299C Drive Speed Sensor Bypass?

[JASON SIMON]

Next finding. The right sensor harness has a connector just under the cab. The connector going back to the computer has the expected color coding : yellow, white, brown, Grey. So I think someone replaced that section with the wrong cable.

 

[Nige]

Maybe I need to reconfirm, the right side is when I'm sitting in the machine and to my right?

Correct.

bit more. I went out to move the machine to move it to a place I can wash it down to do the troubleshooting and got both the 2687-8 AND the 2688-8 codes. Coincidence?

The additional 2688 Code would have been created when you pulled the speed sensor harness connector apart. There are two separate sensors in a common body. The #1 sensor creates 2687 Codes, the #2 sensor creates 2688 Codes.

Next finding. The right sensor harness has a connector just under the cab. The connector going back to the computer has the expected color coding : yellow, white, brown, Grey. So I think someone replaced that section with the wrong cable.

That solves the mystery. There is an incorrect jumper harness installed but your Code is definitely for the RH side motor sensor.

 

[JASON SIMON]

So I'm trying to follow the wires from the jumper to where those wires terminate at the computer/box. I can't figure out from the schematic where to find the other end of this wire so I can test end to end.
The wires seem to join up with the engine harness, but it's not very clear where it ends up.

 

[Nige]

If you follow either of the signal wires from the speed sensor on the schematic back to the ECM you’ll find a hyperlink in the title of the control above the “box” as it’s shown on the schematic - something like “ECM-machine”. Click that hyperlink and it should take you to an illustration of the machine showing exactly where the ECM is located. I think the illustrations are on the last couple of pages but I could be wrong.

 

[JASON SIMON]

Okay so looks like they connect to the ECM which is behind the seat under the cab. Some mighty big plugs back there. Going out to test now.

 

[JASON SIMON]

Okay, done testing. I checked continuity and for shorts from the ECM connectors and all four pins in the sensor harness. Everything checks out good. My assumption is that this means it must be the sensor that is bad?

The troubleshooting step for testing the sensor seems to indicate that if the sensor is bad it will visually appear damaged. The only way to reach the conclusion that the sensor is bad is if the sensor looks damaged. In my case the sensor looks fine and all the wiring/connections checked out. That's a little confusing. I wish there were a way to test the sensor electronically.

Test Step 4. INSPECT THE SENSOR.
A. The key start switch remains in the OFF position.
B. The sensor remains disconnected from the machine harness.
C. Remove the speed sensor. Inspect the speed sensor for damage.
D. Clean the sensor with lint free cloth.
Expected Result: The sensor is not damaged.
Results:
OK - The sensor is not damaged. Reinstall the speed sensor. Proceed to test step 5.
NOT OK - The sensor is damaged. Repair: Repair the sensor or replace the sensor. Verify that the diagnostic tool has been cleared. Drive the machine above the threshold for the motor speed.

After connecting everything back up, the 2 RH sensor errors are coming up more regularly now, so that's progress.

Wish I could just get back to work, but can't risk the random drive cutouts while working in precarious areas around my pond. I'm even tempted to rent another machine for a couple days to finish my project. That's a sad thought given I own the machine already.
I'm exhausted from running the machine for 5-6 days straight and now 1.5 days troubleshooting. I guess I should enjoy the break and just wait for the new sensor.

 

[Chrisso]

Your biggest issue is that the fault is intermittent. During all those steps wiggle the wires a bit.

 

[Nige]

The troubleshooting step for testing the sensor seems to indicate that if the sensor is bad it will visually appear damaged.

Not necessarily. The obvious thing to look for before starting to test is to remove it and inspect for mechanical damage. If it is damaged then the obvious conclusion at that point is to replace it.

However, having eliminated mechanical damage as a possibility (and TBH not likely in your situation where you started out with only one sensor out of two in the common body throwing faults) then you have to move on to testing the electrical side.

Here's a thought. Do you have any way to get the machine up on blocks so that the tracks can rotate freely.? If so, and if your multimeter will read Duty Cycle (not all do) then you could back-probe one of the wiring harness connectors for the sensor while the motor is turning and compare the Duty Cycle of #1 sensor vs #2. Obviously they should be both reading the same.

Your biggest issue is that the fault is intermittent. During all those steps wiggle the wires a bit.

And therein lies the biggest challenge. Intermittent faults are one of the hardest to track down because Sod's Law will dictate that while you are testing it the electrical system will behave perfectly normally. Hence Chrisso's suggestion of the "Wiggle Test".

 

[JASON SIMON]

Not necessarily. The obvious thing to look for before starting to test is to remove it and inspect for mechanical damage. If it is damaged then the obvious conclusion at that point is to replace it.

However, having eliminated mechanical damage as a possibility (and TBH not likely in your situation where you started out with only one sensor out of two in the common body throwing faults) then you have to move on to testing the electrical side.

Here's a thought. Do you have any way to get the machine up on blocks so that the tracks can rotate freely.? If so, and if your multimeter will read Duty Cycle (not all do) then you could back-probe one of the wiring harness connectors for the sensor while the motor is turning and compare the Duty Cycle of #1 sensor vs #2. Obviously they should be both reading the same.

