I monitored the 12v battery voltage during an OTA update

[scottf200]

The Tesla battery is an AGM deep cycle battery (ie. marine, etc usage). 12 Volt Battery Compendium – TeslaTap

Look at graphics of the DC-DC converter charging voltage
1) before the install
2) during OTA install where the software for the DC-DC convert will be logically be "OFF" and the 12v is on it's own >> key point <<
3) after the install for X number of hours

From article and quote below: Marine Battery Maintenance 101 - e Marine Systems
*** "taken to 14.2-14.4 VDC before it is fully charged"
*** "Once fully charged, marine batteries should be held at a considerably lower voltage to maintain their charge - typically 13.2 to 13.4 volts."

Why Marine Batteries Fail -- When a lead-acid battery is discharged, a soft lead sulfate material forms on the battery plates. During the battery's recharge, this material is lifted off the plates and recombined into the battery's electrolyte solution. If, however, the battery is left in a partial state of discharge for as short as 3 days, the lead sulfate material will begin to harden and crystallize, forming a permanent insulating barrier. As this barrier becomes thicker and thicker, the battery's ability to accept a charge or deliver energy is diminished, resulting in the perception that the battery is no longer usable. The accumulation of such deposits, otherwise known as sulfation, is the most destructive process in the life of any lead-acid battery.

Multi-stage Battery Charging -- A typical 12-volt lead-acid battery must be taken to approximately 14.2-14.4 VDC before it is fully charged. (For 24 volt systems, double these figures.) If taken to a lesser voltage level, some of the sulfate deposits that form during discharge will remain on the plates. Over time, these deposits will cause a 200 amp-hour battery to act more like a 100 amp-hour battery, and battery life will be considerably shortened. Once fully charged, marine batteries should be held at a considerably lower voltage to maintain their charge - typically 13.2 to 13.4 volts. Higher voltage levels will "gas" the battery and boil off electrolyte, again shortening battery life.

I used the following device during the 2020.40.4 update a couple nights ago on a 12v battery that is only a few days old.
Aside: During a multi-state road trip my 2020.40.3 left my 12v battery in a bad state. I was jumped and then the install finished. The next day I drove to a Service Center that installed a new battery. Much longer version to that story. I was in NH that doesn't have Service Centers (as it turns out).

https://smile.amazon.com/QUICKLYNKS-Battery-Monitor-Bluetooth-Device/dp/B01MT4583U

Graphic below: https://i.imgur.com/PmFNdFS.jpg
White Xs were used to 'find' 14v. Blue line shows 14v relative to green app data plotted line.

Graphic below: https://i.imgur.com/rmGaBi2.jpg
White Xs were used to 'find' 14v. Blue line shows 14v relative to green app data plotted line.

 

[brkaus]

Our M3 died Tuesday night around 7pm. Installed 2020.40.4 on Monday night. Car purchased Oct 2018.

Good experience with roadside assistance, towing, and prompt service center response this morning.

Would have been better had they been able to dispatch roadside fix. Then I wouldn't have to deal with pickup today.

Windows were down, not a great place to be parked overnight, and I didn't want to mess with jumping it and driving it home to deal with today. The said they couldn't dispatch roadside to replace battery because the car was not online to confirm a 12v battery failure....

Thanks for the graphs!

 

[scottf200]

...
Good experience with roadside assistance, towing, and prompt service center response this morning.
Would have been better had they been able to dispatch roadside fix. Then I wouldn't have to deal with pickup today.

I was in Lincoln, NH. There are *no* service centers in the whole state of NH. We could not get mobile service to come to our rescue even tho they service the whole state but who knows where all they come from to do that (ie. probably some locals and some from other states).

Luckily, I decided on my own to try jumping the 12v via local auto mechanic after looked at YouTube videos to see how to get into my frunk to do it (I was locked out of my whole car). The auto mech brought a battery booster pack and after hooking it up, my X came to life (sentry mode horns blaring) and it completed the install starting at 60% progress bar and took 45 min-1 hr. The next day at a supercharge I got the 'replace 12v battery or your car could shut down message'. Long story but found an open (Sat) service center that we could barely make it to before they closed. We literally made it at 2:57p and they said we'd have to be there by 3pm. It was replaced by 4pm. Worse case I would have bought the 12v battery from them (no core charge) and found a way to get it installed even if I had to buy tools at Walmart!

 

[toolman335]

Are you guys on Wifi while traveling to get the updates?
There's no way I'd do a software update unless I was at home.

 

[HankLloydRight]

Are you guys on Wifi while traveling to get the updates?

If you wait long enough, FW updates can be sent over LTE.

