First off I want to say we really appreciate this site and what it has contributed to so many equipment related issues that keep business's like us going and making ends meet. I'll be real blunt about this subject because it is a sore one for me but emissions is absolutely a fricking joke !!! I was just at the Service dept. at my CAT store. I randomly asked one of the mechanics how much of the work is related to emissions ? He said, 80 % at least ! This is such BS !! Do we have any idea how much money Cat and all the others are making on parts that go out that's related to emissions? Its astronomical I would bet. Anyway.. had to get that off my chest. I feel better now. Back to task at hand... This is what I have done so far. First off, For the last several days I have been on phone with Cat "techs" from at least 3 different dealerships and I get conflicting info from each and everyone of them. So I wanted to try this approach and see if anyone could point us in the right direction. It all started with a 3138-12 code like the pic above. (actual pic) Getting directly to this sensor was a little time taking off the stack or air filter assembly so we wanted to check the power and CAN Bus data wire coming to that via the firewall at CONN 39, We started by checking voltage coming too pins 1,2 and 11,12. Pins 1 and 2 should be battery voltage and checked out good at 12volts. Pins 11 and 12 are CAN high and low. We checked voltage for pin 11, (phase to ground) 2.55 volts and pin 12, (phase to ground) 2.45 volts. Also checked the ohms across phases at approx. 112-115 ohms but of course we are one 120 ohm resister short because of the CONN 39 being disconnected. At this point we decided that it was a bad NH3 sensor and replaced it. Thought this would definitely be the problem but after reassembling everything still the same codes as the above pic. So decided to make absolutely sure no wires were open from CONN 39 at firewall to the 4 plug sensor. We made sure both connectors were disassembled and using a power supply verified all 4 wires could carry voltage. Check mark and good. Next we plugged CONN 39 back together and did the same test as described above on CONN 39 except this time we had the other 120 ohm resister "in the loop". So as we expected ohms reading across pins 2 and 3 at sensor was 58.7 ohms which is correct. Voltage on Pin 2 and 3 at sensor were more less identical at conn 39, being pin 2, 2.56 volts and pin 3, 2.46 volts. Also checked Pin 1 and 2 for battery voltage and was 12 volts as expected.
421-7787: Sensor and Exhaust Controller, Could it be this ??
bump
New Module # 421-7787 installed and still exact same codes as before. Anyone have any ideas at all on where we could look ? Would greatly appreciate it.
I don't know a thing about this issue. But to do any sort of diagnostics you, have to have all the correct information on how it works, what all situations can set the codes, a good schematic and a nice trouble shooting list also helps. Always start by verifying the powers and grounds for the modules etc. Verify signals at and from and or too the sensors, sometimes a scope is needed for that. A corroded connector, an internally broken wire etc can be the cause, and of course check all fuses. Battery terminals to where ever the wires go need to be clean, and have to be disconnected to clean well. Exhaust sensor? Explain how it works? What does the book say? Where does the signal go and what is it used for? is the CAN signal clean?
You obviously have the electrical schematic for the machine. In all the communications you had with dealer techs did anyone supply you with the attached troubleshooting procedure.? The only thing I can suggest is an oscilloscope, something like a Picoscope, on both of the CAN wires somewhere in the machine harness and look at the CAN waveform. Then start disconnecting ECMs one a time and see what happens to it. In Post #24 on this thread you can see a sample image of what a good CAN signal should look like. HeavyEquipmentForums I never worked with emissions systems but I can look up procedures.
Thanks Nige. I had the Cat mechanic come out and hook his Scanner up an hour ago. It took some time to finally get the original code cleared (3138-12) directly related to the ammonia sensor and module. Of which both items were replaced. Now it has a (3609-12) Nox level sensor code which has not showed up at all until the ammonia sensor code went away. That particular part is $1000.00. I am a little concerned about just throwing parts at this when the initial problem may be elsewhere. But as I type my daughter is going to Cat to get the NOX senor for outlet. By the way.. Cat updated the Nox inlet and outlet sensors to a different part number. They said they would recommend replacing both nox sensors to the upgraded part number so they will communicate better. So the 2 upgraded sensors are 2008.00 if I decided to go that route. The mechanic said it might work for a while but just not sure. Anyway.. Ill try and keep you guys updated. Thanks for the input !
BTW... def looking into oscilloscope for sure !
Sorry I uploaded the incorrect troubleshooting procedure above, although in the long run it may be useful. See this one. The troubleshooting for your current 3609-12 is the exact same Datalink sensor procedure.
The four channel Pico is the one that seems very popular with diagnosticians. Not cheap, but seems to be rapidly becoming a must have for all electronic gremlins, not just for aftertreatment systems.
Thanks for posting that post #24 of a Cheaper Scopemeter and some of the info that came with it. Was busy working when that went through and never seen. Good for catching all those intermittent problems that go undetected generally. Simon C
There's a fella I follow on YT who specialises in automotive diagnostics. Whenever anyone questions the value of an oscilloscope I usually refer them to this video where it took him just minutes to track down the problem by watching the CAN signal on the scope and disconnecting modules one at a time. You'll laugh when you see exactly what was causing the problem.......
