Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Roadster 2.5 - Vin 1277 Failure - ID981 State Transition Fault

This site may earn commission on affiliate links.
Hello,
Yesterday I got up to find the vehicle would attempt to charge, then stop with an ID 981 "State Transition Fault" after about 5 seconds. Then about 10 seconds later, it would try again.

The charge screen would show something ridiculously small, like 37 volts, on the HPC. Long before it showed any current trying to charge.

I also tried the 110 V plug, it don't do anything. The ring won't even change to blue when I plug it in.

When I unplug it, close the door and turn the car on, it follows with a ID 1138, "Voltage on charge port line 2 during drive".

At that stage, if I press Drive, it also gives me an error ID 971 "Phase B desat fault". (what is "desat"?)

So...
I assume charge port line 2 is the same as Phase B. I guess line2/phase B is also the single phase used for the 110V, which is why it won't even give me a blue ring.

About 10 months ago, Gruber Motors did a rebuild of the PEM. New IGBJT insulators, new capacitors, new connector blocks, etc... It was working fine up to this point.

I've got the cover off the PEM, and verified that what Gruber said they rebuild was actually done. I was thinking of examining the contactors so see if something failed, or backtracking to the driver circuits that open/close them. But if anyone has any recomendations before doing that, I'd appreciate any advice.

(I'm an EE, and last year I had a precharge failure, so I took them PEM apart and tested the precharge circuit transistor IRF540 thinking it might be the cause... It wasn't so I sent the car to Gruber for the overhaul)

The car is now just parked and the system inhibited.

Any advice would be appreciated.

-Scott
 
I am sorry to hear about the troubles with your car. If you by inhibited mean inhibited the APS, you may in addition consider pulling the service plug to preserve ESS charge and removing the auxiliary battery fuse to avoid draining that also.

I notice that the message that you attributed ID 971 is something different from what is said here: Tesla Roadster VDS Error Code Descriptions | Gruber Motors

Anyway, according to my understanding Phase B would refer to one of the phases of the motor. It is driven by one of the megapole boards in the PEM and desat fault is short for desaturation fault which is a failure condition in driving the IGBTs on the megapole board.
 
I am sorry to hear about the troubles with your car. If you by inhibited mean inhibited the APS, you may in addition consider pulling the service plug to preserve ESS charge and removing the auxiliary battery fuse to avoid draining that also.

I notice that the message that you attributed ID 971 is something different from what is said here: Tesla Roadster VDS Error Code Descriptions | Gruber Motors

Sorry.
The last one is ID 910. PhaseB High Side Desat Fault. I thought I proofed this post.

Yes. APS inhibited. I've seen the interlock on the battery and will pull that if the car is going to sit too long.

Anyway, according to my understanding Phase B would refer to one of the phases of the motor. It is driven by one of the megapole boards in the PEM and desat fault is short for desaturation fault which is a failure condition in driving the IGBTs on the megapole board.

Useful information.

I have since measured the 150A fuses with a DVM, and see that they are both completely open (over 5 meg while in circuit). Both of them.

I could replace them, but if it's the IGBT's, then the fuses would go again...

It's not beyond my ability to pull those boards and replace the transistors either. However if that's not the problem, it will eat more 150A fuses. yes? At over $100 each, I'd prefer to be sure it's fixed.
 
I suppose the blown fuses can be interpreted as a problem in the internal charger, but it is difficult to tell. When I had the insulation issue with mine (Roadster PEM failure), the first indication of a problem was that the car refused to charge (red light and complaints) despite that the charger in the PEM later appeared to be fine. I pulled the logs and got some help here to arrive at a diagnosis credible enough to attempt repair. Hopefully someone here more experienced than I can chime in.
 
Fuses are much cheaper, I got a pair for about $50 delivered on e-bay. Mine blew just plugging into a bad destination charger. I think for the cost you should start with the fuses, they are there to charge the battery not power the mega pole boards. That being said I’m only going on my limited experience and would listen to much more educated advise when they chime in.
 
