-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
What exactly is the 1146 Alert?
- Thread starter gregd
- Start date
Thanks. Any instructions for removing the PEM? Sounds like a big job. I'm a electronics engineer turned programmer so I don't have that sort of attention span any more.:-( Also been a long time since I've soldered anything bigger than SMC. That ticket might be an investment
it seems a shame for it to be a sticky and still be overlooked. The thread is here
DIY: Roadster 2.5 PEM cleaning - step by step
There are a lot of other threads regarding the fan plugs and alternative pins. The search function on this part of TMC works incredibly well due to the low S/N ratio.
I would probably start here
#1146 DMC Motor Fan Problem?
warden_clyffe
Member
both errors(1144/46) can also come up when you accelerate heavy - then have your ac coolant and/or pump checked
Interesting, but I don't expect that's what's happening here. The errors are only 1146. I haven't had any 1144s since August, when the Service Center over-temped the HVAC compressor by putting the car in the 100+F sun (so the battery was nice and toasty too) and then charging it at 70 amps, and the errors occur both while driving and charging (though mostly its while driving now that the ambient temps have dropped to the 60's).
Hi folks,
So, time for an update. I brought the car in for a deep analysis of what was going on, and was prepared to follow whatever conclusions they came to. Tests were run, specifically a contact "pull test" to determine that the connector pins were making positive connection, voltage tests at the motor, and resistance tests of the motor itself, and probably a few things I didn't write down. All tests verified that everything was within spec, and operating as expected. A test drive while in diagnostic mode showed the same.
The factory folks have apparently had some past experience with devices similar to the OVMSv3 that I have (no model specifics were given), where they have caused similar issues to what I have seen: 1146 errors and similar, when no such error was warranted. So, they pronounced the car as perfectly healthy, and blame the module.
Now, before you a) throw out your OVMS, b) yell at Mark, or c) yell and scream that a passive monitor such as the OVMS can't be the cause, don't shoot the messenger. What I reported above is what they told me.
But the story doesn't end there.
I got home with the car this afternoon, got it all charged back up and happy, and grabbed the logs again. Aaaand, guess what... The alerts are now not happening, or at least, they did not occur during their test drive or my drive home, where they did occur during drives just prior to the visit. My own conclusion is that this proves the connector was probably the culprit, and that the removal and repeated scraping of the contacts during the "pull test" cleaned off some little bit of oxidation, and all is right with the world again. At least until they oxidize again or the next annual service has to fiddle with it. I may wait 2 years for the next "annual" just to forestall the inevitable. Oddly, the errors didn't occur yesterday morning on the drive into the SC, and that was before they fiddled with things, so that shoots down my conclusion. The term "perplexed" doesn't do this situation justice.
What I am going to do next is to connect back with the SC and (through them) the factory experts, and try to wrestle this one more layer down. The objective here is to either clear the good name of the OVMSv3, and to confirm with them that, yeah, it was the connector all along. Or if, on the other hand, they do offer up some evidence of how the OVMS module could possibly be the actual cause, I want to know what it could be doing so that we can fix the module to not do that anymore.
I will probably wait a few days before sending off that email, to be sure the errors really have gone away, and not be a temporary bit of sanity in an otherwise perplexing world.
So, the story continues...
So, time for an update. I brought the car in for a deep analysis of what was going on, and was prepared to follow whatever conclusions they came to. Tests were run, specifically a contact "pull test" to determine that the connector pins were making positive connection, voltage tests at the motor, and resistance tests of the motor itself, and probably a few things I didn't write down. All tests verified that everything was within spec, and operating as expected. A test drive while in diagnostic mode showed the same.
The factory folks have apparently had some past experience with devices similar to the OVMSv3 that I have (no model specifics were given), where they have caused similar issues to what I have seen: 1146 errors and similar, when no such error was warranted. So, they pronounced the car as perfectly healthy, and blame the module.
Now, before you a) throw out your OVMS, b) yell at Mark, or c) yell and scream that a passive monitor such as the OVMS can't be the cause, don't shoot the messenger. What I reported above is what they told me.
But the story doesn't end there.
