bitcoinonymous

Non-Toxic #Bitcoin Maximalist since 2011

[bitcoinonymous @bitcoinonymous]

@callmeed There’s a middle ground. Call the desk from your room, say “checking out of room 317”, and hang up. Definitely don’t wait in line in the lobby or navigate a phone menu or app, but if they answer immediately (and they usually do) it costs almost nothing and is a nice thing to do.

[bitcoinonymous @bitcoinonymous]

FSD 14 is amazing @aelluswamy. One complicated exception. There’s a road on my commute that is one way at rush hour. Any ideas for how to solve this safely?

[bitcoinonymous @bitcoinonymous]

This is the final boss for highway driving. Never trust a big rig. Never hang out near it. Pass it or wait or stay away. None of these half measures, it’s the only time I take over on the highway. The problem is that is trained on the highway driving behavior of idiots as well as smart people. Any clip where you are directly alongside a big rig at greater than 35 mph for more than five seconds should be disqualified from being used for training. Those are bad/unsafe drivers. Any autonomous vehicle software should be trained on clips from people who understand defensive driving. Never assume the other guy will do the right thing.

[bitcoinonymous @bitcoinonymous]

@aubreystrobel I agree 100%. I once took the summer off to pursue some non work interests and was both humbled and delighted by how little anyone cared or even noticed. The uncomfortable and liberating truth is that nobody is thinking about you.

[AMERICAN HODL 🇺🇸 @americanhodl8]

I used to be a Bitcoin maxi. I still am, but I used to be too.

[bitcoinonymous @bitcoinonymous]

@AP_Abacus @MorganStanley Why does this not seem to apply to E*Trade? Aren’t they owned by MS?

[bitcoinonymous @bitcoinonymous]

@AP_Abacus Who offers an in kind path for $1,000,000 of IBIT shares?

[bitcoinonymous @bitcoinonymous]

We should limit RF broadcast to whatever threshold can escape the solar system. When this goes into effect, it will appear that our civ died. I suspect this will attract science first civilizations who want to know what happened, and naturally exclude civilizations that want to engage with us to prove some superiority to themselves.

[bitcoinonymous @bitcoinonymous]

@hodlonaut Great write up. Thank you.

[bitcoinonymous @bitcoinonymous]

@americanhodl8 @Xentagz At one point my wife overheard you and Odell and Marty on a podcast talking about all 4 of those and said to me “I thought you said you were listening to a Bitcoin podcast.” Smart lady that one.

[bitcoinonymous @bitcoinonymous]

@DocumentingBTC @HaileyLennonBTC Quite a claim!

[bitcoinonymous @bitcoinonymous]

I tried this, but their tooling for annual renewal with bitcoin was so broken that I eventually surrendered and gave them my credit card. Maybe @lopp has improved the pseudo-anon Bitcoin payment renewal process since then, but when I tried not to dox myself it was not a viable path.

[bitcoinonymous @bitcoinonymous]

@bensig It just always seemed too complicated for the "security" of the air gap. Not saying air gaps are bad, but too many ways to mess it up. I can't claim it was because I thought the entropy was low. Just lucky to have avoided it.

[Sky @skysupersonic]

Dear COLDCARD attackers,

Congratulations. You successfully found a bug, exploited, and have collectively gathered around 1,500 bitcoin.

Now you are sitting on one of the most blacklisted coin stacks in history. It will be nearly impossible for you to exchange for currency or deal with any institution. Every move you make will be highly scrutinized and tracked. Presumably for the rest of your lives, you will be looking over your shoulder.

I suggest you plea with CoinKite to return all stolen funds in return for an audit fee of 5% of the loot. This way your work is rewarded and you are able to enjoy the fruits of your labor.

Respectfully,
hodlers worldwide

[bitcoinonymous @bitcoinonymous]

@BLKMDL3 @wholemars @ChuckCook @DirtyTesLa

[bitcoinonymous @bitcoinonymous]

BMS_a029, BMS_a079, BMS_a117, BMS_w117, BMS_a074, BMS_d265 on my 2021 Model Y: the docs said the degraded pack would charge to 50%. Reality was more interesting.

2021 Model Y, 63,000 miles, on a family road trip in the Southeast, 500 miles from home in Northern Virginia. Woke up one morning to a stack of HV battery alerts. Put the family on the train, drove the car back myself, and logged brick voltages the whole way.

The car was excellent throughout. It caught the fault early, contained it, said "safe to drive," and got me home without a single unplanned stop. What made the trip harder than it needed to be was the documentation: it says the cap corresponds to 50% charge, and what the pack actually delivered was 34%, then 32%, then 30%, then 28%. This post is about that gap, what I learned working around it, and a few things I'd love to see, in case it catches the eye of someone at Tesla who works on this.

If you're sitting at a Supercharger right now wondering why your car won't charge, skip down to FIELD GUIDE.

