Project Lithium Soother Test

Haven't seen it yet, but it might be worth a try.

 

1: I understand you didn't '@' me on this one, but I'm still responding.
2: I agree, you are certainly allowed to post in your own thread.

In chronological order:

...

I think my bigger issue is that you failed to ever address any scientific theory I stated anywhere. Instead you posted a litany of insults (see above), and disregarded clear-as-day facts. Typically in a scientific discussion we discuss scientific things, rather than petty name calling (see above).

These are all readily apparent facts. I haven't hidden anything you listed here. I've also already addressed the "mudder is a competitor" angle at least four times in the past four days. I'm curious how thoroughly you read my threads, because you've never responded directly to anything scientific I've said.

Note that later in your post you contradict the statement you made above.

More proof you haven't read my posts, as I've already stated the following twice:
1: the correct word for written falsehoods is 'libel', not 'slander', and;
2: libel is false written statements, and;
3: I haven't made any false statements.

Given that your thread is titled "Project Lithium Soother Test", I would propose that it was appropriate to post my test results here. I do find it odd that you waited until now to assert that you do not think I posted appropriately... quite a roller coaster over the past few days and yet this is the first I'm hearing that you didn't actually want me posting in here.

Regardless, I respect this forum's moderators if they don't agree my posts are appropriate here.
In that case, I would ask that the moderators:
-move all posts including and after #67 to a new thread titled: Project Lithium's "Signal Soother" is Ill-Conceived and DANGEROUS!
-remove any posts that don't contain a qualified scientific statement.

I actually like that idea and have created said thread here. This will greatly increase the signal-to-noise ratio. Great idea, @AzusaPrius!

Your comments were off-topic, contained profanity, and you actually had multiple comments removed because even after the first time you still didn't get the message.

Should I include this in the above list?

 

I find that unlikely, unless you and/or I get banned.
Another option is you can add me to your 'ignore' list. Since you'll no longer see my posts, that'll probably save us both time, while allowing me to continue sharing my scientific findings. I personally endorse this option.

I disagree with this statement, but you are entitled to your opinion.

It's possible the profanity was posted by @black_jmyntrn. I will need to look at the screen shots I took prior to deletion, but can't do this until I get back from my vacation next week. Regardless of whether or not yours was the post with profanity, both posts were off-topic ad-hominem attacks. I trust that PriusChat's moderators evaluated my report fairly prior to deleting your posts.

Again I have the screen shots on my desktop at home, and will reference them next week when I get back from vacation. You are probably correct that you didn't cuss in my thread (and so @black_jmyntrn did), but regardless your post was out of line, and apparently whichever moderator honored my request agreed. And even after that, y'all continued posting off-topic name calling, hence the further moderation removal requests.

It would certainly distill the technical facts.
However, I don't mind preserving your comments in my own thread, and have asked the moderators to do so for posterity. Once the comments are in my thread, I suppose it's up to me to determine what is out-of-bounds. Note that I've already contacted the moderators regarding this bulk post relocation request, so no action needed on your end, unless you just want to overload the moderators more. I don't know them yet, but huge thanks to the moderators for whatever work they end up deciding is appropriate.

I do not contest this fact:
Yes, I asked the moderators to remove the off-top, ad hominem attacks you and black posted in my LiBSU thread.

I wouldn't need to retype anything; I keep archives of pretty much everything I post on the internet.

In due time, if these posts are ultimately removed from said thread. Of course, I will still link to your thread from my own, since I am ultimately contesting the validity of claims you made in your original post.

 

Posted without further comment.

Posted without further comment.

 

I've read through this whole thread and just got up to speed on the latest information tonight.

Thank you to @mudder for bringing these items to light and data to the discussion.

Thank you to the @ChapmanF who summarizes mudder's video. Post #142

As an independent hybrid tech who has had to remove failed Nexcell system (and possibly more of them in the future) I am quite concerned about the Theory laid out in post #82.

Can an existing smoother user like @AzusaPrius please take a moment to confirm/debunk this theory ASAP? (See below)

I tell techs to always trust but verify when it comes to thinking that a HV system has been discharged. But even I have been guilty of skipping this step.

Please someone check ASAP and report back.

