My Project Lithium Battery Caught Fire

Seems to get turned into some kind of s9e.github.io iframe link.

Maybe it's like what PriusChat does to Amazon links, which you can prevent if you just make your own link to it where you specify the link text, instead of just dropping the URL in the post.

 

This thread is exactly how companies can be lost practically overnight.............information originally kept in the shadows suddenly moves into the light and whether it's amplified 100x more than reality ,or not,....it becomes impossible to overcome...

I hope it gets resolved in a good way.

 

I would have thought that those would have been at least singly latching connectors, or better yet, connectors that latch on both sides. Very hard to tell from the picture, but that looks like the usual sort of header to pin connector that just stays on by friction. I too have seen friction connectors work loose on shipping. For instance, the large motherboard ATX connector that goes to PC motherboards sometimes come off in transit (or possibly they were never fully on when assembled, either is bad). Worst of all are nonlatching SATA connectors, which often pull out off of drives enough to cause errors if the cable is routed in such a way that it applies torque that rotates the connector out of its socket.

I don't know if I would trust either hot glue or silicone glue to hold long term in an automotive application like this. A solid mechanical clamp of some sort would likely be better, and might not be more expensive if it was easier to install.

 

No it's a latching connector. He's making stuff up for PR. The real reason the fire occurred was because of the signal soother. He essentially removed the Prius's ability to detect voltage and resistance properly so the Battery ECU could never detect the actual failure that was occurring... which drove the module to catastrophic failure.

 

Can somebody describe this Signal Soother in more detail?

I remember the first time I heard that name it made my eyes squinch up, but it would still be interesting to know more of the details.

 

Read my original post earlier in this thread... there is a direct link to the post below with further links in that specific post that describe it in detail:

Description of How Signal Soother works and the entire rationalization for why it is the worst idea imaginable.

 

Ok, here's an attempt at a link to the reddit content that doesn't turn into s9e.github.io iframe javascript headaches:

TheLastMojojomo comments on Project Lithium fire reddit thread

Has anyone posted a photo of this Signal Soother, or a schematic, or anything? The linked reddit content kind of just says it interferes with the signals seen by the battery ECU, but isn't heavy on detail.

 

Are you using an internet browser or the app? If you use an internet browser to view the post... it turns out fine. The app is basically defunct and should not be used at this point. It formats everything incorrectly and causes errors. Website is designed very well for Internet Browser use on mobile and works much better overall.

 

I use an internet browser with careful security settings, which means posting a reddit link that gets turned into an s9e.github.io iframe to content that won't render without JavaScript is a pretty good way of making sure I won't see it. I could relax multiple security settings and reload, but if I wanted to do that, I wouldn't have them set that way in the first place.

Giving a straight URL to a page that just shows the content without funny business is usually the best way to make it accessible to the greatest number of readers.

 

The exact thing I have asked to see for years. Of course the earlier comparator fet circuit is now some sort of analog emulator. Given it is sold it can’t be secret.

 

Looks like AzusaPrius at least posted a photo a while back.



and gave a description, in the same post, that could sound like it "will balance them better" (the modules), or could just mean it makes them look like "both at 16.50 each".

@sworzeh, is this "Signal Soother" in your failed battery? Is it still in photographable condition? Could you attach photos?

By the way, attaching photos right to your PriusChat posts makes them easiest for the most people to view without problems; imgur won't work for everybody. There was an issue here making attachments difficult for a while, but I believe it's fixed now.

 

Off topic, but a piece of related history. When the first generation Honda Civic Hybrid (HCH1) came out in 2003 it soon became apparent that the packs didn't last nearly as long as those on a Prius. The pack on the HCH1 looks something like a nuclear reactor. There is a rubber matrix which has a series of parallel cylindrical holes, and these are occupied by "sticks", which are several cylindrical NiMH cells spot welded to each other. (To first order, Civic stick equals Prius module.) There were connectors on each side of the matrix to put all of the sticks in series. The HCH1 run ended in 2005, for the next model the sticks changed to a sort of double stick, and shortly thereafter Honda started replacing failed packs, including those still under warranty, with ones made with sticks scavenged from old packs. That is, no new sticks, only old ones "restored" by discharge/recharge, similar to what people sometimes do with Prius modules. As soon as packs with new sticks stopped being available from the manufacturer several small companies appeared to try to service this market. Like with the Prius, some were pack rebuilders (which is what the manufacturer had become), and some started using new sticks made in China. In the original packs each stick was shrink wrapped. One of these companies thought that the shrink wrap wasn't necessary and used "naked" sticks in the matrix. At least one fire resulted. At that point somebody thought to check the conductance of the matrix, and contrary to most people's expectations, it did in fact conduct. So no surprise that there was eventually a fire. The new Chinese sticks were very hit and miss, I had an IMA pack that used them and it was very short lived. Between the pack issues and the horribly unreliable CVT this was an expensive car to keep running. These days I only very rarely see one, even though they were pretty common in my area at one time. (The city of Pasadena bought a bunch of them for parking enforcement in 2004 or so, but I have not seen one of their cars in at least a decade.)