 

[brkaus]

Are you guys on Wifi while traveling to get the updates?
There's no way I'd do a software update unless I was at home.

Nope. I see when updates are ready when I'm away, but only get them at home.

As a general rule, I avoid updates when I'm traveling.

 

[scottf200]

Are you guys on Wifi while traveling to get the updates?
There's no way I'd do a software update unless I was at home.

In my case, I was at an AirBnB for a couple of days and it had a simple password. I connected it to the car to get quick routing and mapping when we first got in the car and did things (hikes, scenic drives, etc) that day.

I never did an OTA update while traveling for 5 years. After seeing a couple 1000 on TeslaFI install that OTA then that meant there were many many 1000s getting it. No TMC threads or reddit threads were showing problems so I did it for the highway speed limit sign reading (construction zones, etc). This would not have been an issue if my 12v battery wasn't one the edge. Back to my hard rule of thumb and only admitting it here to triple remind fellow road trippers and pay it forward.

The main point of my thread was to provide empirical data and graphs showing how Tesla is maintaining the 12v battery. It seems very legit for a AGM deep cycle (~13v maintain and ~14v charge to full after things like an OTA).

 

[AlanSubie4Life]

The Tesla battery is an AGM deep cycle battery (ie. marine, etc usage). 12 Volt Battery Compendium – TeslaTap

Look at graphics of the DC-DC converter charging voltage
1) before the install
2) during OTA install where the software for the DC-DC convert will be logically be "OFF" and the 12v is on it's own >> key point <<
3) after the install for X number of hours

From article and quote below: Marine Battery Maintenance 101 - e Marine Systems
*** "taken to 14.2-14.4 VDC before it is fully charged"
*** "Once fully charged, marine batteries should be held at a considerably lower voltage to maintain their charge - typically 13.2 to 13.4 volts."




I used the following device during the 2020.40.4 update a couple nights ago on a 12v battery that is only a few days old.
Aside: During a multi-state road trip my 2020.40.3 left my 12v battery in a bad state. I was jumped and then the install finished. The next day I drove to a Service Center that installed a new battery. Much longer version to that story. I was in NH that doesn't have Service Centers (as it turns out).

https://smile.amazon.com/QUICKLYNKS-Battery-Monitor-Bluetooth-Device/dp/B01MT4583U

Graphic below: https://i.imgur.com/PmFNdFS.jpg
White Xs were used to 'find' 14v. Blue line shows 14v relative to green app data plotted line.

Graphic below: https://i.imgur.com/rmGaBi2.jpg
White Xs were used to 'find' 14v. Blue line shows 14v relative to green app data plotted line.

The Tesla battery is an AGM deep cycle battery (ie. marine, etc usage). 12 Volt Battery Compendium – TeslaTap

Look at graphics of the DC-DC converter charging voltage
1) before the install
2) during OTA install where the software for the DC-DC convert will be logically be "OFF" and the 12v is on it's own >> key point <<
3) after the install for X number of hours

From article and quote below: Marine Battery Maintenance 101 - e Marine Systems
*** "taken to 14.2-14.4 VDC before it is fully charged"
*** "Once fully charged, marine batteries should be held at a considerably lower voltage to maintain their charge - typically 13.2 to 13.4 volts."




I used the following device during the 2020.40.4 update a couple nights ago on a 12v battery that is only a few days old.
Aside: During a multi-state road trip my 2020.40.3 left my 12v battery in a bad state. I was jumped and then the install finished. The next day I drove to a Service Center that installed a new battery. Much longer version to that story. I was in NH that doesn't have Service Centers (as it turns out).

https://smile.amazon.com/QUICKLYNKS-Battery-Monitor-Bluetooth-Device/dp/B01MT4583U

Graphic below: https://i.imgur.com/PmFNdFS.jpg
White Xs were used to 'find' 14v. Blue line shows 14v relative to green app data plotted line.

Graphic below: https://i.imgur.com/rmGaBi2.jpg
White Xs were used to 'find' 14v. Blue line shows 14v relative to green app data plotted line.

It's very interesting how different your battery charging behavior is, compared to Model 3 (which may not have an AGM battery - unbelievably, no one knows for sure - we need someone to disassemble one!). I don't really understand why yours has so little cycling!

Here's mine. This capture is for a period where the car is not plugged in. I do generally have higher vampire drain, for unknown reasons. Obviously it's in the garage, so it's close to me, and I didn't take any special steps to turn off the phone key or anything like that.

I just wonder whether others see as many daily charging intervals as I'm seeing?

