Loss of ODB2 Port network connection, I think it's a TIPM problem

[MichaelOverfield]

Well it seems the TIPM stories abound, and I may have a new twist. After looking around, I have not seen anyone else with similar issues, and I'm now about certain it's the TIPM.

So to the point, then I'll follow with history and how I think I've nailed it to the TIPM:

When connecting any OBD 2 scanner of device, it initially works after engine start and all seems well.. but then at some point the device (any device and I've tried about 4 completely different devices), it losses connection to the CAN Bus (ground and voltage to the pin connector are fine and verified).

The CAN connectors, 4 and 16 if I recall, the Service manual states should show a 60 Ohm resistance, and they do, when working or when not working, so connection to the TIPM is not the issue either.

So reading how the CAN network is set up on these vehicles, the TIPM is basically the network hub, the CAN C (engine, ABS, critical items) and CAN B (Body and less critical items, like radio wireless module etc). All connect to the TIPM, and then the TIPM makes the diagnostic CAN D (basically a connection to either of the networks) available at the OBD 2 port on the dash.

When the network connection to the OBD 2 port is lost, the engine and all other car components appear to work fine, I've seen none of the other typical TIPM issues like fuel pump, horns, lights, etc. It appears all other parts of the TIPM are working fine. I even took it out, opened it and did a detailed inspection, and it was dry, clean, and free of any obvious fault, all solder joints looked fine, pins, etc.

Now a bit of truck past history that now seems to line up with bad computer glitch behavior from the TIPM..

Long ago, when still under waranty I got the MIL/CEL light. The code was a strange one, and the dealership looked at it and cleared it.. when I asked what it was they said it was a code for the AUTOMATIC TRANSMISSION.. but I have a G56 Manual! Odd as it was, it never came back.. but the selection of manual or automatic transmission is a TIPM parameter!

Then I added an Ultraguage OBD-2 monitor. I've used them on several other cars and they work well and are simple, cheap and reliable. But from the beginning, it would sometimes appear to shut off in the truck. At first thought it was the ultra guage, but it can be commanded on with the truck not running, and it comes on, but sees no ECU, so power is fine, connections are fine.. it has been that was for some time.. Then it seemed to work pretty well most of the time. so I didn't trace it further..

When I put in an aftermarket radio, and assciated interface box, that had some issues, I assumed the problems with the OBD 2 and CAN bus might be from that interface box, which is connected via the CAN B network to the TIPM.. So I got a new radio and interface box, thinking that might fix it.. but no..

Now with new radio, that shuts off, a handy feature for sure.. The OBD 2 port is still unreliable.

Next issue, the massive current demand of the Grid heaters.. I know they normally cause stress on the charging systems, but I'm convinced mine was excessive.. I have dual 220A alternators, and I have output current meters.. those heaters are pulling about 250Amps! Seems excessive to me.. and they causes a voltage spike when they shut off, with one or both alternators running.. so I disconnected them, and that did seem to help keep the TIPM from shutting down the OBD 2 network connections, somewhat, though it has not completely fixed it as it shut off on one of 2 trips today. So I'm still dealing with that OBD 2 port shutting off at random.

I'm looking to get another TIPM, but they are expensive and not in stock most places.. when I found one at a price that seemed reasonable, I ordered it but won't see it until August... maybe..

In any case just sharing that here for others to comment and share from my situation and perhaps someone has additional insight or thoughts. I'm going to make a TIPM saver to isolate the trailer harness, a known area that has killed TIPMS, and go with lower rated fuses for associated outputs.. Also, any relay is going to get a Diode to kill off any inductive kick upon shut off.. I'm going to go extra cautious on this TIPM, and the next.. and I have no intention to return the core, because seems having a spare is a good idea seeing as they are hard to get, fail often, and cripple a truck. This one is an annoyance with the OBD 2 connection, but everything else seems to work.

 

[sag2]

OK, to be clear, the tool stays powered up but you loose communication with the truck?

Before I spent the money on a TIPM I would replace the DLC connector pins. I have seen it before where it appeared t be a module, but was in fact a spread terminal at the DLC. The complete repair kit (5183480AB) is kind of expensive, but you could pick up any DLC at the wrecking yard and replace the diagnostic CAN C terminals.