 

Yep, I used to own one of those. Pack gave out at around 72K; and was replaced under warranty. Promptly sold the car after the pack replacement. Got top dollar and gave the battery replacement paperwork to the new owner. I never got better than 42 mpg and tried everything - even sticking to 55 mph on the highway! I've owned straight stripped down ICE only Honda civic 5 speeds that got that without the high price. Really disappointed with that car!

 

I was the one who wanted to use silicone glue, was overruled by my colleagues - mainly because hot glue was cheaper. The connector in question on my product was a latched connector. We found that the rapid temperature changes along with shaking during transport would work those connectors lose. The three dabs of glue would prevent those connectors from popping lose. One dab on each end and one directly on top of the latch. The beauty of this is that the glue pops off using a dental pick; so you can actually work on the equipment. If silicone glue was used, it would be harder to clean off the connector - but stretchy enough to maintain a positive connection under all conditions.

 

Reading that thread it's clear @AzusaPrius does not know what they're talking about. The signal soother does not actually balance the battery... It fakes signals to make it look more balanced. Nor does it increase power output. Power output is solely regulated and rate limited by the Hybrid Control ECU.

The Signal soother is placed on the cable that measures Internal Resistance and Voltage of the Modules that connects to the Battery ECU... Charge does not flow through this cable... signals do... or more correctly extremely extremely low amounts of power... which is why it's called a signal soother.

If you want to verify what I'm saying... put your Toyota Hybrid into Ignition On Mode (2 clicks of power button with foot off brake) and connect to Dr. Prius while the ECU's are running. You are getting measurements from the Battery ECU without the Hybrid Battery being connected to the rest of the Hybrid System because each individual module is being measured by the Battery ECU. This is how the Prius determines whether it is safe to start the car... or to throw an error code based off of these signals.

The Signal Soother interupts the normal signal circuit and then transmits it's own false signals to the Battery ECU... while the battery Modules themselves actually could be wildly out of balance or suffering from high Internal resistance or extreme voltage difference. So the Battery ECU is essentially blinded... forcing the Prius to function normally while an underlying problem continues to get worse.

This is why @sworzeh's Battery caught on fire... THE BATTERY ECU WAS BLINDED TO THE SEVERE ERRORS THAT WERE OCCURING... resulting in a fire.

The whole reason Jack sent the signal soother was because the Battery ECU was detecting Hybrid Battery errors like increased internal resistance or voltage difference.

Instead of actually doing any sort of leg work to diagnose whether the Battery ECU is actually detecting errors (which it was and the errors were real) Jack sent a signal soother to be installed and blinded the Battery ECU resulting in a module melting because the original error that was detected was not fixed... and then unable to be detected after signal soother installation... which then resulted in a catastrophic melt down.

You guys do not understand how severe of an issue this is... this could literally kill someone and is lawsuit material. The fact that Jack made and implemented this shows his fundamental understanding of the Hybrid System is so flawed that all of these Batteries should never have been sold and should be immediately recalled before someone gets hurt or killed.

I have just messaged this to @jacktheripper and I hope he sees the logic in my statements and will take actions to prevent more dangerous circumstances from occuring. This Signal Soother needs to be recalled before somebody gets hurt or killed. I have been very worked up over the past 48-72 hrs seeing the Signal Soother implemented and the end result of it... and have more so been attacking Jack rather than being diplomatic and that is going to change right now.

Jack made me aware of the Signal Soother's existence back in November of 2023 when I was a beta tester and immediately ended beta testing with Jack because I couldn't believe he was implementing this as a solution to his Hybrid Battery Issues... I was so stunned I immediately quit beta testing without telling him why. My thought process was that "if his understanding of the Hybrid System is so flawed... why even bother..." and then all of a sudden all that emotion came roaring back 3 days ago when I saw the implementation and end result of the Signal Soother. I needed to get all the emotion out of my system.

With all that being said... I am literally just a dumb mailman... but I have analyzed the Hybrid System to an extreme degree. If anybody wants to weeve a counter narrative and explain why I'm wrong... I will listen. But I believe that I am right. I can be persuaded if the logic is sound. But I will retort with evidence of my own if the logic is not sound.

 

separately... are you familiar with how the ECU in a Prius auto adjusts itself based on the signals it receives from not just the battery but the engine as well? by my logic, a signal soother would send adjusted signals based on the battery parameters so parameters of the ECU can do their magic to better adjust to the battery that's being used. There might be one key piece to all this you are missing and that's dissecting the ECU and how it functions in relation to what its being plugged into!
I need to find a 2017 ecu so I can really solidify this logic and how my 2013 PHEV can obtain the 80mph in EV mode like the Prime does with ECU and additional clutch hardware.

 

Yes, I can see that in the photo.

That would depend entirely on the internals of the Signal Soother. There is nothing about that sense wire harness that limits the current that can flow through those wires (up to the ampacity limit of the wires themselves, of course). This lesson is occasionally driven home when a couple pins at the connector get shorted by corrosion.