I'll turn off some features (data sharing, cabin overheat, phone bluetooth, etc.) to see if I can change the frequency. I just wonder whether other people see as much cycling. It's bad for the contactors, if nothing else! I could see this happening if the 12V had high self-discharge. Or, it would happen if something else had high discharge when in sleep mode.

I've noticed not every wake cycle is associated with a need for charging, though most appear to be (it pretty much appears to be every one in this particular plot). You can see one wake cycle somehow did not do any charging (I have no idea how that could physically happen, honestly).

Odd. Perhaps an important question for people is what sort of decay rate do they see after the DC-DC for the battery turns off? That should correlate to the total level of drain the battery sees (total of both self-discharge and sleep mode discharge).

Looks like mine decays at about 0.5V per hour (in the first hour). To some extent there is rapid decay of voltage naturally after a charging event in a 12V battery, from what I understand/observe. So maybe that's all that is happening there. Still, seems kind of fast compared to what I see on my ICE vehicle (which of course has minuscule drain on the battery, probably 500x to 1000x less).

(Incidentally, just checked and my DC-DC has been on from 9:45 to now, over 5 hours, at 13.4V (not 14.3V). And the car is making that quiet whining noise. It's not plugged in.)

However, here's another plot. In this one, you see some periods of time where the battery voltage is completely flat, but at a different voltage!

It's all very interesting. I know exactly when I was driving yesterday, and I don't see evidence the battery charging (14.3V) turned on at all in those periods (a bit of a surprise to me, but might make sense - it's not like a regular car which will always charge the battery when driving, I suppose). It looks like the DC-DC is turned on while driving, but is often set to a low voltage to avoid damaging the battery. But oddly, after parking the car and leaving at 5:30PM, the DC-DC stayed on, and did not turn off for a good hour. Explains why the voltage was so flat at ~13.3V.

So for Model 3:

- to me it looks like there is a charging voltage of 14.3V
- it looks like there is a maintain voltage of 13.3-13.4V
- Maybe there is some lower set point of ~13V as well? I can't explain longish flat periods otherwise. (2:30-4:30 AM for example) I understand that at some points in the discharge curve the voltage may "flatten" - but to me this seems a bit too flat, given the decay rate seen at other times/voltages. Maybe I'm wrong.

 

[brkaus]

Love this data. Looks like I need to buy one

 

[scottf200]

It's very interesting how different your battery charging behavior is, compared to Model 3 (which may not have an AGM battery - unbelievably, no one knows for sure - we need someone to disassemble one!). I don't really understand why yours has so little cycling!
...
So for Model 3:

- to me it looks like there is a charging voltage of 14.3V
- it looks like there is a maintain voltage of 13.3-13.4V
- Maybe there is some lower set point of ~13V as well? I can't explain longish flat periods otherwise. (2:30-4:30 AM for example) I understand that at some points in the discharge curve the voltage may "flatten" - but to me this seems a bit too flat, given the decay rate seen at other times/voltages. Maybe I'm wrong.

Wow, those charging characteristics are so very different than my X. My old battery had similar characteristics as my new battery (ie. 'look' of the battery monitor graph).

We need more empirical data from TM3s. It appears that when your TM3 12v battery hits about the 13v 'floor' that it charges back up. Maybe it can't hold a charge. I'd be pretty tempted to get a new battery if I were you. Your at 2 years (off production line) it seems.

My TMX doesn't sleep. I keep my FOB in an Altoid box from the early bug days of BT comm causing a problem IIRC. Maybe my DC-DC is always running and maintaining ~13.3v? Odd. I do have ScanMyTesla and I've never seen the DC-DC not working but I've only looked when I was in the car as I recall. If yours sleeps and you have something like TeslaFI that lets you look at the minute by minute raw data then you could see if any sleep period corresponds with your graph.

Your driving voltage is the only thing that makes sense to me.

 

[scottf200]

... Model 3 (which may not have an AGM battery - unbelievably, no one knows for sure - we need someone to disassemble one!) ...

Crimines, now I don't even know about my X battery after reading the TOO post.

The 12 Volt battery in the Model 3, not AGM? - Tesla Owners Online
Via: TOO member 'android04'

I've looked in the Hankook batteries website before and couldn't find specific info on the Model 3 battery on their lists, but didn't think downloading their catalog PDF would help. It obviously did help in finding more info, thanks for sharing. Here's a screenshot of the relevant page. The three batteries with (*1) next to them are in Teslas. The first one goes in the Model S and has a AGM in the name, the second goes in the Model X and only has MF (maintenance-free) in the name, and the third goes in Model 3/Y and also has MF in the name. So you seem to be right that our Model 3 batteries are not AGM. The question now is does it matter for longevity in our application?