I got home with the car this afternoon, got it all charged back up and happy, and grabbed the logs again. Aaaand, guess what... The alerts are now not happening, or at least, they did not occur during their test drive or my drive home, where they did occur during drives just prior to the visit. My own conclusion is that this proves the connector was probably the culprit, and that the removal and repeated scraping of the contacts during the "pull test" cleaned off some little bit of oxidation, and all is right with the world again. At least until they oxidize again or the next annual service has to fiddle with it. I may wait 2 years for the next "annual" just to forestall the inevitable. Oddly, the errors didn't occur yesterday morning on the drive into the SC, and that was before they fiddled with things, so that shoots down my conclusion. The term "perplexed" doesn't do this situation justice.
What I am going to do next is to connect back with the SC and (through them) the factory experts, and try to wrestle this one more layer down. The objective here is to either clear the good name of the OVMSv3, and to confirm with them that, yeah, it was the connector all along. Or if, on the other hand, they do offer up some evidence of how the OVMS module could possibly be the actual cause, I want to know what it could be doing so that we can fix the module to not do that anymore.
I will probably wait a few days before sending off that email, to be sure the errors really have gone away, and not be a temporary bit of sanity in an otherwise perplexing world.
So, the story continues...
Probably because without OVMS connected, as this is a debug only error, you don't see it. Still happening, but invisible. Apart from Tattler, is there any other device like OVMS?
Definitive test would be, on a car that shows this in VDS debug mode, to disconnect OVMS then see if the fault code shows. I'd bet money on it still showing even with OVMS disconnected.
Definitive test would be, on a car that shows this in VDS debug mode, to disconnect OVMS then see if the fault code shows. I'd bet money on it still showing even with OVMS disconnected.
Hi Mark,
Yep, I did that test. Disconnected the OVMSv3, drove, charged, pulled the logs. The 1146 errors were still present. I informed Tesla about this test, but they apparently ignored it.
Playing engineer, the only connection I can fathom between the errors and the OVMS is if somehow a command from OVMS to, say, lock the car collided with (delayed, corrupted, whatever) another message on the bus, perhaps that might have put some other component on the bus in a "bad" state. Once in that state, the "badness" would persist until some other event reset it, hence removing the OVMS wouldn't change things. The problem with this scenario is that the 1146 error is the result of some sort of measurement the PEM is making on the dynamic voltage/current profile of the fan motor, something that is not done via that CAN bus. In fact, nothing regarding the drive train is done on that CAN bus. Blaming the OVMS is clearly blaming the messenger.
Since writing the update a month ago, and recovering the GPS clock so that the log is sane, I've continued to NOT see the 1146 errors in the log. The ambient temps have dropped significantly since late summer when they started, reducing the demand on the fan, so this may be somewhat seasonal. My hunch at this point leans towards the fan motor and its brushes. The whole system was stressed severely during the drive home after the annual service (they didn't reattach the PEM Fan air duct properly), which would have stressed the motor. DC motor commutator brushes tend to re-seat and reform themselves over time, and if so, perhaps that's what "healed" the system last month. {shrug} It's my only theory at this time, and not worth $1,000 to replace the motor to find out. Besides, they did some tests on the motor, and pronounced it in good health.
I sent the email off to the SC and the Roadster NA service email address yesterday, detailing all this. Hoping for a reply at some point, though I'm guessing I will have to keep pushing on them.
OVMS doesn't send anything that is not also sent by the VDS. It basically acts as a 'remote' VDS. This is on the instrumentation CAN bus, with the PEM (and its fans) on a completely separate bus (that we don't even initialise, or connect to, at the moment).
Yep. Which is why I turned off these DMC notification messages at the server level.
Exactly. Perhaps "conclusion" is a better word than "hunch". Still not believing it's worth spending $1,000 on one that's otherwise working just fine, if the only net effect is removing the alerts. I can ignore them if I know that it's ok to do so.See post #11
For anyone scanning this in the future, in the course of investigating further "fan troubles" (See Think I need a new PEM Fan... Anything else to check?), I can report the answer to this thread.
