12v battery issue, Tesla unsatisfactory response

[webbah]

View attachment 592725 View attachment 592726 To anyone running the Ancel or Quicklinks Battery Monitor (@webbah @scottf200 and @Watts_Up ) have you noticed inaccuracies in the Voltage percentages from reality (I.e. side by side testing with a voltmeter and what you get with the app)? Was going to pull the trigger on one however there were numerous reviews on both with reviewers posting photos of the side by sides and there were big differences. In essence they (and I) were saying what’s the point if (a) the voltage percentages are not within the variance tolerance spec the manufacturers list and (b) quite often the monitor would list a trigger for a low voltage warning and the owner would voltage test with an actual meter and it would be fine? (I.e. the Voltage output indicated thus triggering a warning would be way less than actual voltage in reality after a vmeter test).

Any input appreciated. Really would like this unit but want it to be accurate...as I have a enough to worry about without constant false readings. After all I’m purchasing to alert me to “pre-failure” of the 12V so I don’t get stranded.

Also is a power draw oh 1 mA - 1.5mA enough to slowly drain the 12V when parked let’s say for 4 days and NOT plugged in and charging?


Regards,

Ski

I haven't seen this personally, but I am only comparing what SMT and this show. And I'm more interested to know if voltage drops significantly and be alerted via a push message when I approach the car. I wish there was a way to get the 12v info from the Tesla API and into a Grafana dashboard with Teslamate, but to my knowledge this is not possible.

Here's a few screenshots of my app for the last day. I'm not close enough to the car so it will show disconnected but the data is there. You can see the power fluctuations between when the car is at rest vs driving/charging. Keep in mind I have an Ohmmu and not the OEM AGM battery. So mine maintains stable voltage much better than the OEM. I did that as I run 2 12v devices constantly (Blackvue DR900S dual channel HD cameras and Teltonika RUT850 Wifi/LTE Automotive router, each pulling 5 watts).

 

[AmokTime]

I’ve seen a lead-acid battery explode once while being jumped. I was about 10 or 11. It was a van (I was a passenger in it). The hood was up, obviously, but there was a space between the base of the windshield and bottom edge of the hood, a line-of-sight where you could peer into the engine compartment from inside the van. The battery detonated, spraying a significant amount of acid up onto the lower windshield. Every time I jump a car or change a battery I think about that in terms of safety.

I’ve also seen jumper cables melt from too much current with a *really* dead battery (probably shorted internally) and an overly-persistent attempt to jump it.

12 - 13V won’t shock you, but sulfuric acid will burn you, possibly severely (especially your eyes), and a lead that has had too much current flow through it can burn the crap out of you.

 

[one2many]

Well I blew up a car battery, one hell of a bang, jump starting it. Lucky I got out
without harm. One shorted cell and it went up not sideways. So it can happen.

 

[Skione65]

And I'm more interested to know if voltage drops significantly and be alerted via a push message when I approach the car.

You’re quote above is exactly what I’m talking about though. What’s the point if the unit is “inaccurate” and when you walk up to the car it gives you the ‘low voltage trigger warning’ due to the inaccuracies if in fact the voltage percentage is actually higher in reality and ok?
On another note what are the comparisons showing between your SMT and this BM? Are they close, similar, spot on? Just curious. I need to get that SMT and install already.

Regards,

Ski

 

[webbah]

You’re quote above is exactly what I’m talking about though. What’s the point if the unit is “inaccurate” and when you walk up to the car it gives you the ‘low voltage trigger warning’ due to the inaccuracies if in fact the voltage percentage is actually higher in reality and ok?
On another note what are the comparisons showing between your SMT and this BM? Are they close, similar, spot on? Just curious. I need to get that SMT and install already.

Regards,

Ski

Almost identical. 0.01-0.02 difference. Solution works great for me. I don't have access to a home charger as I live in an apartment so the car is almost always unplugged. Never had any issues to date and the car is now going on 15 months old or so. You need to find what works for you and this gives me peace of mind. It allows me to make decisions about enabling Sentry, using the kill switches I installed to turn off the Blackvue or Router, etc. Especially in colder winter.

 

[scottf200]

I haven't seen this personally, but I am only comparing what SMT and this show. And I'm more interested to know if voltage drops significantly and be alerted via a push message when I approach the car. I wish there was a way to get the 12v info from the Tesla API and into a Grafana dashboard with Teslamate, but to my knowledge this is not possible.

Here's a few screenshots of my app for the last day. I'm not close enough to the car so it will show disconnected but the data is there. You can see the power fluctuations between when the car is at rest vs driving/charging. Keep in mind I have an Ohmmu and not the OEM AGM battery. So mine maintains stable voltage much better than the OEM. I did that as I run 2 12v devices constantly (Blackvue DR900S dual channel HD cameras and Teltonika RUT850 Wifi/LTE Automotive router, each pulling 5 watts).

You’re quote above is exactly what I’m talking about though. What’s the point if the unit is “inaccurate” and when you walk up to the car it gives you the ‘low voltage trigger warning’ due to the inaccuracies if in fact the voltage percentage is actually higher in reality and ok?
On another note what are the comparisons showing between your SMT and this BM? Are they close, similar, spot on? Just curious. I need to get that SMT and install already. Regards, Ski

Re: Battery Monitor 2 device on 12v.
a) That app was using a lot of my phone battery so I turned off the 2 notifications and set the app to not use battery when in the background via the Android OS. Working great still and when I'm within 20ish feet of my X then it still connects and uploads the data from the hardware device to my phone.

b) I've been testing since yesterday at 11:30am and had my fridge on as well as the blackvue dashcam which is always on. I also drove to pickup takeout last night (> 30 minutes).
My X seems to be working different than the 3 or I am missing something as I have yet to see a spike in voltage from the DC-DC converter 'charging' my 12v OEM battery.
There seems to just be a steady state of maintaining the voltage as it only goes from like 13.19v to 13.24 volts constantly. No major jumps up or down.