 

[AlanSubie4Life]

My TMX doesn't sleep.

That explains everything about your plot. Seems like a problem for it not to sleep. Your vampire drain must be insane.

Maybe my DC-DC is always running and maintaining ~13.3v?

Yeah it looks like that is exactly what is happening.

If yours sleeps

Mine does. I hear the contactors opening and closing frequently; it's very easy to tell when they are opening and when they are closing, the two sounds are totally different. I'll pay closer attention but fairly sure them closing corresponds to the jumps up to 14.3V or whatever in the graph, typically.

If yours never sleeps, it would be hard to compare these graphs and the decay rate to see whether there is something wrong with the battery.

Hopefully someone with a Model 3 can post a similar plot and we can compare. I might post in that Model 3 battery thread to get more info from people.

 

[AlanSubie4Life]

Maybe it can't hold a charge

That is the key question. Need to see what a new battery looks like when the car is sleeping and the DC-DC is off.

 

[LowlyOilBurner]

I take it, this is all null and void if the car is plugged in? Or if the 12v batt is healthy?

 

[brkaus]

I take it, this is all null and void if the car is plugged in? Or if the 12v batt is healthy?

Even when plugged in, the car will control the 12v charging system. It doesn’t keep the high voltage pack energized all the time.

if the 12v battery is healthy is should fall slower, but there will still be some change.

 

[brkaus]

That is the key question. Need to see what a new battery looks like when the car is sleeping and the DC-DC is off.

I just ordered the device. And picked up our 3 from Tesla yesterday with a new 12v installed.

 

[AlanSubie4Life]

I just ordered the device. And picked up our 3 from Tesla yesterday with a new 12v installed.

Great. Should be interesting. Here's what it looks like with Sentry Mode on. For some reason my car thought it was away from home and left Sentry on for over 24 hours.

Rock solid at 13.3 to 13.4V, the maintain voltage.

The 14.3V is the charge voltage.

But you can see a rapid cycling of charging and discharging after I parked this afternoon. However, the car is plugged in right now. Will be interesting to see how yours looks (hopefully your car sleeps well and such).

Ideally, we'd figure out a pattern to watch for which indicates a battery that is about to fail...

I'm turning off all the data-sharing. Tomorrow I can unplug and see what difference it makes (as far as I can tell so far it doesn't seem to change much).

Would be good to annotate:

Plugged in/not plugged in
Configuration (no Summon Standby of course) Data Sharing? Cabin Overheat? Etc.

 

[power.saver]

@AlanSubie4Life Just installed today, M3 SR+ 9/2019 build with original battery. Prior to 5:30pm I had it on a 12v supply to test and configure Bluetooth. Car was on until 6pm, then sleeping with Sentry off. Scheduled charging began at 10pm for about 30 minutes. I'll post another graph tomorrow showing what happened overnight.

 

[AlanSubie4Life]

@AlanSubie4Life Just installed today, M3 SR+ 9/2019 build with original battery. Prior to 5:30pm I had it on a 12v supply to test and configure Bluetooth. Car was on until 6pm, then sleeping with Sentry off. Scheduled charging began at 10pm for about 30 minutes. I'll post another graph tomorrow showing what happened overnight.

View attachment 599642

Thanks. Looks like it is doing a bit better than mine! Lol. Odd that open circuit initial voltage for you is lower than it is for me though. Looks like 13.1 or 13.2V for you.

Mine starts its rapid decay from about 13.5V.

This is unplugged, sitting in garage, with everything vampire-like turned off I can think of.

Tonight I turned off Sentry Mode entirely (it was already off at home but now it is really totally off) and unplugged my USB key and USB pass through. And I have disabled the phone Bluetooth. Just to make sure.

In theory it, should now sleep soundly. But I doubt it will. Will leave it unplugged.

Would definitely be interesting to see a “before and after” battery replacement. Seems like this sort of thing should be easily detected by Tesla though, if it is really a sign of a bad battery. Guess I might be contacting Tesla soon, to ask.

Not a huge fan of the car cycling the contactors 20+ times a day.

 

[power.saver]

All that charging is wearing out your main contactors!

Charging is 14.3v. Initial state immediately after charging stops is 13.2v.

Your decay doesn't look right for a healthy SLA battery. It should drop to resting voltage fairly quickly after charging stops, then slowly decline over time. Yours appears to drop somewhat, then have some kind of an RC time constant or step drop. I think your 12v battery is about gone. I wonder if the charging trigger is a voltage level, or a rate of drop to a voltage level. My battery got to 12.9v or so just before the schedule charge boosted it up again. That appears to be lower than yours gets to. We'll see what happens overnight.