Wondering if it is way better than some of the Fluke units like 120 series meters. Just a Fluke 87V is close to $1000 these days in Canadian funds and can catch some things on Peak Hold but not recorded on a graph like these scope meters.. They all cost money but so do some sensors and harnesses that get changed out regularly with the problem still there. If nothing glitches on a graph then at least a guy can look elsewhere. Do you know of any other techs using the Flukes and are they as good or better or is it the price.? Simon C Thanks will look at later. Simon C
Wondering if it is way better than some of the Fluke units like 120 series meters. Just a Fluke 87V is close to $1000 these days in Canadian funds and can catch some things on Peak Hold but not recorded on a graph like these scope meters.. They all cost money but so do some sensors and harnesses that get changed out regularly with the problem still there. If nothing glitches on a graph then at least a guy can look elsewhere. Do you know of any other techs using the Flukes and are they as good or better or is it the price.? Simon C
WOW. Watched the whole thing. Just when he thought he had it figured all out he decides to check more of the Can Harness to find one section lying in a pool of washer fluid in left rear quarter panel. Scopes will soon be the norm is my opinion. Would of bought a pile of Control modules and harnesses to find that one without the Laptop and scope. Simon C
His two mantras are - "Data-driven diagnostics" followed closely by - "Fixed. No parts Required."
I didn't watch the link, it must be Pine Hollow Auto Diagnostics. Yes top guy for electrical on all types of machinery, cars etc. Pico ? I'm just stuck with antique Tektronix myself. A breakout box is nice to have as well. But I don't think equipment has an OBD standard like cars etc.
For cat you make your own breakout box by getting the male end of the diag connector and some sure seal Sockets and wire Of course that gets you global CAN only. The 2 channel $350 picoscope works great
Was just looking at a fluke 87v although just a decent Pico like Nige had mentioned might end up going with. Definitely need one these days for sure.
I have an 87V and an 87-111 and both have saved me a lot of money on my furnace repairs and airconditioning trouble shooting. They can read up to 1MS speed so can catch quite a bit and are way easier to move about and connect, but are not a graphic screen with multiple events plotted like these scopes do, with the super high speed. Each have their areas where they shine. In minus 30 my Flukes only last 30 minutes maximum, those scopes would be frozen up before the testing was done. Simon C
Why does this machine still have operator inducements and low nox conversion events active? Were they never cleared, or did they all come back after the NH3 sensor replacement? Has all the software been checked/updated? Have any diagnostic tests been ran? (aftertreatment system functional test) Did the functional test fail with any errors? If you had the large old style NOX sensors that take an adapter bracket to replace, those both need to be changed as they have a high rate of failure. You will also need to update software above a certain point for them to function. And ammonia slip could have destroyed the nh3 sensor, so all this may go back to a different root cause.
Dang Brother.... minus 30 !! I was just complaining to wife a few minutes ago about the temp being 12 degrees. Lol. Hell 12 would be like a heat wave compared to minus 30. That's colder than a witch's tit !!
This is somewhat a long story so wanted to follow up so this info might be useful to someone else. First this all started with a rodent. Chewed some wires over a period of a few months on the same job site. The first code they noticed and im going off memory was I believe was the NH3 sensor. Not long after that there were other codes. The hydraulic temp, fuel gauge and one that was related to the heater in the def tank for heating the def up to temps. So once we got the machine to the shop we found under the floor plate where a squirrel had chewed those wires that I just described. So I very carefully soldered those back and got rid of all the codes except the 3138-12 NH3 sensor. So in my mind im still chasing an open wire from the root cause of all this being a rodent. But after testing best to my ability and especially without an oscilloscope everything was pointing to a bad ammonia sensor. So I replaced that and no change in the active codes. So got on phone with another tech and he suggested the module that plugged into the sensor was bad. So replaced the module and still the same code 3138-12. In my mind if the sensor is good that particular code would go away besides having to reset the ECM with ET scanner for non related sensor codes, ie SCR related and aftertreatment. I had noticed several times when unplugging some other sensors or even 12 pin plugs and have to turn key on I would get many new codes but the minute you plug everything back in they would disappear. So my thoughts were why isnt this code for this particular sensor going away if everything is checking out. Also had the correct voltage on the CAN high and low and the correct ohm readings as well as power to that sensor but the code is still active. I even tried the ole battery disconnect overnight and come back next day with same thing. At that point i was stumped and knew I better get the mechanic with the ET scanner. So he came out and was thinking that everything would clear up after this. He ran an aftertreatment system functional test and we both were standing at the machine and he was pointing out that everything seemed to be working until a few minutes later and now we get a NOX outlet sensor code and seems the NH3 code has cleared itself. So... next thing was to spend over $1000 for yet another sensor. So first thing this morning I installed the new NOX outlet sensor and he came back over and ran another aftertreatment functionality test and forced regen. This time all codes cleared and so far everything seems to be good. So on the subject of the NOX sensor. Our machine still had both the old ones in it and Not the updated ones. He had mentioned that the updated NOX sensor might not communicate with the older stlye NOX which was the inlet. And CAT recommends replacing both. But I didnt want to drop another $1000 and told him we could try the new NOX first and see if it works. So far so good. So im curious what caused all this ? Was it just a random bad Ammonia sensor and NOX Sensor at the same time? Still not sure why the code that was directly related to the sensor not clear like others after being replaced. Did it have something to do with the chewed wires on the def heater ?? I always like to know the root cause of something if for nothing else the peace of mind. Any theories or ideas greatly appreciated !
Thanks for posting what you found to the benefit of others. Simon C
Attached the info regarding the updated NOX sensors. Note the requirement for updated engine software. I have heard the same in regard to replacing the NOX sensors in pairs but it does not appear in this article. Also interesting that there are Reman NOX sensors available, probably at around half the price of new. I bet the dealer never told you that.
The 11R sensors are a brand new sensor with a core. The worst problem is half the time you can't get the nuts loose with a wrench, so i cut the harness and put an impact on it to get it out, but then you don't get your core anyway.
Bugger.