2023-05-25 EMC

and can in the injection clamp at 750 mm.
And now I've performed a threshold analysis on exactly this fail position.
And this is now in the presentation on the next slide.
And you can see we went up 1 dB microamp
And 102 dB microamps is category two.
The ECU has passed the test, and there was no DTC entry,
which is, I think, for this analysis, very interesting to check when the DTC starts.
The next level also no DTC with 104 dB microamps.
We have the DTC visible in the memory after the test.
So, we start to have some issues with the internal safety mechanism from the deviation of cell voltages and main current it's okay.
So, this test we would evaluate still as a pass. The next level has the same behaviors,
some more deviation, especially for cell voltage, and then at 106 dB microamps we have the failure at 384 megahertz,
some more disturbance visible on the measured signals in the CAN environment,
and of course it's caused by the DTC-TI internal error. Understood.
Which means for evaluation, I would say 103 dB is a clear pass, this setup. And at 104, 105,
if we make whatever kind of testing and keeping it half an hour on the most critical frequency,
I don't know if the error would then qualify. So it would also fail.
So this is, I would call it the grey zone here. So this test, the standard testing has passed,
but could be we can provoke a failure here. Understood, yeah. But nevertheless, so based on your conclusion, with 126 milliamps,
we are category 2, 200 is definitely category 3. So yeah, we are quite close to category 3,
So this is, to be honest, exactly what I was hoping to come closer to the Category 3 threshold.
Yeah, and I bet if we change the setup,
wiring and whatever, we might even be able to become better, maybe even worse.
Yeah, there's a lot of setup dependency as we found in the C1 test activities,
internal wiring harness and external one.
So I think, yeah, you should definitely test inside of the vehicle what happens in the real environment.
Yeah, we discussed about it. I think this would be kind of hard to test it inside the vehicle,
especially. That's why I wanted to have this very close approach on component level or detailed test. Yeah, okay.
But nevertheless, thanks for the results. Okay, is there anything else?
Yes, there's something. Jason, since you're on the line,
I would also like to share the test results from yesterday until afternoon with the two special batteries, because this is interesting.
We had the one nanofarad capacitor placed parallel to each NTC on the wiring harness,
manually placed, and we have a 10 nanofarad capacitor parallel to each NTC.
And we have tested only 750 millimeters position, because this was almost the most critical,
and starting with terminal 30 and CAN. And we can see with one nanofarad, it is already a bit better.
So we have passed the test at 106 dB, category 3.
And then we increased by 2 dB, and we got a failure.
On the other hand, we already see the DTC entered at category 3, and also at the 104 dB microns.
We already see the DTC. So if you compare this, there's not much change to the battery load.
To the battery without any cap, okay.
Right. But the 10 nanofarads was very interesting, because we could even go up to 110 dB microns.
Which corresponds to 360 milliamps, okay, way over category 3, okay.
Way over category 3, and differential testing and whatever, and we don't have the DTC. Okay.
The real question is if you need it inside the vehicle, which I cannot judge. But as we have already discussed,
the norm is very hard here. And other norms are, I think,
even below 100 milliamps injection current. But if there ever is an issue, this could be, besides the ferrite, another option.
Potential solution. This would mean, basically, that in order to, let's say, to place this additional caps to each NTC,
we need to change this flex board, basically. Yes.
New layout and so on, okay.
Yeah, and keeping in mind that an MLCC on a flex board is not mechanically robust, yeah.
So, I would not make a change like this just to get some improvement here, because the disadvantages can be quite bad.
Yeah, but I think there are caps which, let's say, do have this flex termination,
and they tend to be open motor also, yeah. Fail open or fail short, yeah, and I think we can find solution,
if it comes, indeed, to the necessity to place these caps, yeah.
Yeah, so this is just, yeah, investigation test, only PCI, nothing else.
So, I cannot judge what kind of effects there would be on other kinds of tests.
Okay. So, Wolfram, hi. Hi, sorry for the delay.
Don't worry. Maybe, Robert, you can switch back to slide 14 again. Yeah, of course.
So, yeah, for your information, Wolfram, just a second, Jack is joining.
Here is now, you can see an overview of the, yeah, close injected current, basically, on the BCI test.
