Can Lambda Error Code

[jigga009]

This is a Link G4+ Xtreme application on the latest available firmware.

I've unknowingly had the "Lamb48- APE Overvoltage" issue for a while. Only found out after poking through some old datalogs.

Device was installed per the installation guide included in the box using the suggested method that does not involve the capacitor.

Did some digging, and tech support suggested that my sensor was on its way out (or a fake). Since this was the original sensor supplied by Link in the box with the Canlambda, and with less than 20 miles of use on the sensor and the CanLambda itself, I thought it was a little unusual for a sensor to check out so soon, but hey...They suggest that I look for Bosch 17025 which is compatible with the Can Lambda.

I purchased a new original Bosch 17025 from Amazon. Once I fired up the engine, same result - "Lamb48 APE Over Voltage". The code was present as soon as I started cranking the engine. 

Talked to Tech support at Link again, and they suggested that I may have purchased a fake sensor from Amazon. 

Purchased another new Bosch 17025 sensor locally from a trusted automotive parts supply company, installed the sensor, and upon cranking the engine over (during a battery load test), I get the same error code pop up as soon as the starter motor starts turning the engine over.

I'm not sure how to proceed at this point, as the CanLambda was installed per the installation guide, and these sensors aren't exactly cheap. If I allow the engine to run, it does appear to work, but with the error code ever-present.

I tried applying some Stabilant 22A to the electrical terminals connectors of the sensor and Canlambda, but same thing occurred.

 

My plan now is to check wiring, but I'm unsure as to what exactly I am checking with my multimeter

- Since it seems as if the problem is at Pin 1 of the Canlambda->Wideband sensor interface, I'm a little unsure how to "test" this connection short of switching in new sensors (which I have done twice already now).

I'm including a copy of my ECU map, as well as a few datalogs in case they show something I'm missing.

 

ECU Map - https://www.dropbox.com/s/a3fv5f4ywarjaf4/Map 3.1 - Cranking map - ignition and fuel turned off.pclr?dl=0

Yes, ignition and injection were turned off on this map, as I was conducting a battery load test.

 

Datalog with sensor #2 purchased from Amazon - https://www.dropbox.com/s/erc4wmu532c2hz4/Startup log file Sept 6 2021.llg?dl=0

 

Datalogs with sensor #3 purchased locally from an automotive parts supplier - 

https://www.dropbox.com/s/zo8eb3lz2bmaz7r/Log 2021-09-18 12%3B57%3B23 pm - cranking test 1.llg?dl=0

https://www.dropbox.com/s/ukbi1h7wo7w9lig/Log 2021-09-18 12%3B58%3B40 pm - cranking test 2.llg?dl=0

These particular 2 logs are both engine cranking tests, as I was actually performing a battery load test at the time while the logs were taken with the sensor 3 in place, but they do show the Lambda error pop up with sensor #3 as soon as the engine starts cranking over.

Any assistance would be very much appreciated.

Thanks.

 

[Adamw]

As a quick test, try adding the capacitor or moving to a different power supply.  

[jigga009]

12 hours ago, Adamw said:

As a quick test, try adding the capacitor or moving to a different power supply.  

I’ll pick up the capacitor tomorrow on Monday. In the mean time, I took a couple more cranking logs

https://www.dropbox.com/s/eyzeh3px8rfpvcq/Log 2021-09-19 3%3B24%3B17 am.llg?dl=0

 

And 

https://www.dropbox.com/s/xlmj2cn3njd2qv3/Log 2021-09-19 3%3B20%3B07 am.llg?dl=0

 

Interesting to see that on some cranking events, no error shows, while on subsequent events, an error pops up as soon as the engine starts turning over

the battery looks to be sound, as voltage measured at the terminals drop to 11.2v during a 10 second cranking session. 

I’ll report back soon.

 

thanks!

[jigga009]

Ok, I added the capacitor, and same thing happened.. On first cranking attempt, everything seemed okay. On the second cranking attempt, the code popped up.

https://www.dropbox.com/s/hymjy5bef5dsain/Log 2021-09-20 6%3B13%3B21 pm.llg?dl=0

That's a datalog showing what happened. 

[Adamw]

Whats all the weird CAN stuff set up in your original file attached above?  The lambda error often occurs at the same time as CAN transmit errors.  You have some very odd stuff like stream 7 & 8 being set up as both receive and transmit which wont work. 