THE ALERTS

Five, inside one second, timestamped one morning at 7:15 am:

- BMS_w117_SW_Delta_SOC_Weak_Short
- BMS_a117 (silent limiting)
- BMS_a029 (max charge level reduced)
- BMS_d265_replaceBatteryPack
- BMS_a079_Max_Charge_Level_Exceeded

I think I also saw BMS_a074_Max_Charge_Level_Reduced at some point during the trip, but I'm not 100% sure.

Well inside the 8yr/120k battery warranty. Every one of them said "Safe to drive - Schedule service." The car never claimed to be disabled or dangerous. It said, in effect: I have a problem, I've contained it, get it looked at. That posture is what the rest of this post is about.

a029's description, verbatim:

"The Battery Management System (BMS) detects a condition internal to the high voltage (HV) battery that requires the max allowed brick voltage to be limited. Potential Impact: The target max brick voltage for charge is limited to the voltage corresponding to 50% SOC."

The BMS found one brick group misbehaving and clamped max brick voltage to 3.74V, which it says corresponds to 50%.

Credit first: the detection is good. w117 isn't watching a voltage threshold. It's watching one brick's state of charge diverge from its neighbors over time. It flagged a bad cell group in a car that drove completely normally, and d265 went ahead and called for a pack replacement on its own. In my data, the same brick (index 88) showed up as the pack minimum at every single reading across two days and hundreds of miles, while the maximum rotated among the healthy bricks. The BMS knew exactly which cell group it was talking about.

THE FIRST DAY: FIGURING OUT THE RULES

The alert payload that first morning read maxBrickV = 4.06, maxBrickVTarget = 3.74. The pack was sitting well above the new ceiling, so the car simply refused to charge. A 1 kW trickle charger, then a Supercharger, both declined. The local service center turned out not to do HV battery work at all, so the question became whether this car could get itself home.

Nothing to do but drive the charge down and find out where the ceiling really was. My working assumption, from the documentation, was somewhere around 50%.

At 52% displayed: max brick 3.82V. Still can't charge.

At 45%: 3.80V. Still can't charge.

At 41%: 3.73V. Under the target at last, but with almost no headroom to work with.

The first session that actually accepted charge started around 32% and stopped at 34%. That was the real ceiling. Not 50%.

THE CORRIDOR: A RHYTHM OF SMALL STOPS

From there it became a pattern, repeated up the interstate through the Carolinas and into Virginia: drive about 45 miles, plug in, charge for a short stretch, unplug, repeat.

The stops settled into a shape. Fast charging for the first stretch, then the instant a brick touched 3.74V, a hard drop into a taper like a healthy pack floating near 100%, except at a third of the charge.

Two things emerged from the rhythm.

The ceiling kept dropping. 34% the first time. Then 32%. Then 30%. By Northern Virginia, 28%. I don't have a confident explanation. One observation: the first Supercharger stop peaked at 90 kW and a later one at 114 kW, so the car was allowing more current as the day warmed up. If the ceiling is really about voltage under charging load, more current means hitting it sooner. That's my speculation, offered in case it's useful to someone who can see the other side of the data. The BMS tightening the target as it learns would fit equally well.

Charging current mattered enormously. The clearest single data point of the trip came near the end: plugged in at 10% SOC, peaked at 114 kW, and hit the 3.74V ceiling in under a minute. At ten percent. Forty points below where the cap is nominally supposed to bite. Meanwhile an overnight Level 2 session at a hotel, at far lower current, quietly reached 32% right after a Supercharger session had stopped at 30%, with the ceiling trending down at the time. Same pack, same day, better result at lower power.

Put those together and the picture is fairly clear: the 3.74V test is being applied to cell voltage WHILE charging current is flowing, and the voltage rise from that current is most of the gap between the documented 50% and the 28% I was getting.

THE ERROR MESSAGE IS NOT THE TRUTH. THE CHARGE RATE IS.

This cost me the most confusion of anything, so it gets its own section.

Once you hit the cap, the car throws "Unable to charge - Maximum charge level reached." It stays on screen. It looks final.

I attached a photo of that banner displayed while the car reports 43 kW, 173 mi/hr, +6 kWh, at 20% SOC. It was charging fine.

Ignore the banner. Watch the kW. If power is flowing you're still gaining, and it can keep flowing at meaningful rates well after the message appears. There's no other reliable indicator. The charge slider still travels to 100%, and the SOC percentage on the screen tells you nothing about where the cap sits.

SMALL ODDITIES FROM THE ROAD

Collected here because someone at Tesla might recognize them instantly, even though I can't explain them:

Regen trips the alert. Pulling away from a charger near the ceiling, the first real braking event gave three beeps and the same "unable to charge" popup. Every time, at Superchargers and Level 2 departures alike. Makes sense in hindsight (regen is charging, and pushes bricks over 3.74V), but it's startling, and I never saw the dashed limited-regen line on the power meter through any of it.