 

[MichaelOverfield]

OK, to be clear, the tool stays powered up but you loose communication with the truck?

Before I spent the money on a TIPM I would replace the DLC connector pins. I have seen it before where it appeared t be a module, but was in fact a spread terminal at the DLC. The complete repair kit (5183480AB) is kind of expensive, but you could pick up any DLC at the wrecking yard and replace the diagnostic CAN C terminals.

Thanks for the reply. It's not the pins, if it were I'd not see the 60 Ohm reading. It's clear the TIPM is just stopping the communication to the connector.. it is bizzare, and I have not seen anyone else with this issue. Think if that part of the TIPM as a nework HUB or a better comparison, a router on a LAN.. it appears to be stopping traffic to the CAN D (the connection from the TIPM to the OBD2 port (DLC). It has been a problem for a long time, but after rulling out any problem with the radio interface box, I moved to the TIPM..

Also, I just put a DORMAN REMAN unit in (all are REMANS, the OEM wants like $1300 (can you say total rip-off!)). O'Rielly had a 15% off promotion going, so I used that and got one for about $400 (after I return the $200 core). I'd kind like to have a spare, but I have another lower cost A-1 CARDONE unit on order from Summit Racing for $290 + $130 core.. but it's due to ship in August, turns out these TIPMs are in short supply, so I wanted to get what I could get at a resonable price.

I just put it in earlier today, at first thought I'd have to program the TIPM, it's supposed to auto learn after about 12 seconds, and a key cycle, but it did'nt at first try, then appears to have done it second cycle. I have the Auto Enginuity PC Scan Tool, so can access the TIPM directly if needed, but it worked, and I then used my tool to do a full system test and so far all appears to be well.. but I'll need multiple long drives to confirm.. so I'll do one on Thursday and see if the problem has been corrected, I suspect it will solve the problem.

 

[MichaelOverfield]

Well, it appears to be the gift the keeps on giving.. I am really getting to hate anything computer related or controlled..

So the TIPM replacement appeared to solve the problem.. several start cycles, all was well, so I did a 45 minute drive... all was well! I'm thinking, great, its fixed! (Bad idea to think that way it seems).. then about 30 minutes into the 45 minute drive home.. loss of connection for the ultragage, it shuts off (still has power at OBD2 port), and I tried the bluetooth OBD2 LX, same result no network! I pulled over at an exit, shut off the truck.. paused a few seconds, started back up, all is working again.. It is bizarre. So I scanned to pull codes with Torque Pro.. got P00FF and U3FFF. Niether code appears in the RAM DTC list, service manual, or online! I'll have to run AutoEnginuity later and see if it has any other codes... I went ahead and turned in the core original TIPM, over $200 core charge for that thing, and while it is now not for sure fixed, it is defeninitly improved.. so the original TIPM had some part to play, but this thing is perplexing for sure, anyone with additional ideas, please share. I'm thinking about tracing out the network connections between the TIPM and ECU (CAN C bus). They can't be checked with a standard multimeter per the service manual, so it gets more tricky, and even then it is clearly an intermittant issue.. but makes no sense that it is less an issue now with the new TIPM.. seems it's a network communication problem for sure.

 

[seafish]

Is it possible that the ultra gauge is part of the problem?

Also, while I generally do NOT recommend an oem software reflash from a dealership, this might be a case where you should consider one (of course I am assuming that your truck hasn't one since this problem manifested)

 

[MichaelOverfield]

Is it possible that the ultra gauge is part of the problem?

Also, while I generally do NOT recommend an oem software reflash from a dealership, this might be a case where you should consider one (of course I am assuming that your truck hasn't one since this problem manifested)

I ruled out the ultragage, becasue it powers up in override, so it has power, but it is desiged to shut off when it losses network connection. I tried the OBD 2 LX Bluetooth scan tool, it also has no network connection. To futher confirm, on prior testing I used my AutoEnginuity scan tool also.. none of them can connect. When it is off, it is off.. I also checked pins, and wiring to the TIPM, all good with the 60 Ohm per the service manual. Now that I say that... I will try to leave the ultragage disconnected, and use the OBD 2 LX instead and see if it does the same, where it drops off after some time.. I long thought the network could be being cut off due to interferance of some kind.. if the ultragage is causing that.. perhaps.. but that would be surprising if it were. I've used the OBD 3 LX for years on my Cruze Diesels without issues.