Can you unplug all the other CAN devices, turn off all the user streams and try with just the CAN lambda.  

[jigga009]

The setup of the CAN might be odd, but you helped me set it up a while back. I have in the meantime made a few alterations as to include more data, but the general architecture that you set up is as is. I'm not sure if it is something about how CAN works on the MXG 1.0 devices.

I had to set up stream 7 and 8 to receive and send in order to be able to add more parameters to be sent to the dash without the ECU complaining about "too many streams in use" or something with just around 30 parameters being sent to my AIM dash. Doing it this way, I was able to increase the number of parameters it was sending to the dash without it complaining.

The AIM dash is connected on CAN 1, and CAN 2 has my Link CAN Lambda, AEM 4channel wideband, and the CAN 2 connection to the MXG. I can disconnect the AEM from the CAN 2 bus. I can also try disconnecting the aux harness from behind the AIM in order to take the dash out of the CAN 2 bus as well.

[jigga009]

Ok,

I turned off ALL streams on CAN 1 and CAN 2 the EXCEPT Canlambda, and disconnected everything on the CAN 2 bus (which is where Canlambda lives).

I also pulled the harness from the AEM wideband which lives on CAN 2 in order to rule it out (even though it is on a relay that the ECU turns on about 15s after engine start).

I also pulled the MXG from the car so as to rule it out since it lives on CAN 1 and is also patched into CAN 2

Same result upon cranking the engine over - Lamb48 APE over voltage.

Map as tested - https://www.dropbox.com/s/g4xscuet7e56lvc/Map 3.2 - same as 3.1 but with all streams off except canlambda.pclr?dl=0

Log file - https://www.dropbox.com/s/qca595uzs64fnrr/Log 2021-09-20 - log with all streams off except canlambda.llg?dl=0

 

[Adamw]

Ok, map attached with some extra channels set up to help diagnose, and I have set up ECU logging to record some relevant parameters too. 

The Lambda will start up as soon as the ignition is on with this map. 

So can you try this:

1. Load this map, store. Clear ECU logs.  power cycle.
2. Allow to rest with ignition on for say 30sec, confirm lambda is working.
3. Crank engine for say 5 sec twice as you have done before.  
4.  Download ECU log.

Test 2:

1. Enable fuel & Ign.  Store power cycle.
2. Allow to rest with ignition on for say 30sec, confirm lambda is working.
3. Start engine, allow to run for say 30sec.
4. Stop engine, download ECU log.  

Lambda test.pclr

[jigga009]

I was able to perform Test 1

Internal Log: https://www.dropbox.com/s/mpjrwkgodpgrop0/Log 2021-09-20 10%3B54%3B35 pm - LambdaTest 1 log results.llg?dl=0

No errors at all from what I could see, oddly enough. Sensor heated up, as you said it would. Threw a few extra attempts to crank the engine over to see if anything popped up, and nothing. 

I'll have to wait on performing Test 2. My neighbors will shoot me if I start the car at this time of the day. 

 

Thanks very much for your assistance so far. I'm very appreciative.

[Adamw]

Im suspicious it is maybe "browning out" due to low voltage during cranking.  If you look at CAN ANV 2 in the log below, this is the voltage received at the CAN lambda, it drops to 8.5V during cranking, that is with the sensor already warmed up so the heater is pulling little current.  The "Battery voltage" channel in the ecu is heavily filtered so that doesnt show the drop. 

 

 

Can you also do another similar test, but this time change the offset on stream 1 to 0, this will mean the lambda doesnt turn on until RPM is seen, and we will see what effect it has on voltage drop. 

  

[jigga009]

Log with CAN RPM offset stream set at 0 - https://www.dropbox.com/s/4fei0l4c5lis9m6/Log 2021-09-21 1%3B44%3B35 pm - new log with CAN RPM offset of 0.llg?dl=0

I did not notice any errors pop up here either.

Here is a video I shot visually monitoring battery voltage and amp draw during a cranking session, as I was curious as to what the voltage drop and amp draw were at the battery during engine cranking.

I have noticed that the battery voltage measured by the ECU in PC Link is lower than what is going on at the battery terminals:

https://www.dropbox.com/s/g5yw7pwcw69i21y/IMG_0392.MOV?dl=0

This is the exact battery and model in use - https://www.northstarbattery.com/product/nsb-agm34-78

1AWG battery cables on positive and negative sides running to the engine bay where the CAN Lambda, main fuse box, AEM 4CH wideband, CDI, starter motor, etc. is fed 12V from.