I suppose the blown fuses can be interpreted as a problem in the internal charger, but it is difficult to tell. When I had the insulation issue with mine (Roadster PEM failure), the first indication of a problem was that the car refused to charge (red light and complaints) despite that the charger in the PEM later appeared to be fine. I pulled the logs and got some help here to arrive at a diagnosis credible enough to attempt repair. Hopefully someone here more experienced than I can chime in.

I've got the logs, that's the easy part :)

Who did you get to help with the diagnosis?
 
Fuses are much cheaper, I got a pair for about $50 delivered on e-bay. Mine blew just plugging into a bad destination charger. I think for the cost you should start with the fuses, they are there to charge the battery not power the mega pole boards. That being said I’m only going on my limited experience and would listen to much more educated advise when they chime in.
I found some there , more than $50, but at least not $128 each.

I'll get some advice before installing them. Logically they should allow it to charge without engaging the motor circuits.

But it appears that the system died during an overnight charge because the battery wasn't filled all the way, but also wasn't at the level from my drive that afternoon. So I'm not sure how the IGBT driver board relates to those fuses (they are inline with the charge port lines, not with the battery)
 
Did it drive with those error codes. I forget what my codes were but I know it was the fuses because as soon as it would start to build voltage it faulted. It took me a while to work out but it threw some of the bad PEM codes as well, it just took new fuses.
 
B1A36A8F-F50E-4F06-98FA-E58BCCF5C0C5.png
 
Did it drive with those error codes. I forget what my codes were but I know it was the fuses because as soon as it would start to build voltage it faulted. It took me a while to work out but it threw some of the bad PEM codes as well, it just took new fuses.
On Friday it drove fine. No error codes at all.

The overnight charge cycle started. When I got up Saturday, I heard the HPC kicking on and off every 20 seconds or so. I saw that is had partially recharged, but didn't fill up. So overnight something failed during the charge.
 
I've got the logs, that's the easy part :)

Who did you get to help with the diagnosis?
I used the vmsparser to extract the messages and included them in the post that I referred to. The help I got is in the posts from people on this forum following in that thread. In addition I got a couple of pm's. I see that you already received some advice. The big question for me was wether or not to deal with it myself or leave it to Tesla. I opted for the first alternative when I became convinced that I had the well known insulation issue (and failed to initiate a satisfactory technical dialogue with the SC). It is not clear for me what is the best action for you. It is a tricky risk assessment, I guess. Will pm some more info.
 
Update:
The car had essentially a catastrophic failure of the power IGBT's in one of the driver phase boards. They are basically one of three phases of an H-Bridge driver, two "half H-Bridges" on each board. They have 7 transistors all in parallel (14 total, upper and lower), 28 one each board . Since they are in parallel, we couldn't measure which one had failed while in circuit, but it was prudent to replace all of them on all three boards,

Carl Medlock of Medlock and Sons fixed it for me. I shipped it off, and followed it up to his shop for some education.

I am very impressed with the attention to detail and work ethic of him and his team.

There is an issue with the retaining nut in the motor connection block. This was from Gruber.

A year earlier, Gruber Motors has replaced the power and motor connection blocks with new 3-D printed blocks. When I got the PEM back, I noticed that I could not secure the lower screws on the motor connection cover. The nuts in the block were missing. (I could see one on the right, but it would never grab the screw)

MotorBlock1.jpg


I decided that I should investigate this, as a loose nut rolling around inside the PEM was not a good situation to have. I opened the PEM, to examine the block.
The right side one was there, so I attempted to push it out from the outside, and it is visible here.
BackRight.jpg


I thought about where the other one might be. Well it was shipped standing on end, so I looked there. And it was tucked away in the corner.
Bolt.jpg


If it has shorted out the power rails, that would have been a disaster.