As to a reflash, I know it had one some time back, when there was a update for the high idle to mitigate soot issues in the DPF, I believe the early issues I mentioned in the OP, like the odd MIL for an automatic transmission.. that I think was pre-flash. Now that you ask, I think I'll pull up the service records and verify. The odd behavior with the OBD 2 shutting off goes back many years, to when I first started with the ultragage, and at first I assumed it was the device, then maybe power to the OBD2, but rulled both out, and know know for sure, the network is just not there, since multiple scantools can't connect. It's always there for a bit after start-up, then after some time, it just shuts off... It really is bizarre. Now if I didn't have a monitory on the OBD 2 port, there is no way I'd even know there was a problem.. unless I tried to hook one up and found it did not connect, but most people would assume it was the scantool, as I did.. but multiple tools, that all work on other vehicles.. and the connection to the TIPM is sound electrically, and there is voltage at the OBD 2 port... there are actually only 4 wires of the 16 pins on the OBD 2 on this vehicle.. Ground, 12 V, CAN D+, and CAN D -, that is all there is on that 16 pin connector!

 

[sag2]

Are you front probing the pins on the DLC to check for voltage? If you are, I still say to replace the two BUS terminals. When you are sure those are perfect, get a dummy DLC connector like this one Amazon product ASIN B078XTBJSZand you can then probe the wiring from the back side without sticking a probe in the female DLC pins. I have seen a couple vehicles where they lost communication with the tool and it was just the DLC.
Once your sure that side is good you should confirm the diagnostic BUS connection at the TIPM. Just don't front probe any terminals.

 

[MichaelOverfield]

Are you front probing the pins on the DLC to check for voltage? If you are, I still say to replace the two BUS terminals. When you are sure those are perfect, get a dummy DLC connector like this one Amazon product ASIN B078XTBJSZand you can then probe the wiring from the back side without sticking a probe in the female DLC pins. I have seen a couple vehicles where they lost communication with the tool and it was just the DLC.
Once your sure that side is good you should confirm the diagnostic BUS connection at the TIPM. Just don't front probe any terminals.

I sure wish it would be the terminals.... just because it's easy, I'll look at them again, but it's very unlikely to be the cause.

I don't jam a probe into the connector when I check, to be sure to not damage the pins, I have a very fine point tip probe that I can use to avoid that problem.

I just ordered an O-scope, so I can check the CAN C bus with the O-scope for normal signal patterns. If it's not a TIPM issue, I could have an issue between the ECU and TIPM. The article I was reading about CAN made some very intersesting point about it being critical that the connections be twisted pair wire, and shielded, with the shield to ground on only one end of the wire. It's apparantly susceptable to interfereance and electrical noise.. That might be the issue actually.

 

[MichaelOverfield]

Found this about a CAN bus "off-state". It appears as if the TIPM is shutting off the network connection to the OBD2 DTC port.. perhaps there is interferance on the line.. In any case thought this was interesting. When I get the 2 channel O-Scope and have a chance to look at the CAN signals in real time.. might have more to post. Will also be doing a very, very hard look at ever possible CAN network wire and connection I can track down.. Updates to follow. https://www.embeddedrelated.com/thread/939/what-is-can-bus-off-state

 

[MichaelOverfield]

Re-verified terminals, pulled them out of the connector and inspected. No issues with the terminals. When off it gets the 60Ohm as per specifications. Waiting on an O-scope, but checked "on" voltage to ground on CAN H and CAN L. Both are within expected voltage (best I can do with a multi-meter). I re-checked wiring to the TIPM.. looked for any possible issues.. none found. Just to verify it was not an issue, I replaced the OBD 2 splitter (I have one to plug in the diagnostics without unplugging the UG or OBD2Link LX). After several start cycles, all seemed well, but it appears it takes time to appear. I did a short drive.. about 10 minutes in, it shuts down the port. Did the same on return.. always comes up at start. Next phase, I have the UG unplugged... perhaps I'll be lucky, and somehow it is the problem.. but I generally don't get so lucky.. more to follow.

