was falling off ?

 

[KBuilt]

Checking the filter and screen for the vtc is a good step to take, were you running an air filter for the track weekend ?

 

[Bjorn]

Do you have vtc solenoid duty cycle data? Would be interesting to compare duty cycle to vtc angle changes. It certainly seems like it could be an oil pressure/ flow issue. It's possible there are other contributing factors too. Where is your oil pressure sensor located?

Could electrical noise be interfering under load but picking up correctly when you're off the throttle?

Are you plugs and coils all healthy?

Are all you're engine grounds present with good connections? Maybe a bad ground has vibrations under load at high rpm that are causing poor sensor readings

 

[LotusElise]

That's an very interesting issue @Red_Z. I also see a dependency of engine speed and the issue. This could, like @Bjorn mentioned, indicate an EMC root cause. Background would be e.g. spark plugs with an worn resistor, causing a high EMC by air and harness into the ECU. To investigate this just the spark plug ohmic resistance have to be measured. Voltmeter, spark plug, measure, ready to read the result, done . Otherwise, VTC sensor data points are missed in the log, which could lead to missing grounding (temporarily), like @Bjorn mentioned.

What makes me thinking to another root cause is following:

• the ramps up and down of VTC is achievable for the stock system,
• EMC noise would have a frequency correlation of engine speed, can't see this in the VTC movement.
• oil pressure-engine speed correlation brakes off when VTC is actuating crazily, indicates the oil pressure system sees temporarily added consumers/filters/...
• the driver feedback of lower power is a good indicator for real VTC movement

The engine just didn't feel as strong at the top-end, like dragging an anchor or fighting a huge headwind vs. early in the weekend. Other than that, there weren't really any other big symptoms - engine started and idled fine, no weird noises, etc.

What I would like to see is lambda = indicator for the VTC movement, ign. timing = indicator for VTC movement and battery voltage. I have an idea, which need to be fed more with data.

BTW, there are 50° VTC peaks. Don't you limit VTC to around 50° with the DC 3.2 intake cam? Does the piston accommodate this?

 

[Red_Z]

What makes me thinking to another root cause is following:

• the ramps up and down of VTC is achievable for the stock system,
• EMC noise would have a frequency correlation of engine speed, can't see this in the VTC movement.
• oil pressure-engine speed correlation brakes off when VTC is actuating crazily, indicates the oil pressure system sees temporarily added consumers/filters/...
• the driver feedback of lower power is a good indicator for real VTC movement

What I would like to see is lambda = indicator for the VTC movement, ign. timing = indicator for VTC movement and battery voltage. I have an idea, which need to be fed more with data.

BTW, there are 50° VTC peaks. Don't you limit VTC to around 50° with the DC 3.2 intake cam? Does the piston accommodate this?

Thank you all for the quick feedback. This is all great stuff for me to look at. I am going to pull the engine and will double check all grounds to check the EMC grounding possibility. I'm leaning towards @LotusElise thinking it is some other root cause.

To answer the question on coils & plugs - these are all brand new with the entire package having less than 400 miles or so. Sounds easy enough to Ohm out the plugs just to be sure. What resistance would I be looking for on those?

I have all of the additional data requested around duty cycle, lambda, etc. Looks like the full Motec logging files are too large to post, but I can send directly if that is of interest to look through?

Here are some screenshots to try to show the additional data:
1) VTC M1 logger: The same M1 logger file with Lambda, Voltage, etc added. You can see same segments of good following vs. bad
2) The same data but logged through the Motec C127 data logger. This is from the same session (Sunday) as the M1 file where things weren't working well. For some reason the CAM position comes over slightly differently (is called CAM Position instead of VTC Phase) and shows up differently but I think represents the same thing.
3) Motec C127 data from Friday session where everything worked perfectly

@LotusElise I think the 50 deg VTC peaks are hopefully not real. The Aim values are not set to 50 deg and under normal traces, it never goes there. So I'm hoping this isn't physically what is happening.