I had to lift the circuit board and mounting frame and was able to hold the nut in place using a bent scribe, while screwing it in from the front.
hook.jpg

And I had to insert it on the left and follow the same process...
BackLeft1.jpg


That's fine for this time, but what about next time the PEM comes out... I don't want to have to open it up and struggle hold it in place so I can tighten the motor block connection cover.
Instead, I globbed some epoxy behind the nuts to hold them in place.

Car works great now.
 
Last edited:
There is an issue with the retaining nut in the motor connection block. This was from Gruber.
@scotty2541 The motor output block was designed with the jack nuts to remain glued in their respective hexagon-shaped channels. Given sufficient external force, as you noted on the right jack nut you pushed in, will break the adhesive bonds between the jack nut and plastic channel. In the future, I would recommend.
  1. Not “[attempting] to push [the jack nuts] out from the outside”. With sufficient force, anything can be broken loose.
  2. Using your full weight when inserting/removing screws smaller than a few millimeters in diameter
  3. Use the right bit. The wrong Philips bit can cause the bit to cam out of the screw head on the motor cable safety cover. You wouldn’t believe how many Roadsters come in with those milled slots smeared like Playdoh. There are at least seven different types of Cruciform shaped screw heads that come in a variety of sizes.
Good idea epoxying the jack nuts back into their respective channels!

Lastly, given the numerous PEMs I see on a near weekly basis, the majority of them have cracked or unusable motor output blocks. Therefore, a design improvement was necessary rather than reprinting the original Tesla designed motor output block. The motor output block you have installed now relies on pressing pem nuts into the bus bars instead to ensure a longer lasting part with every annual preventive maintenance and PEM cleaning. We will continue to improve this vital part to keep PEMs functioning, and I appreciate you pointing out an unintended use case scenario.
 
@scotty2541 The motor output block was designed with the jack nuts to remain glued in their respective hexagon-shaped channels. Given sufficient external force, as you noted on the right jack nut you pushed in, will break the adhesive bonds between the jack nut and plastic channel. In the future, I would recommend.
  1. Not “[attempting] to push [the jack nuts] out from the outside”. With sufficient force, anything can be broken loose.
  2. Using your full weight when inserting/removing screws smaller than a few millimeters in diameter
  3. Use the right bit. The wrong Philips bit can cause the bit to cam out of the screw head on the motor cable safety cover. You wouldn’t believe how many Roadsters come in with those milled slots smeared like Playdoh. There are at least seven different types of Cruciform shaped screw heads that come in a variety of sizes.
Good idea epoxying the jack nuts back into their respective channels!

Lastly, given the numerous PEMs I see on a near weekly basis, the majority of them have cracked or unusable motor output blocks. Therefore, a design improvement was necessary rather than reprinting the original Tesla designed motor output block. The motor output block you have installed now relies on pressing pem nuts into the bus bars instead to ensure a longer lasting part with every annual preventive maintenance and PEM cleaning. We will continue to improve this vital part to keep PEMs functioning, and I appreciate you pointing out an unintended use case scenario.
Peter,
The block is as a unit is fine. I thinks it's good value add.

In this case, I'm not sure why but both nuts were extremely loose. There was virtually no force needed. When I opened the cover I could slide the right one back in easily, and then the screw pushed it right back out again. Just not all the way (I could have slid it all the way out if I wanted to). The left one wasn't in the hole at all.

If it wasn't too large originally, perhaps something happened with age/heat that allowed it to become loose. The extrusion material is very slick, so I think it's wise to use something (epoxy, glyptal, hot glue, Loctite,...) to hold it in place. I hadn't touched the PEM since I returned the car for you to install those blocks.
 
lol C, but it was sitting on the last blue block, I usually only see it on the second to last unless I’m in range mode. My problem was I looked at the OVMS and saw the actual temperature, without that knowledge I would have been in peaceful ignorant bliss.