 

[seafish]

So have you removed BOTH the Ultra Gauge and also the Port splitter and then drive to see if it happens with ONLY a diagnostic reader plugged in??

If not, then I would DEFINITELY try doing that and drive around for awhile to see what happens!!

 

[MichaelOverfield]

So have you removed BOTH the Ultra Gauge and also the Port splitter and then drive to see if it happens with ONLY a diagnostic reader plugged in??

If not, then I would DEFINITELY try doing that and drive around for awhile to see what happens!!

When it shuts off, I can unplug everything, plug in my AutoEnginuity Scan tool, and it will not connect, it sees no network. When I get the O-scope, I'll be looking at the network signal directly to see if there is some kind of interferance on the line, so to speak. I'll also be looking at it as soon as it shuts off, to see that is going on there.

When this all started, I did not have the splitter, or other OBD2Link LX, just the UG, and I thought it was the problem, or power at the plug.. but both ruled out... that was years ago now. This has been an issue for a long time.

It consistently resets and comes back if I cycle off the engine and pull the key, then restart. That would not be consistent with a wiring issue, so it leads me to beleive its an intentional network shutdown to that port, which is between the TIPM and plug, the TIPM is the "router" in this network. Since all the other items appear to we working normally, yet I say appear, as I have no way to monitor when the DTC is off... but it seems the rest of the network is communicating, or I'd have more serious problems. The odd non-sensical codes come up periodically, the latest was a C2300, before that I had P00FF and UFFF. None of these make any sense, and the U3FFF points to a network issue, but it's not a code that exists in the DTC list for this vehicle.

I will be driving with only the OBD2Link LX and the Torque Pro app on the new radio and see if it holds up, I'm hoping it does and it is somehow a glitch in the UG, but I generally don't get that lucky.

 

[MichaelOverfield]

deleted duplicate post

 

[MichaelOverfield]

So, saga continues. Drove the truck again today to see if UG unplugged would solve the problem.. Nope.. Same thing, about 10 minutes in, OBD2 network signal dropped to the OBD2Link LX.. a device proven reliable with many, many hours on my other vehicles... so it's not the plugged in devices. The CAN bus is shutting off to the DLC. It's bizzarre. The O-scope came in today, so I'll have to get that working to look at the actual network signal to see if there is some noise on the network causing a shutdown. I might hold off a few days to get the break out box, which will make it easier to connect, and it may require a science project to keep it hooked up and logging data and do several drive cycles to get it to act up, it does not always shut down, the drive home it worked fine.

 

[MichaelOverfield]

OK, Big update. It appears it is NOT, and has never been the TIPM. As far as I can tell. It appears the ECU is in fact the most likely culprit of the problem. Here is how I came to that conclusion, though I still have some work to do on verifying the ABS isn't causing some network issue.. so more to follow.

So, I have have the wiTECH Micropod 2 adapter and software, and got a 7 day Mopar TechAuthority subscription. The tool and software as a bargain, at about $150, the 7 day subscription is $70. With that I have the capability to do whatever the dealership can do, including flash any computers in need of an update. (I highly recommend this tool, it does everything my AutoEnginuity can do, and more, and even in off-line mode which does not require a subscription to be active). Then if you need a flash, you can do a 1, 3, or 7 Day TechAuthority subscription at a cost well below a dealership doing it, which is very handy.

To make all this work, I had to ensure I could maintain 13-13.5volts with the ignition on, engine not running, turns out you need to have about 30Amps to hold the voltage when the various electrical items actuate, so I made a power supply for about $20 to ensure I had that covered, a new tool that can be used for any of my cars when needed.

I also now have a LAN drop in my Garage, because the Micropod is computer and network based, and it needs a Windows 7 computer, just so happens my aging laptop that can't update to Windows 10 fits the bill perfectly. I did not want an issue during flash with the network, so used the Cat 5 cable to connect vs. WiFi, which is available but not great signal in my driveway.

Included my efforts was an OBD2 Breakout box, and 2 Channel O-scope, also connected to my LapTop.