1. There are two independent "drivers" within the PEM, supplying regulated power to the blower assembly under the rear of the car. One driver is for what was the original PEM cooling fan, the other for the traction motor fan. The "dual fan upgrade" (which my car has) combines both into one assembly, with both drivers wired in parallel, and driven by software to act together. 1144 alerts are specifically related to the PEM-side of the circuit (pins 1 & 3 on the PEM connector), and 1146 to the motor side (pins 2 & 4).
2. The 1144 errors get reported to the cabin, the 1146 generally do not (but they're in the logs if you pull them)
3. The alerts can be for a variety of things, but generally relate to a sensed current that is different than expected, for the voltage that is being sourced. In the specific case of December's 1146 errors, they were an under-current sensed, and in hind sight, were likely an early warning of the failure of pin #2 in that connector, which has since totally failed. These were decoded by a software tool that Gruber Motors has developed.
4. Please see the rest of the above referenced thread for the thrilling saga and conclusion to what was really going on here. You won't want to miss it!
1. There are two independent "drivers" within the PEM, supplying regulated power to the blower assembly under the rear of the car. One driver is for what was the original PEM cooling fan, the other for the traction motor fan. The "dual fan upgrade" (which my car has) combines both into one assembly, with both drivers wired in parallel, and driven by software to act together. 1144 alerts are specifically related to the PEM-side of the circuit (pins 1 & 3 on the PEM connector), and 1146 to the motor side (pins 2 & 4).
2. The 1144 errors get reported to the cabin, the 1146 generally do not (but they're in the logs if you pull them)
3. The alerts can be for a variety of things, but generally relate to a sensed current that is different than expected, for the voltage that is being sourced. In the specific case of December's 1146 errors, they were an under-current sensed, and in hind sight, were likely an early warning of the failure of pin #2 in that connector, which has since totally failed. These were decoded by a software tool that Gruber Motors has developed.
4. Please see the rest of the above referenced thread for the thrilling saga and conclusion to what was really going on here. You won't want to miss it!
Bunnak
Member
No, long story, still evolving. The easy answer is to replace the connector and cable at the underside of the PEM, but given that this keeps happening, and since I otherwise have the time, I'm looking for a deeper cause. The thread I referenced above gives the long sorted details, including discussions I had with Gruber and others.
It’s brave of you, mine just had the connector fault and I was just happy changing the plug fixed it and I can hear my fans. I hawk the temp in vitals when I drive now and notice driving in range the temps are way higher, uncomfortably/unbelievably high. I put it in normal mode and I hear the fans step up and watch the temperatures drop to less sphincter clenching numbers. I always wondered if the development of the PEM cooling system was at some point meant to be part of the battery cooling. For a long time the way to cool a computer is water cooling. In the early 90’s before that was around I ran mine in an apartment fridge.the PEM seems to be passively cooled, and by that I mean using ambient temperatures rather than active cooling. The whole design seems restrictive and inefficient compared to the battery cooling. Digging deeper is probably a good way of finding out what they were thinking. I often thought about additional cooling but that would probably slow the other fans down making it more problematic than it needs to be. Looking at my temps and fan speeds there is obviously different algorithms written for range, performance and normal rather than one written dependent on just the thermistor feedback.
I believe part of using range mode is a somewhat wider range for temps, saving energy in the process. The PEM has always been cooled from ambient air, and that seems to be just fine under most circumstances, even here in Northern California's Central Valley. It certainly runs warm, but stays at the top of the Blue zone in Standard mode. I've not looked at it when in Range mode. Probably should!
It would probably be fairly easy to rig up a small set of coils (salvage from an ICE heater core?) on the air path from the fan to the PEM, and plumb it into the battery coolant loop. Not sure it'd help much, and I certainly wouldn't want it to heat (or freeze in winter) the battery as a side effect. Running the A/C to do the cooling is a bad idea, too. Maybe a swamp cooler? {shrug} I'm just looking to restore the regular air flow. I think that is best.
I was seeing 103 degrees celcius on the motor and 50-54 degrees on the PEM, in normal mode I have 40-42 PEM 84-88 Motor. I will pull the motors and seal them better if/when I get the level 3 charger