We can see that we are, let's say, at 200 milliamps, we are closer to category three, but nevertheless,
for the tests with smaller current, yeah, 158 and 178 milliamps, we do get a DTC, but a circuit breaker does not open.
What does TI internal error mean? It's just a DTC.
This is the DTC that takes place. It's the feature of the TI. It compares internally cell voltage from different channel.
So, this is for functional safety, the check, and when the DTC is triggered,
sometimes there's an error, there's a mismatch, but there's a timing set behind that.
So, some margin for errors is allowed for a certain time.
Okay. The only solution is with now putting a parallel capacitor, right?
So, basically, we have now the status where, basically, exactly at the threshold of category three,
we do get a fail, but we are quite close to category three, yeah? So, the overview shows with,
let's say, 178 milliamps or 105 decibels, we get a pass, yeah? So, the circuit breaker does not open, yeah?
On the other hand, Robert has tested on the next slide,
has tested this possible solution where we placed additional ceramic caps to the NTCs,
yeah? And as you can see on the left side with one nano cap,
we get almost the same results as without having any caps in parallel to the NTCs, but with 10 nano,
we get a pass which is way over category three. So, we are shooting there with even 110 decibels, yeah? So,
this corresponds to over 300 milliamps of current, and we get no DTC either.
So, this means that the 10 nano caps work indeed. Yeah, but would be a change, of course, and I don't,
so it's up to EBE to comment. I think there's not a lot of space in the flex PCB and so on.
Possible solution path, yeah, if required.
Is it maybe necessary to have here risk assessment just for us to argument with our EMC specialists that we not or don't need this change here,
that we might say we can have different behavior in measuring the temperature of the cells
or maybe we have much more effort in producing the flex PCB,
just that we have it in kind of a written form to show it. I was thinking to do this or to do this together,
this risk assessment, because we know basically what this hardware change would mean in the current development stage,
yeah. I mean, just because if we are now presenting this to the EMC department, they will, of course, say, let's make this change.
Sure, but you have to countermeasure that. You will have mechanical problems,
and the other thing is that you have, I don't know from another project, they said they need to test it.
Yeah, so we need additional, I know, vibration tests, lifetime tests, endurance tests,
and so on. They need to be able to electrically test the capacitors.
At the end of line, I know. I know the topic, I know the project you're referring to.
And that is a problem as far as I've heard. Yeah, but there is, as I said, if we really,
let's say, are obligated to implement this change, there is a solution.
So, we might use caps which have a fail short failing type and do have a flexible termination,
so they will definitely withstand mechanical stress way better than normal caps,
and by the fact that they are failing short, you are able to detect this at the end of line testing,
because they will short basically the NTC, so you will not get any temperature measurement.
So, basically, in this way, you will be able to ensure that each battery which leaves the production line is healthy,
let's say. But nevertheless, we know what the hardware change brings up, so additional testing,
yeah, which currently, and since we are basically in the D sample, we cannot allow us any further hardware changes.
So, my target is here together with our EMC department from the system level to get argumentation or to get a deviation acceptance on the threshold,
so basically on the slide 14. I will definitely show this, yeah, but as Karina said,
we will definitely basically say or, yeah, additionally write what are the pros and the cons, yeah,
so what are the advantages and disadvantages by improving this, and basically in this time now,
yeah, close to series start, we are not able to implement it, even if we are, if we want it,
so this will definitely be a change for a next sample phase, which I don't think currently is planned.
Okay, any further questions for the EMC results? Yes. Of course. Yes.
The one nano is a close, oh, wait a second, this one is wrong, I did not, so sorry,
I need to update this one. So, this is 106 db microns, this is category three,
this has passed the test, so this way it's correct.
So, with one nano, we can just narrowly pass the test, but it's very close pass.
Exactly, so if we put a capacitor, I would say 10 nano farad is working well,
and one nano farad does not make enough difference to justify the change.
Okay, they didn't fully understand, but if you have further questions, we can also continue discussion in another round.
This is 158, so now it should be correct here.
Late afternoon yesterday.
Could you please switch again to slide number 14? Yes.
Just to make a screenshot to have the documentation here in the next one, LOP. Thanks.