So first thing I did was connect to vehicle with the Micropod and wiTECH software, and log in to the TechAuthority so I could verify the ECU flash. Turns out one possible issue in the TSBs listed per the VIN was a CCN (Cabin Compartment Node, think dash cluster). There was a need for some vehicles to be re-flashed following the J35 Emmissions Recall flash of the ECU, thought maybe that could be it, but NO, my CCN is up to date, no newer flash available there.

My ECU did have the most recent flash, but it happened to be the only computer that had the option to re-flash, even when it already showed it had the most recent version.. that is a bit interesting and tells me that it must be a thing that is done on a regular basis, and I can think of a few reasons as to why, but in any case thought that I'd go ahead and re-flash to see if it had any effect.

The tool works very well, and the re-flash completed without issue.

So, next up was the O-scope. From what I could find online, I should see symmetrical square wave like digital signals between the CAN H and CAN L connectors.

So I hooked up the O-scope and after getting the right time scale and volt scale I could see signals, but not always symmetrical, and the CAN L amplitude was often about HALF the amplitude as the CAN H. Very weird, but more concerning, many signals were not digital square wave forms, and then there were some wild fluctuations periodically that are not normal.

Yet despite that, all seemed to be working, and the wiTECH software comms test between all modules was showing proper communications.

The problem of the loss of OBD2 port was not in play, YET.. so I left the key on, and diagnostics connected for an extended period of time, also had the truck running for over an hour to see if I could get the condition to happen and be able to see something.

Well, after 6 hours.. finally something. the wiTECH showed the ECU (on the diagram it shows as PCU, but that's just a software one size fits all issue, for the Cummins the proper term is ECU, but I digress). In any case this situation IS the condition that led to the OP, and the TIPM replacement, but the wiTECH was still connected to all the CAN B modules, verified by running tests, the only questionable connection is the ABS Module, that runs on CAN C, with the ECU.. but the ECU was in red, and completely not responsive, and all data for ECU parameters was unavailable on the CAN D connector. So, I know know that the CAN D line, and DTC/OBD2 connector is not in fact the issue, the ECU is dropping out of communication with the rest of the car.. I't is basically deciding to not play, but oddly the rest of the modules apparently don't care! No DTCs on any other module!

Now during this portion of my marathon troubleshooting saga.. I had the O-scope running as well to see what was going on there, and I noted a trend, when there was communication with the ECU, the voltage spikes and other improper waveforms were more frequent, so when the ECU is talking, it appears to be putting out some interference on the line.. interesting.

So to get the ECU back in communication, just like I have observed over the many years of this issue, shut it all down, wait a minute, and restart, just like the old MicroSoft reboot, it comes back up and is working again..

So, I proceeded to do another re-flash, and reset everything.

During the second re-flash, I monitored the signal on the O-scope, this is intensive comms on the line with the ECU, and it was going crazy with improper waveforms, but it did take the flash, so it seems comms are getting through, but there is something just not right.

I'm going to double check the pin to OBD 2 break out box today to verify it's not an issue with the test set up, and look at the ABS module connectors, because I noted that despite the wiTECH saying that the module was communicating, during the ECU loss of comms, I tried a pump test and it would not respond, yet it did at other times.. so there still could be something there, and if it's causing the fluctuations, it's possible the ECU goes into a communication off state due to interference on the line.... yes, a longshot, but this entire saga has been all longshots it seems.

 

[seafish]

Please remind me if you gave already checked and cleaned the two ECM connectors and pins....IIRC they are prone to corrosion which could well be the source of your signal interference.

 

[MichaelOverfield]

Please remind me if you gave already checked and cleaned the two ECM connectors and pins....IIRC they are prone to corrosion which could well be the source of your signal interference.

Actually, not yet, but good idea. The truck is low miles, clean, and never been driven on salty roads, so I would be surprised if that were the issue, but certainly worth looking at. In fact I was thinking of pulling the ECU anyway to get the actual numbers off the unit itself. It should be a CM2100 Cummins ECU, and the price range is all over the place for that unit, and most way over priced and it ticks me off there is such a market of rip off artists out there, but it certainly exists.. How can one company provide one for $600, and another wants an insane $1250, even a "repair service is about $250, and frankly I'd bet the repair would be test, see it works, and ship it back, nothing fixed. It's intermittent and took almost all day to get it to act up.. but I would LOVE it to be something as simple as the connectors.