1AWG Negative cable from battery is grounded at the engine block next to the starter motor, with another 1AWG wire running from that ground point to the chassis in the engine bay. Each head is also ground via 8AWG wire to the chassis.

Another 1AWG ground is run to the chassis in the back from the negative battery terminal. A 200A magnetic breaker is also installed near the battery. It's a bit surprising that voltage could be dropping to anywhere near 8 volts at any time.

It's a bit surprising that voltage anywhere in the system could be dropping to the 8V range at any point.

 

[jigga009]

On 9/21/2021 at 6:04 AM, Adamw said:

Im suspicious it is maybe "browning out" due to low voltage during cranking.  If you look at CAN ANV 2 in the log below, this is the voltage received at the CAN lambda, it drops to 8.5V during cranking, that is with the sensor already warmed up so the heater is pulling little current.  The "Battery voltage" channel in the ecu is heavily filtered so that doesnt show the drop. 

 

 

Can you also do another similar test, but this time change the offset on stream 1 to 0, this will mean the lambda doesnt turn on until RPM is seen, and we will see what effect it has on voltage drop. 

  

I would have thought that the capacitor that you had me install would smooth out the power supply to the Can lambda. Are you thinking the problem here is that the CanLambda (as it comes out of the box) turns on at the wrong time? Or perhaps an issue with the relay in use? Weak battery (even though it looks to have no issues turning the engine over)?

Thanks again for your assistance!

[Adamw]

The capacitor is to filter out high frequency spikes.  One possibility is the voltage at the lambda controller is dropping too low during cranking so it browns out.  But it may also be something noise from the CDI or similar which is why I wanted a log with the engine starting too.  

 

[jigga009]

59 minutes ago, Adamw said:

The capacitor is to filter out high frequency spikes.  One possibility is the voltage at the lambda controller is dropping too low during cranking so it browns out.  But it may also be something noise from the CDI or similar which is why I wanted a log with the engine starting too.  

 

Understood. I'll should be able to grab one in a couple of hours once people are up and about.

I will revert the RPM offset back to 500 and then take a log with the engine running. 

Thanks again.

[jigga009]

13 hours ago, Adamw said:

The capacitor is to filter out high frequency spikes.  One possibility is the voltage at the lambda controller is dropping too low during cranking so it browns out.  But it may also be something noise from the CDI or similar which is why I wanted a log with the engine starting too.  

 

Ok I was able to get a log for Test 2 using the "Lambda Test" map you attached earlier. I was able to confirm that the wideband was warming up before starting the engine. I could be wrong, but I don't believe I saw any errors this time either:

https://www.dropbox.com/s/nzasemjhh1eth3b/Log 2021-09-23 6%3B41%3B36 pm - TEST 2 Log.llg?dl=0

Prior to taking the log though, I took the liberty of conducting a proper load test on my battery using a carbon pile load tester just to rule out a weak or defective battery:

https://www.dropbox.com/s/gt5qsoxguh1l9jg/IMG_0398.MOV?dl=0

Thanks again for your help

 

 

[Adamw]

Yeah, it looks happy in that log.  Can you now do a few starts with the RPM offset back at zero so the lambda work as normal.  

[jigga009]

13 hours ago, Adamw said:

Yeah, it looks happy in that log.  Can you now do a few starts with the RPM offset back at zero so the lambda work as normal.  

Log 1: Data log with a few starts with the RPM offset back at zero:

https://drive.google.com/file/d/1B5i-EGwB5WGZOvAGPR-xrz-7Xgg8C6ti/view?usp=sharing

Seemed to capture only one starting event, so I took some more logs

Log 2: Took a second datalog with multiple starts, but I realized after that forgot to disconnect my battery charger during Log 2:

https://drive.google.com/file/d/1XSM0ce1JLy_3cNCsTwa-8CyczfVoxupx/view?usp=sharing

Log 3: Third time's a charm - Took another log without the charger attached. Was able to capture multiple starts:

https://drive.google.com/file/d/1cJ6aN0BMoriHsKp3MPe3DATy4JLeJksP/view?usp=sharing

[Adamw]

So it seems it has not errored for the last few tests you have done.  Do you want to put the CAN set up back to CAN lambda now and see what happens.  Im still suspicious about the power supply, it still drops to 9.3V during cranking in the log with the battery charger connected.