When it is plugged directly into the battery ECU the way God and Toyota intended, the currents that flow depend on how the battery ECU is built, and the ECU doesn't do anything but high-impedance voltage measurement. When plugged into a Signal Soother instead, that's going to determine what happens.

I'm familiar with what the battery ECU does.

I understand what you believe the Signal Soother does.

It would be great to see some teardown photos with enough detail to decide if that's what I believe the Signal Soother does.

 

Without a complete analysis of the signal soother I cannot say whether this one is a good or bad. The idea is not intrinsically bad for adapting a new battery technology for a battery controller which is expecting a different technology - so long as it still communicates to that battery controller when there is a problem.

For instance, imagine that an LFP module had a fail safe mode that detected an internal error and completely bypassed the cells it contained, perhaps using something like a power FET or even a relay (it only needs to close once and could latch, the module would need to be serviced or replaced, and the relay could be replaced at that time). The intermediary device could indicate that that module was just a low voltage, not a a zero voltage, and make up the difference across the other cells. The car would throw a battery related code (with a different meaning than that for the original pack) but wouldn't drop dead while driving. (This assumes of course that the total voltage across the remaining modules is sufficient for the car to keep operating.) Similarly, if the internal LFP controllers balanced the internal cells in a manner that resulted in significant variation in the total module voltage, but wasn't harmful, the intermediary might reasonably lie to the BCU and present it with a picture of an older technology pack which would result in the appropriate charging/discharging rates, and no error code.

Also, isn't an intermediary device mandatory in any situation where the number of modules is reduced? I can't imagine the BCU would be happy with half the voltage sense lines reporting no or spurious voltages.

 

Yes it's not the wires that set the current limit.. it's the Battery ECU that does when it's connected to it.

The Battery ECU is basically just a mutimeter with 14 inputs instead of 1 with all of them set to measure internal resistance and voltage. A mutimeter has extremely high resistance so current flow is almost non-existence when measuring voltage.

Expanding on this example further... Take your red positive multimeter lead and splice the signal soother into the middle of that wire. Then connect the two red ends as normal... one in the multimeter and the other touching the positive terminal of a 12v Battery and of course complete the circuit by grounding the other black lead.

The level of current that can flow through the signal soother in this made up scenario is rate limited by the resistance of the multimeter... the Battery ECU is like the multimeter and the signal soother is spliced into the middle. If power flows any other direction besides through the signal soother and then through the multimeter... it entirely defeats the purpose of the Battery ECU and would be considered a short. You even described this in your comment with the detail of a shorted battery ECU pins.

You do not want any more current flowing then what is supposed to be allowed by the internal resistance of the Battery ECU. Any other current flow creates a short and will render the Battery ECU inert and useless effecting readings even more so because of electricity taking the path of least resistance to the short instead of through the Battery ECU. If somehow the signal soother is designed to use more current... this makes it even more dangerous.

Yes... but the Signal Soother will still be rate limited in current flow by the Battery ECU's impedance because it's part of that circuit... unless it's shorting. More current can't flow through the signal soother than the Battery ECU normally allows due to it's impedance as you stated. The only thing the signal soother can do is increase resistance further and decrease current and the circuitry components can alter voltage uniformity and resistance along the way.

The only example that would lead me to believe that this isn't how the signal soother works is that one of the pins in the orange Battery ECU Connector is considered a direct ground and not meant for measuring voltage and the signal soother somehow uses this for a purpose. Is there a schematic somewhere that can prove this? I might have one actually. I'll check.

 

The whole point of the Signal Soother was to prevent error codes being read by the Battery ECU which were real and present in the Modules... that is how it came into existence and why it was implemented. This is what you need to fundamentally understand. The Battery ECU would detect errors of Internal Resistance. I experienced this during beta testing. Below is a text I sent to Jack a few weeks ago.



The Signal Soother was created to stop exactly what you said it should not stop... that is why the logic of the device is so flawed. It stops communication of errors that are real and occuring and was it's entire purpose. A signal soother wouldn't have been needed if there were no errors.

Yes... Jack did do exactly what you said shouldn't be done... the reason was he got it in his head that the error codes were glitches when they were not.

I think this is primarily because the Battery ECU does not transmit Internal Resistance readings as viewable parameter via OBDII until the Battery is above 70°F... but is still actively measuring Resistance in the background even though it's not viewable in Dr. Prius. So the Battery Modules were experiencing increased resistance due to an underlying flaw in the Module and the Battery ECU was detecting it correctly... but internal resistance would not show the accurate # as viewable parameter in Dr. Prius until 70°F was hit for pack temperature... even though it would still record it accurately in the background. I tried explaining this to him at the time... but he wouldn't listen:

You'll need to clarify... there are not reduced number of active sense lines with Project Lithium... sure OEM NiMh had 28 modules and Project Lithium had 14... but 2 Nimh Modules are the equivalent of a Project Lithium Module. There's an internal bus bar that connects the Lithium Module to itself.

NiMh there's 1 voltage sense wire for every 2 Modules with a total of 28 modules...

Project Lithium There's 1 voltage sense wire for every 1 module with a total of 14 modules.

1 Project Lithium Module is the equivalent of 2 NiMh Module.