And therein lies the biggest challenge. Intermittent faults are one of the hardest to track down because Sod's Law will dictate that while you are testing it the electrical system will behave perfectly normally. Hence Chrisso's suggestion of the "Wiggle Test".

Thanks for the replies guys. I agree that the intermittent behavior is hard to explain if it is just a sensor.

I put a lot of wiggle in during the continuity/short testing, so I feel like I got that mostly covered.

Unfortunately I don't have a meter to test duty cycle, so it looks like I need to get one.

The new sensor arrives on Monday. I guess I'll just wait and throw that in and pray it works.

I suppose I could order a jumper harness and new meter as a backup plan...

 

[Nige]

Unfortunately I don't have a meter to test duty cycle, so it looks like I need to get one.

It would be a "nice to have" if it had that capability but if you have a new sensor arriving Monday it's not really worth it at this time. Although having said that a meter with Duty Cycle capability is very useful with all the PWM sensors on that machine. There are more than just the travel motor speed sensors I can assure you. https://www.circuitbread.com/ee-faq/what-is-a-pwm-signal

 

[JASON SIMON]

Thanks for the link on pwm. That helps explain the measurement of duty cycle really well. Got a meter ordered.

I'm wondering if you can do the duty cycle testing on the ground... assuming the test leads can be secured away from the tracks.
When you test on blocks are you testing each sensor at the exact same throttle position?

Just thinking this through...
If this is just a pesky connection/wire problem (not the sensor), it must be the 8v or return wires right? As a test, maybe I tap into/splice each of the 2 wires at the sensor and run 2 parallel wires up to splices in the respective wires at the ecu.

 

[Nige]

When you test on blocks are you testing each sensor at the exact same throttle position?

You have 2 sensors in a common body. What you’re trying to do is measure the duty cycle of one against the other when the motor is rotating at a given speed. If both are, say, at 47% or thereabouts then you would be good. If one is at 60% and the other is at 25% not so much. You test the output of the “good” one first then look at the other.

 

[JASON SIMON]

UPDATE: Got the new sensor and had it swapped out in 15 mins this time (as opposed to 4 hours the first time). Drove it around and still throwing codes for right side sensor.
I swapped the connections for Left and Right under the cab. Sure enough the code gets thrown for the left side. So that means it has to be the wiring.
Next I pulled the the jumper harness out and lo and behold inside the pass through the wiring had chaffed all the way through the sheathing and cut through one of the wires. It must have been shorting out on a corner of metal. I have no idea why my testing did not pick up this up.
I cleaned up the wiring really well and repaired it. It tested good on my meter. Put it back in the machine and drove it around and it is still throwing the right side codes.
I'm going to just order the jumper harness. That's got to work.
Such a bummer that this saga is not over yet. Ugh.

EDIT:
Was just re-reading another speed sensor thread and there was a gentleman who said it took 30 mins for his codes to clear out. Is that common? Maybe mine is actually fixed and the 5 minutes of driving it I did wasn't enough?

 

[Nige]

Was just re-reading another speed sensor thread and there was a gentleman who said it took 30 mins for his codes to clear out. Is that common? Maybe mine is actually fixed and the 5 minutes of driving it I did wasn't enough?

Is the Code still coming & going like it was when you first posted or is it on permanently.?

 

[JASON SIMON]

The 2687-8 code was showing steady now. Not intermittent like before.

 

[Nige]

Try running it for longer. If that doesn't make the Code go away then you are back to the troubleshooting procedure I posted on Page 1.

If the sensor is new and switching the wiring side to side moves the Code from RH to LH or vice versa then it has to be in the wiring. Maybe look again at the area of the machine harness where you previously found damage and repaired it.?

 

[JASON SIMON]

Ran it for another while and seems like it is re-throwing the code as soon as I go from stop to go. I'm going to test the wiring a little bit more and I guess just go ahead and order that $70 +$20 jumper harness if I can't fix it any further.

 

[Nige]

Correctly executed the troubleshooting procedure should direct you to exactly where the problem lies, especially considering that the Code is now permanently active rather than being intermittent.

If you want to load the parts cannon and throw money at it to make the problem go away then don't let fear stop you.

 

[JASON SIMON]

Correctly executed the troubleshooting procedure should direct you to exactly where the problem lies, especially considering that the Code is now permanently active rather than being intermittent.

If you want to load the parts cannon and throw money at it to make the problem go away then don't let fear stop you.

I know it sounds ridiculous to do that. We are seriously talking about 4 simple wires and their connectors that is suspect here.
I'm not an e. engineer, but I know enough about electronics and soldering to know how simple this should be. It baffles my mind that my resistance testing came up totally fine for all 4 wires prior to finding the damaged wire.
And testing the lines again today - all 4 check out perfectly. No shorts, no resistance. I still have the new sensor in so that seems like less of a culprit.
I'm going to repeat some tests when the temp cools down a bit - including the left/right swap.

 

[Nige]

I know you had problems getting the old sensor out but I would still suggest to swap the new sensor to the LH side and see what happens to the Diagnostic Code. There is always a (slim) possibility that the new sensor could also be bad. It wouldn’t hurt to rule it out.