The theory...
Theory A: Signal Soother violates service plug safety element. This presents an electrocution hazard even when the service plug is removed!
A1: Remove service plug
A2: Remove battery cover
A3: Measure voltage from HVDC+ to HVDC-. You should measure a non-referenced (i.e. floating) voltage, which typically manifests by the voltage meter slowly drifting around until it reaches 0 volts. However, I propose you'll read close to the actual pack voltage... dangerous!

 

Given that nobody on Team Jack responded to my several proposed test methods yet, I will perform this test as I see fit in the lab when I get home in a few days. After I perform this test, I will then perform my other proposal (also in the lab), which will almost certainly destroy some combination of parts near B4+ & B5+.

 

As an EV tech, I have certainly forgotten to replace the service plug before turning at the car at times. But it's never had a risk of blowing out battery components.
If/when I come across any other cars with Nexcell and smoothers I will certainly double/triple check that the plug is in.

 

If my theory is correct, then the only component at risk is the Signal Soother itself... no loss if you blow it up, as you absolutely shouldn't let a customer leave with the SS still installed. So no risk to pulling the service plug on a customer's car... worst case the SS lets out magic smoke when the pre-charge contactor fires at keyOn.

Meh... just pull the plug out and turn the keyOn... send the SS to heaven.

 

Ok. That sounds better. I don't mind blowing up a soother.
Just don't want to damage the car.

 

There is a very slight risk that the OEM BMS wire harness to the cells could overheat during the event. However, it is much more likely that only the Signal Soother's opamps/resistors/diodes will erupt into a brief flash, which should remain contained inside the heat shrink tube wrapped around the SS PCB; followed by a small volume of smoke.

The Prius' precharge routine should quickly detect that the HVDC bus voltage is collapsing (because you're attempting to pull massive current through the SS PCB), which should prevent the primary contactor from firing.

Worst case, you'd have to replace customer's BMS harness, too. I estimate this outcome is very unlikely.

 

off topic but just curious how the Tesla Powerwall defect report you made to the Tesla community a few months ago is progressing?

 

BAM! 2nd thread hi jack attempt.

 

Here are my Signal Soother benchtop test results:

As I predicted, the Signal Soother was damaged when the service disconnect was removed while a load was placed on the HVDC bus.
This test verifies both theories I proposed previously:
1: Yes, the Signal Soother has no galvanic isolation between the two pack segments when the service disconnector is removed. Therefore, the service technician is subjected to the entire HVDC pack voltage even when the service disconnect is removed.
2: Yes, the Signal Soother is damaged if any load is placed on the HVDC bus with the service disconnect removed.

IIRC, the service disconnect also contains the HVDC fuse, so then if that blows out while driving, that condition would also destroy the Signal Soother.

Fortunately, the failure that occurred when I removed the service disconnect was entirely self contained...
...unfortunately, the sparking (across R41 & R51) was so brief that my camera didn't capture it. I repeated this test three times (with replacement resistors) and was unable to capture the spark on film, even at 960 frames per second. It's very fast. For completeness, here's my second attempt, which (also) doesn't show anything:

Obviously each subsequent test I perform is less and less faithful to the "SS was previously as-found" standard. Ideally I'd test again with another Signal Soother.

I did physically witness said sparking during each test. After each test both R41 & R51 were visibly damaged, and also measured above 100 megaohms (they were previously 1 kOhm resistors). Here's the original R51 component after my first test (R41 looked identical):


During my third test, I placed a high speed differential probe across D5, which is in parallel with the 'V5' opamp, and also a 100 nF capacitor. I had this probe set to trigger on a rising edge at 39 volts. When I repeated the service disconnect test, the probe did NOT trigger. That means the 36 volt zener diode is heroically shunting ~216 mA ([252 - 36] / 1000 ohms), which is ~61 times it's nominal 3.5 mA rating. This means the voltage rail driving this opamp never exceeded 39 volts, which means the opamp probably isn't damaged.
Note that the opamp's Vabs(max) is 36 volts, so technically we exceeded its ratings... but both opamps (V4 & V5) worked again each time after replacing the 1000 ohms resistors. It's possible there is residual die damage to these parts, which is difficult to test for without a much larger sample size (because internal high voltage part failures don't always immediately present themselves). However, I didn't observe opamp failure during my testing; only damage to the high-side 1000 resistors.