 

[seafish]

Actually, not yet, but good idea. The truck is low miles, clean, and never been driven on salty roads, so I would be surprised if that were the issue, but certainly worth looking at. .

Given how clean the truck is, this suggestion might not be necessary... and also judging from your meticulousness, I am sure that you already plan to, lightly clean the area around the ECU and connectors with a light brushing and/or solvent and compressed air to prevent any dirt from getting into the connectors when you remove them from the ECU.

just my .02

 

[MichaelOverfield]

Ok, some progress, maybe. I pulled the ECU, and verified pins and connectors were in good condition, they were, pictures attached. I also wanted to verify the part number, because I might have found one for $180 plus shipping, which would be awesome if it works out.

Wanting to be sure I was looking at the CAN signals on the O-scope correctly, I took the set up, and moved to my 2018 Cruze Diesel, and hooked it up. I got solid and proper CAN signal patterns after a few adjustments, which might have been off for my initial reads on the truck, but I did note the truck is 11 bit 500 kbaud CAN, the Cruze is 29Bit 500 kbaud CAN, so I needed the 50ms setting on the Cruze, I had been using a much, much smaller time interval on the truck, and was basically looking too short a time to see a full pattern... That said, I moved back to truck, adjusted and found 200ms time setting worked about right, and the signals are more like what they should be, but they the are dirty (noisy) compared to the Cruze, they are often not properly symmetrical, as they should be, so there is still something off, but its not as bad as I thought. It is also interesting to see how much MORE data passes around on the 2018 Cruze compared to the 2009 Truck, the signal is constantly sending on the Cruze, the Truck has lots of idle time where no signals are being sent. I think that is normal..

I also pulled the ABS plug, and verified connections, and wires were in good condition, no smoking gun there either. With either the ECU or ABS disconnected the basic pattern of CAN signals seems unchanged. I had both the O-scope running and the wiTECH reading information in real time as I was disconnecting the ECU and ABS. The ECU can be disconnected at the large connector just inboard from the brake master cylinder, that disconnects the entire engine from the rest of the car, the ABS is on the other side of the master cylinder, and on mine under the clutch hydraulics.

With all that said, best I can tell all signs still point to the ECU. I know there is a reputation that the Cummins ECUs don't die, and since it is not technically dead, I guess it's technically not going to upset that reputation, but it is certainly shutting it's communication off to the CAN D bus. Now that I think of it, it's just the CAN D bus that is not getting the ECU, I assume the other modules are still communicating with the ECU, because if I disconnect the ECU, there are DTCs set, but when the CAN D goes down, and the wiTECH stops reading it, there are no DTCs on the other modules... The more I think of this the more it still looks like a network, and TIPM issue, because the only place the CAN D (DTC, OBD 2 port) goes is to the TIPM, from there it routes traffic on the CAN C (in my case ABS and ECU) and CAN B (everything else). Yet, I have replaced the TIPM already... ARRGH.. my 1962 Land Rover with a grand total of 2 fuses and hand crank start option is looking better and better all the time. These computerized vehicles are great when the WORK properly, but not so much when you get a fault that the system can't help you figure out.. which is what this saga is.

 

[MichaelOverfield]

So looking at the stupid cost of ECUs, I was able to find a used one at a reasonable price, ironically it's shipping from China (likely that is where they were originally made..). Since I have the Mopar Flash tool already, and know it works.. now all I need is the security four digit pin and I'll be able to swap ECU's and find out what is going on.. in worst case, if the used one has issues, it will be a "core" if I go with a more pricey unit later.. frankly I don't want to lose the one I have, despite the possible comms issues on the CAN D, it has never failed to run the engine reliably, so I'd rather have a spare on hand, especially when it's clear the price has gone full stupid, on this and the TIPM. Seems keeping the vehicle it might be wise to grab a spare, perhaps that is driving the costs as well.. Anyhow, picture of the one that is being sent, it is the correct part number and connectors, it appears to be a lower serial number, so older than my current ECU. This one was $180 plus $27 shipping and a transaction fee, for about $220 total.. compared to the best price found elsewhere at $700, and many prices in the $1200 range.. that is almost a steal, and it is about the average for the core charge for many units out there.