 

[Bjorn]

I think you're getting a sensor reading error. The cam position error when it's faulting is getting up to 1500 degrees. It shouldn't be able to fault more than 50 degrees if it's a mechanical/ oil pressure issue.

 

[Red_Z]

I think you're getting a sensor reading error. The cam position error when it's faulting is getting up to 1500 degrees. It shouldn't be able to fault more than 50 degrees if it's a mechanical/ oil pressure issue.

Thanks @Bjorn . That's what I was suspecting/hoping that was all it was. So do you think that is purely sensor and/or wiring to the sensor vs. something internal giving it weird signals somehow? Can't think what that might be but maybe trash in cover somehow throwing it off? Seems too consistent at specific times though to be random trash...

I plan to pull the valve cover off and inspect just to be safe but wondering if that would even cause the crazy numbers like 1500, etc.

 

[LotusElise]

What resistance would I be looking for on those?

Around 5 kOhm they should have new. With increasing electrode gap and miles on the plugs the resistor wears out. Some tend to decrease in ohmic resistance, some increase. Depends on the material is used to create this resistance and what it is wear mechanism.

but I can send directly if that is of interest to look through?

I would be interested, but I need logs and calibration. I will PM you.

I think you're getting a sensor reading error.

For that there is a higher likeliness. Error calculation of it shows the sensor feedback of the VTC angle is completely wrong. So there is a likeliness that e.g. the VTC angle sensor has a not fully brocken wire which get loosen beyond 7000 rpm because of the vibrations.

Further observations:

• Ignition timing wasn't corresponding to VTC, maybe it was not tuned for every single VTC step.
• Battery voltage is untypically low. Did you alter the voltage aim for the alternator control? Normally around 14.2 V are on the list there, but here the assumed average is below or around 13 V.
• Lambda probe signal was following the VTC signal. Seems not a battery voltage driven thing as this was curve-wise of the voltage the same before and after. So we must assume there are real VE changes, which relates to real changes in VTC. We see 5 to 8 changes per second, which means roughly 0.20-0.13 s peak to peak, which would be realistic for response of a AEM X-Series or a MOTEC PLM or maybe an Innovate, but not that likely for the stock stuff. Lambda was flat before the issue came up, as far as the huge lambda range and bad resolution may show it.

So my conclusion would be, yes, there is an issue of the VTC sensor (the one on the intake cam side), maybe a not fully broken wire of it or a too big distance (loosen bolt of the sensor bracket), but also the VTC, the VTC duty cycle, the oil pressure, the lambda probe and the driver experience tells us the VTC was moving as it may shown in the logs. I would definitely check following:

1. starting with to check the VTC sensor wiring, form pin to pin, sometimes brocken wires are hard to find as conditions like heat, vibration can't be simulated during an ohmic resistance measurement of the wire and visual inspection. Therefore I would check the wiring of it very carefully, noddling, pulling and stress it and test of course the sensor voltage response
2. Check the alignment of the cams and TC
3. Inspect the visually the pistons on intake side and the valve stems. Maybe you have an endoscope? VTC movement was very likely not aligning the aimed VTC angle, there is an not known likeliness of PtV-contact. You may bypass pulling the head by an deep dive leakdown test. But I highly recommend to investigate the mechanical healthiness of your engine.
4. We definitely didn't see the full extension of the lambda deviation. A visual liner and oil quality check is a sort of safety check. I assume the calibration wasn't tuned for all VTC angles in all operation fields, so if VE deviates, because VTC didn't follow the injector don't know an difference, so fueling went way way overrich in high speed sections. Liners don't like this. Depending on the milage with the issue existing, this could have an impact on the hone structure. An chemical oil analysis or simple boil off mass check can indicate this. A visual inspection of the liner would be better.
5. I would also check the VTC spooling valve wiring and valve itself, it likely didn't cause it, but it was the long arm of the failure and may suffered on that
6. I would also check the oil filter paper and its particle, just to see if any mechanically went wrong during that high frequency VTC changes
7. I would definitely care about the voltage of the battery, having 13 V in high speed sections is acceptable, 12.5 V is low. Do you have huge electric power consumers like EWP? What is your battery capacity?