If someone else wants to send another Signal Soother, I'll test it again on the bench. Of course, anyone with the Signal Soother still installed (why?) can easily verify my findings by removing the service disconnect and then attempting to start the car.

Here's how the Signal Soother behaved before and after each test (results were identical all three times):
Blade# | deltaVout (before) | deltaVout (after) | circuit damaged?
B1 | 18 | 18 | No
B2 | 18 | 18 | No
B3 | 18 | 18 | No
B4 | 18 | 0 | Yes, R41 visibly damaged, measures above 100 megaohms
B5 | 18 | 0 | Yes, R51 visibly damaged, measures above 100 megaohms
B6 | 18 | 18 | No
B7 | 18 | 18 | No
B8 | 18 | 18 | No
B9 | 18 | 18 | No
B10 | 18 | 18 | No
B11 | 18 | 18 | No
B12 | 18 | 18 | No
B13 | 18 | 18 | No
B14 | 18 | 18 | No

These results are unsurprising. When the entire pack voltage is applied to each opamp circuit on the signal soother, the 36 volt zener diode clamps each opamps' voltage rail, which caused 216 volts (252 - 36) to develop across both R41 & R51. Each 1 kOhm resistor therefore briefly sinks 46.6 watts, which exceeds the nominal power rating by ~375 times (hence why I saw them flash during each test).

 

Nope, and I previously outlined why (i.e. I don't own a Prius).
You have a Prius though, and IIRC you also have a Signal Soother. Why not test in yours? It would only take you a few minutes... all you need to do is remove the service disconnect. Imagine your (hypothetical) sweet victory post when your Signal Soother doesn't fail and you get to rub it in my face.

Please verify my findings in situ. If you're correct, your Signal Soother will emerge undamaged. If I'm correct, you'll just need to remove your Signal Soother afterwards, which you should do anyway. So then there's no reason for you to not test it. Maybe I'm wrong... we'll never know until someone tests. That's how science works.

Did you have any thoughts regarding how the failure I predicted (across V4 & V5) is exactly where the failure occurred on the bench? It's almost like there are symbols you can write on paper that some people (not you) can use to communicate ideas without physical objects.

Moving goalposts.
Can you please outline which 'credentials' you'd like to see.

Related: I can't wait to:
-review Jack's Signal Soother notes & schematic during discovery, and;
-cross-examine Jack's thought process regarding why he intentionally created a product whose sole purpose was to deceive the last line of defense after his LFP cells failed.

If I prevail on the merits, ultimately Jack would cover my legal costs after my countersuit.
And if I lose, yeah it's all play money at this point.

Thanks, again!

 

@AzusaPrius: Still waiting on your in-vehicle test results.

 

Here's your opportunity to prove the bench testing to be wrong. And it's a very simple test for you to do.
Verify your soother is installed.
Set up a video camera or just have a friend record the test so it can be posted.
Remove your safety plug.
Gain access to where the (+) and (-) cables fasten to the HV battery.
With your right hand, touch the (+) HV battery cable terminal.
While doing that, touch the (-) HV battery cable terminal with your left hand.

You'll get the answer pretty quick, one way or the other.

Let us know how it works out.

Update: Yeah, please..no one actually do this...not even Az. Use a multimeter instead. The results may even vary depending on which version of soother. One version may juice you, another may not. Only testing will tell the tale, but since doing it in a controlled environment is thought by some to be ineffective (Remember this? So no actual real test on a Prius yet? Bench testing in your little lab and trench coat wont hold up buddy.) maybe that someone can do it on a battery that's installed in a car?

 

Since I don't want anyone who missed your sarcasm to win a Darwin Award...
...if my bench top results are correct in situ, please heed the following "YOU WILL DIE" message:

DANGER: IF THE SIGNAL SOOTHER IS INSTALLED AND THE SERVICE DISCONNECT IS REMOVED, TOUCHING THE EXPOSED HIGH VOLTAGE CONTACTS COULD KILL YOU. THE SIGNAL SOOTHER CIRCUITRY REMOVES THE GALVANIC ISOLATION SAFETY MECHANISM THAT WOULD NORMALLY EXIST WHEN THE SERVICE DISCONNECT IS UNPLUGGED. IF THIS OCCURS, THE ENTIRE PACK VOLTAGE COULD EXIST ACROSS THE SERVICE DISCONNECT TERMINALS.