[jigga009]

41 minutes ago, Adamw said:

So it seems it has not errored for the last few tests you have done.  Do you want to put the CAN set up back to CAN lambda now and see what happens.  Im still suspicious about the power supply, it still drops to 9.3V during cranking in the log with the battery charger connected.

 

Thanks for the response!

Do you mean add CANLambda into Channel 3 while retaining the stuff you did on Channels 1 and 2?

[Adamw]

Change channel 1 to CAN lambda, turn off channel 2.  Both those streams are duplicating the normal CAN lambda stream with a bit of extra stuff added in.

[jigga009]

 

Before continuing, I wanted to spend some time scrutinizing the power supply, looking for anything out of the ordinary to explain the apparent low voltage that the Can Lambda is seeing when the engine is turned over. I learned how to perform voltage drop testing, made a set of extended test leads, and performed some voltage drop testing to get an idea as to what and where the voltage losses were when amps were being drawn through the electrical supply.

I performed a voltage drop test on the positive side and observed a voltage drop of 0.56V while cranking the engine over between the battery positive and starter motor. I then made a few changes in the name of efficiency, namely removing these items:

and replacing them with crimped copper lugs and some official Mil-Spec battery terminals, which allow me to crimp copper lugs onto the positive and negative wires, and then bolt them directly onto the terminals. All other connections in the electrical system are now utilizing crimped copper lugs.

After making these changes, I tested again and observed a voltage drop of 0.40v while cranking the engine on the positive side, and 0.159v on the negative side (between the battery and the starter motor). Voltage drop between the positive battery terminal in the boot of the car and the copper lug connected to the fuse box in the engine bay that feeds the can lambda 12v is 0.428V while cranking (a process that draws ~190A initially, and the quickly drops to ~140A). Aside from going with a larger cable size (currently running 1AWG), I'm not sure what else I can do to bring down the voltage drop and increase the voltage Can Lambda sees when the starter motor turns.

Again, the battery (rated at 1000CA/880CCA) has been tested with a carbon pile battery load tester, and it is able to withstand 240A (440CCA) load for 15s without dropping below 10.8v at any point.

Thoughts? Is the voltage drop too high?

Thanks again!

 

 

 

 

 

 

[jigga009]

I took another datalog using "Lambda test.plcr" to see if anything changed with the voltage CANLambda was seeing during engine cranking subsequent to the information gained from testing voltage drop and making the aforementioned changes. Battery is fully charged, and I waited for the wideband sensor to heat up before engaging the starter:

https://drive.google.com/file/d/1sb00XU0eY9scUGkEIf253QlMaI73g6yR/view?usp=sharing

Is the power supply still looking suspicious? The part that I'm still a little confused about is that both the ECU and Canlambda appear to report lower voltages than I am able to repeat when I put an even higher load on the battery than the car does during engine start.

A maximum of around 190A is momentarily pulled during engine cranking (at which time Can Lambda is reporting seeing some rather low voltages, yet when I pull 240A (~440CCA) out of the battery for 15 sec at a time, battery voltage does not drop below 10.8V. 

[MagicMike]

Interesting that my ECU displayed battery voltage is generally about .5v lower than actual battery voltage, and I've been having CAN lambda errors aswell.

[Adamw]

Definitely some noticeable improvement there, drops to a minimum of 9.8v at the start of cranking and sits at 11ish during cranking, so thats come up about 1V from before.  

[jigga009]

9 hours ago, MagicMike said:

Interesting that my ECU displayed battery voltage is generally about .5v lower than actual battery voltage, and I've been having CAN lambda errors aswell.

 I have come across a few other threads where people have mentioned it in passing, but nothing substantive really came of the inquiries. I initially thought they might have been programmed with some sort of offset to take into account voltage drop, but it doesn't explain Can Lambda reading low, as I have used 3 multimeters now to probe the voltage at the power lead for the device. My only possibility is that there is something about the DTM connector on the can lambda harness that causes a bunch of voltage drop. I will have to look closer at it today.

Here is a thread where someone had a similar issue - https://forums.linkecu.com/topic/8905-ecu-voltage-lower-than-battery/

As irritating as it is, it might just be normal.