I am kind of better safe then sorry attitude, went with this perfectly fine as I found often up coming stress or damage for the engine before it happened. A few times I was to much on the safe side and the effort of investigation was too much. But I didn't lost not a single engine up to now throughout my OEM time nor my tuner time. I hope this explains the longer list of inspection tasks 😉.

 

[Red_Z]

Around 5 kOhm they should have new. With increasing electrode gap and miles on the plugs the resistor wears out. Some tend to decrease in ohmic resistance, some increase. Depends on the material is used to create this resistance and what it is wear mechanism.

I would be interested, but I need logs and calibration. I will PM you.

For that there is a higher likeliness. Error calculation of it shows the sensor feedback of the VTC angle is completely wrong. So there is a likeliness that e.g. the VTC angle sensor has a not fully brocken wire which get loosen beyond 7000 rpm because of the vibrations.

Further observations:

• Ignition timing wasn't corresponding to VTC, maybe it was not tuned for every single VTC step.
• Battery voltage is untypically low. Did you alter the voltage aim for the alternator control? Normally around 14.2 V are on the list there, but here the assumed average is below or around 13 V.
• Lambda probe signal was following the VTC signal. Seems not a battery voltage driven thing as this was curve-wise of the voltage the same before and after. So we must assume there are real VE changes, which relates to real changes in VTC. We see 5 to 8 changes per second, which means roughly 0.20-0.13 s peak to peak, which would be realistic for response of a AEM X-Series or a MOTEC PLM or maybe an Innovate, but not that likely for the stock stuff. Lambda was flat before the issue came up, as far as the huge lambda range and bad resolution may show it.

So my conclusion would be, yes, there is an issue of the VTC sensor (the one on the intake cam side), maybe a not fully broken wire of it or a too big distance (loosen bolt of the sensor bracket), but also the VTC, the VTC duty cycle, the oil pressure, the lambda probe and the driver experience tells us the VTC was moving as it may shown in the logs. I would definitely check following:

1. starting with to check the VTC sensor wiring, form pin to pin, sometimes brocken wires are hard to find as conditions like heat, vibration can't be simulated during an ohmic resistance measurement of the wire and visual inspection. Therefore I would check the wiring of it very carefully, noddling, pulling and stress it and test of course the sensor voltage response
2. Check the alignment of the cams and TC
3. Inspect the visually the pistons on intake side and the valve stems. Maybe you have an endoscope? VTC movement was very likely not aligning the aimed VTC angle, there is an not known likeliness of PtV-contact. You may bypass pulling the head by an deep dive leakdown test. But I highly recommend to investigate the mechanical healthiness of your engine.
4. We definitely didn't see the full extension of the lambda deviation. A visual liner and oil quality check is a sort of safety check. I assume the calibration wasn't tuned for all VTC angles in all operation fields, so if VE deviates, because VTC didn't follow the injector don't know an difference, so fueling went way way overrich in high speed sections. Liners don't like this. Depending on the milage with the issue existing, this could have an impact on the hone structure. An chemical oil analysis or simple boil off mass check can indicate this. A visual inspection of the liner would be better.
5. I would also check the VTC spooling valve wiring and valve itself, it likely didn't cause it, but it was the long arm of the failure and may suffered on that
6. I would also check the oil filter paper and its particle, just to see if any mechanically went wrong during that high frequency VTC changes
7. I would definitely care about the voltage of the battery, having 13 V in high speed sections is acceptable, 12.5 V is low. Do you have huge electric power consumers like EWP? What is your battery capacity?

I am kind of better safe then sorry attitude, went with this perfectly fine as I found often up coming stress or damage for the engine before it happened. A few times I was to much on the safe side and the effort of investigation was too much. But I didn't lost not a single engine up to now throughout my OEM time nor my tuner time. I hope this explains the longer list of inspection tasks 😉.