Replugging works sometimes. After a session tags the 3.74 threshold it sometimes stops dead, unplug and replug and it restarts, but the second session never went above 75 to 80 kW for me even when the first went higher. And on the last stop, it peaked at 114 kW and tapered on its own with no replug needed at all. No idea what variables were at play here.

Preconditioning worked against me. Navigating to a Supercharger preheats the pack, which costs range I didn't have and appeared to raise the allowed charge rate. If the ceiling is a voltage-under-load problem, that's the opposite of helpful. I ended up navigating to addresses NEAR Superchargers instead, just to keep the pack cool.

FIELD GUIDE

If you just got these alerts, here's what I wish I'd known that first morning.

Stated SOC means nothing. Watch brick voltage. The alert says 50%. I couldn't charge at all until the pack was at 29%. In Tessie you have to go into the profiler, then the charging section, and turn it on. It's not on by default and it's capped at 10 hours a month, so don't leave it running. Watch BrickVoltageMax. That's the value the a079 payload names, and it's what the cap acts on. Under 3.74V you can charge, over it you can't, regardless of the percentage.

Watch the charge rate, not the error banner. The banner appears while the car is still taking 43 kW.

a079 clears and comes back. It only exists while you're above the cap. Clearing is not improvement.

Short legs. I ran about 45 miles between stops. The bulk of each charge comes fast, but the moment any brick touches 3.74V, the session drops into a taper like a healthy pack floating near 100%, so most of the time at the stall is spent tapering. Unplug once the rate collapses. The last percent or two isn't worth the wait.

Level 2 is slightly better and less annoying. 32% on AC right after a Supercharger session that stopped at 30%, no replug needed.

Don't chase the floor. I planned around 10% and dipped to 8% once. Didn't want to find out what happens at 3% with a brick that's already weak.

If you are using FSD, Sloth mode hit the stated arrival SOC 100% of the time, anything above sloth did not. Speed is not your friend with a degraded pack if your goal is range.

WHAT I'D LOVE TO SEE

None of this needs the cap loosened. It needs the cap to hit its own stated number, or at least closer to its stated number, and the rest of the car to work from what's actually being delivered.

1. Account for the voltage rise under load before comparing to 3.74V. Test the cell's resting voltage, not the terminal voltage while hundreds of amps are flowing. From my data this looks like most of the gap between the documented 50% and the 28% I ended up with.

2. Cap DC charge rate when these alerts are active. Something around 50 kW. Less voltage rise, probably fewer replug cycles, and it costs nothing. My bulk phase lasted under a minute anyway.

3. Skip preconditioning for a charging speed this pack can't use, or limit it to whatever a 50 kW rate actually needs. It currently spends range the driver doesn't have. Maybe 80F instead of 120F? Degraded packs with a software limited capacity of ~30% don't need any heating in my opinion.

4. Make "unable to charge" mean unable to charge. It currently fires while the car is taking 43 kW.

5. Handle regen near the ceiling without a popup. Three beeps and a charging error mid-brake is alarming for something that's working as designed.

6. Plan trips on delivered capacity. The planner DID adjust. It routed me on a 50% pack. I was getting 28%-34%.

That last one sums it up. The degraded-pack mode exists, and word gets around. The word is just 50%. The BMS did its job: found the brick, clamped the voltage, called for its own replacement. The trip planner even got the memo. But everything downstream believes the documented figure, and nothing seems to check it against what the car actually charged to at the last several stops.

CREDIT WHERE IT'S DUE

The car never stranded me. Five hundred miles on a pack it had already flagged for replacement, no unplanned stops, and it told me exactly what was wrong in machine-readable detail. I could watch brick 88 sitting below its neighbors from my phone at every stop of the trip.

And FSD v14 Lite on HW3 did well over 99% of the driving, on a battery running at a fraction of its capacity. Two days, hundreds of miles, a stack of charging stops, and the actual driving was the easy part. Truly remarkable machine.

Pack's under warranty. Service appointment is booked. The system, on the whole, worked. It just deserves a degraded-pack mode as polished as the rest of the car.

Thanks for reading!

@Tesla @teslasi @elonmusk @aelluswamy @yunta_tsai @ClairesEquation @EricETesla @pduan @teslafi @teslascope

[bitcoinonymous @bitcoinonymous]

@deathandbitcoin What tools did you use to make this?

[bitcoinonymous @bitcoinonymous]

@DavidFBailey It's you David, the call is coming from inside the house.

[Donald Trump Jr. @DonaldJTrumpJr]

Remember when they said Trump would get us into a full-blown war with Iran?

Times like these should make us all realize how lucky we were to have Trump, his strength and his resolve.

[bitcoinonymous @bitcoinonymous]

And today they got over 1000 in the fist 60 minutes. 1000btc per hour.

[bitcoinonymous @bitcoinonymous]

$STRC just scooped up 100 $BTC worth over $71 million in exactly 13 minutes. $MSTR @Strategy @saylor