You're welcome. We will upload it also to Panama later.
Please do it as fast as possible, because I will try to summarize all the information to get a status of the current sample,
and present this to our colleague as soon as possible. Thank you.
I have no further questions on this topic. Okay.
I will stop sharing my screen.
I have no further questions on the BCI topic, but Jason, maybe a question to you,
and I see also Jack in the line. Since Tuesday, since our video meeting,
was everything fine so far? Have you started testing with the stripline and so on? Is everything okay?
Another meeting for the other test setups?
Then please send out the meeting invitation ahead of time.
Keep in mind, next week, Monday, June 29, it's a public holiday in Germany.
If we have it on Friday, tomorrow, it should be fine, but please send out the invitation. I hope it's okay.
Is it okay for the BMW participants? Just a sec. Which time exactly again?
8.30 tomorrow morning. 8.30 tomorrow morning. Yeah, we can organize it.
If you don't have any topics for a problem management meeting, we can schedule it for 8.30.
Yeah. Robert, one question though on the BCI topic. Have you made any pictures with the
placed or with the additional caps to the NTCs?
Do you have the possibility to make a picture or something and provide it to me? I got some pictures from Jason or Tina, I think.
Upload them together with the MC report.
I can put it inside, but I cannot produce it because it's in between the cells now.
I understand. It's fine. But nevertheless, upload them.
It's better to have a picture which basically explains what's the exact hardware change.
Yeah, understood.
Okay. So, according to the agenda, we have discussed the EMC topic.
Thanks again for performing all these tests, Robert. Yeah, my hope now is to, as I said,
based on the threshold analysis or very detailed threshold analysis,
that we get this or an acceptance on the failed exactly category three just for the 750 millimeters use case. Yeah, so.
Okay. State and soft implementation. We have here some tickets, plausibility check.
I think we have the ticket for it. Karina, as we discussed yesterday, we wanted to check again this Aurix topic.
Let us see. So, status here on this one, it's not loading quite exactly.
I somehow can't see the status on this ticket.
Yeah, me neither. You can just see here in the solution that it's not done yet.
Okay. So, still an open topic.
Open, let's say, different path for 12-volt supply. Integrated, fertig. Okay.
So, this topic has been retest only possible with C2. I think this has already been done.
Update for some power for how to afford protection. I think, yeah, I think this ticket it's closed. Okay.
So, let me document this. Test this. Reaction check.
Softy ticket has been solved. Implementation done. And tested. Okay.
Minus 40 with the Aurix. It's still ongoing. I think groundless is also ongoing.
Let me check. Yeah, it's still not, it's still open. Indeed.
What's the latest move? It's just important.
Yeah, here it's just important that we have it before starting with the 48-volt test and the PV, but they are delayed nevertheless. So,
we still have a bit of time. Okay. So, let me move this up so that we don't search it again.
So, groundless detection. Okay.
They planned for the i410 software release. But I think it's the third software release or something like that. Okay.
Maybe when you go back to the driver ticket, we can have a quick look.
It's written here. Move to 410. But withitutional release.
Oh, institutional release. Okay. Yeah.
Or maybe let me add this with institutional release. Okay. And i410 software,
Okay, institutional release. Okay, interesting topic. Okay.
So, groundless detection. E07, I think. Okay. Yeah, that's fine.
This was a hardware issue. This has been closed here from the software.
Unplausible CB temperature. Okay, debouncing time.
We need something here. This has also been fixed. Implausible.
Already for the 390. Exactly. CB implausible. Okay. Yeah.
And the time of the table for the discussion.
But I think it's still an ongoing topic in the German meeting.
Yeah, but the ticket itself is closed. Yeah, it's closed.
Okay. So, from our hardware side, we don't need to do anything, I would say.
Okay. Debounce hours increased to two minutes. Okay.
So, basically, still open are these two tickets. And I don't know if the...
And maybe also the current plausibility. Exactly.
Who is here in charge from hardware side? How is he called? Hans Luther.
Hans Luther, yeah. Yeah, yeah. Okay.
He's the one who creates some test setups and to...
Yeah, for the colleagues to the software that they can test it.
And yeah, he's involved. Okay. All right.
Just if we have some questions again, then I will directly contact him.