Thank you Markus! This seems like a very thorough list of potential culprits and things to check out. I am with you on better safe than sorry on this stuff. I think I have access to an endoscope that can hopefully let me inspect the motor when I have the plugs out.

I also noticed that battery fluctuation seeming to be much lower than I would have expected. We run a very light weight, minimalist Li-Ion battery but the alternator at least in the pits seemed to be charging well and keeping Voltage around 13.5V+. I did just find an interesting thing on the voltage, which I'm not sure exactly how to interpret...

I am logging 2 separate Voltage channels: Bat Volts Dash and Bat Volts ECU. The Bat Volts Dash is pretty solid & stable at about 13.5V all the time in all sessions - it never moves very much. The Bat Volts ECU, however, which is what I've shown in the screenshots moves around and bounces between 12.7V - 13.6V and as we've noted seems to get down close to 12.5V quite a bit. So the ECU is getting a lower Voltage Potential for some reason. I'm assuming this isn't normal? Any ideas what to investigate there? Would a bad connection/ground between ECU and battery potentially cause this? The ECU is a Motec M130.

I did drain the oil already and the oil looked really clean. Hardly any metallic shavings in there, etc. I haven't pulled the oil filters yet - we run the stock oil filter as well as an external canister replaceable cartridge filter. I will do that this weekend and see if anything looks off there.

I sent you the data files separately and would love your thoughts on anything else you see there.

 

[LotusElise]

I'm assuming this isn't normal? Any ideas what to investigate there? Would a bad connection/ground between ECU and battery potentially cause this? The ECU is a Motec M130.

Mmmhhh, if the source itself has a stable value of 13.5 V and the ECU sees much wider band there seems to be a consumer or a dancing reference like e.g. an lose ground connection. The biggest current drawers are the injector drivers in the ECU, but there is no correlation between duty cycle, engine speed and the battery fluctuation on the ECU Battery Voltage. What you can investigate is the ECU grounding and the 12 V supply wires to the ECU.

I sent you the data files separately and would love your thoughts on anything else you see there.

Thanks, I skimmed through the two logs and the calibration. There are no indicators in the tune itself causing any related issues or instabilities of the VTC nore the battery, which could avoided and run the engine properly anyway. I apprehend you still have to go through the list anyway.

 

[Red_Z]

Mmmhhh, if the source itself has a stable value of 13.5 V and the ECU sees much wider band there seems to be a consumer or a dancing reference like e.g. an lose ground connection. The biggest current drawers are the injector drivers in the ECU, but there is no correlation between duty cycle, engine speed and the battery fluctuation on the ECU Battery Voltage. What you can investigate is the ECU grounding and the 12 V supply wires to the ECU.

Thanks, I skimmed through the two logs and the calibration. There are no indicators in the tune itself causing any related issues or instabilities of the VTC nore the battery, which could avoided and run the engine properly anyway. I apprehend you still have to go through the list anyway.

Thanks @LotusElise. I will check power & ground to the ECU to see if I can find anything there.

I will replace the CAM position sensor and check all wiring to it. I will also run through all of the other investigations that you listed out just to make sure I don't see anything else bad.

Thanks for the help!
Dave

 

[LotusElise]

...I will check power & ground to the ECU to see if I can find anything there.
...I will replace the CAM position sensor and check all wiring to it. I will also run through all of the other investigations that you listed out just to make sure I don't see anything else bad.

Very good. Let us know the results. Maybe we recognize something more detailed and can alter or reduce the workdown list.

 

[KevinLauria]

Hello There! I don´t have any experience with that but they can run in open loop? to descart the CAM position sensor. If you are sure of the pistons don´t crash with the valves.

 

[Bjorn]

Hello There! I don´t have any experience with that but they can run in open loop? to descart the CAM position sensor. If you are sure of the pistons don´t crash with the valves.

No. a fixed duty cycle won't hold a fixed cam position. The closed loop needs to be in place to achieve the cam phase target.

You could theoretically have an open or closed valve to achieve 2 states of cam position. Full advance and full retard. A very poor compromise on the full adjustability of the factory system but better than nothing.