UPDATE: well my car is also plugged in so perhaps that is why the voltage is steady? will test.

Perhaps the X and 3 treat the 12v charging differently. I should point out that my X doesn't like to sleep tho.
UPDATE: I see @AlanSubie4Life points out in the next post Teslas (3 and others?) will charge the 12v after waking up. Since mine doesn't want to sleep perhaps that is why I see the steady voltage.

2nd graphic below has a red rectangle showing a steady voltage from SMT although amps jump around with the duty cycles of the fridge (this was from a few weeks back but shows steady voltage).

Original post of the below with larger graphics in this post and the one following it. Fridge/Freezer that fits the Model X is on sale

 

[AlanSubie4Life]

Also is a power draw oh 1 mA - 1.5mA enough to slowly drain the 12V when parked let’s say for 4 days and NOT plugged in and charging?

First, you only need to worry about whether the car detects the load you put on it as a bad battery symptom. Not sure how this works, but I think there is some such detection of unmetered draw (current drawn from the battery not going through a path which powers things when in sleep - remember all those other loads are solid-state fused so in theory they can measure all of those loads).

I doubt 1mA is enough to trigger a bad self-leakage detection algorithm (internally leaky batteries leak a lot more than that, see below to see general order of magnitude needed to make a battery bad).

Anyway, that speculation aside:

The car will charge the battery every few hours or so, whenever the car exits sleep. When the car is in sleep, it is drawing 8-12W or so. That is about 660-1000mA. (This is why the battery wears out quickly - 12V batteries really aren’t intended to be drawn down from what I understand - especially batteries that are not AGM - the Model 3 battery is *possibly* not AGM based on available evidence, see above posts.). This draw is obviously way way higher than nearly every other vehicle you could buy - other vehicles can sit for up to months with no effect on the 12V. Anyway, extra 1mA would likely never be noticed unless it has extremely accurate detection of unmetered current flow (it won’t).

It’s a 65Ah battery, so to kill the battery, you can draw 1000mA for 65 hours (very approximate), or 1mA for 65000 hours (7.4 years). Again, all very approximate. Obviously the battery would actually be dead long before 7.4 years.

 

[iwannam3]

Has anyone ever been injured from a hydrogen explosion from a lead acid battery?

62 of the 97 people on the Hindenburg survived so the odds were in your favor.
Seriously though has anyone ever been injured by a hydrogen explosion from a lead acid car battery? Seems so unlikely.

In HS auto shop they had front clips of cars with engine & dash. Teacher would put in a faulty component and you had to get the thing running. A kid was having trouble and kept cranking and then yanked the charger cables off in frustration. All 6 cell caps blew off as the hydrogen exploded! Sounded like a burst of machine gun fire.

 

[SageBrush]

ICE cars have a state wherein the engine is off but the 12v supplies juice to accessories. IIRC it is called 'ACC' and used to be enabled in key systems by turning the key one click.

Can we simulate the same in a Tesla ? Would disconnecting the DC/DC connector do the trick ?

 

[scottf200]

ICE cars have a state wherein the engine is off but the 12v supplies juice to accessories. IIRC it is called 'ACC' and used to be enabled in key systems by turning the key one click.

Can we simulate the same in a Tesla ?

"Camp mode" the equivalent to the olden days car accessory key position (The second position is the ACC/ACCESSORY position, which allows you to use your radio, windshield wipers, and other accessories while the engine is off. )

2019.40.50.1 (Software Update) | Teslascope

Would disconnecting the DC/DC connector do the trick ?

I don't follow this line of thinking. Do you mean the contactor connector under the seat or something else? Seems like disconnecting anything like that shuts down the 12 'rail' all together and powers down all 'subsystems' (threads). Tesla Model 3 Hard Reset | Mountain Pass Performance

 

[SageBrush]

I don't follow this line of thinking. Do you mean the contactor connector under the seat?

Yes

I want to drain the 12v for the capacity test, or perhaps drain the 12v surface charge for an OCV test. I don't want traction battery energy.

 

[AKinDC]

Maybe I missed this somewhere in the thread, but why does a car that’s essentially a battery on wheels need an additional 12v battery?

 

[SageBrush]

New Maybe I missed this somewhere in the thread

yes

 

[AlanSubie4Life]

Maybe I missed this somewhere in the thread, but why does a car that’s essentially a battery on wheels need an additional 12v battery?

Yes, you missed it. Learn to use the advanced search options!

12 Volt Battery Compendium – TeslaTap

Take these articles with a grain of salt, but they are largely correct.

 

[lUtriaNt]

Low cost way to monitor your 12V Battery

no. i used one of these on my old combustion vehicle and the data from the app measured correctly with the dashboard. on the tesla, i get reliable results as well. ive been using this particular model for over a year on my old car and a few months thus far on model 3. i recommend this one.

View attachment 592725 View attachment 592726 To anyone running the Ancel or Quicklinks Battery Monitor (@webbah @scottf200 and @Watts_Up ) have you noticed inaccuracies in the Voltage percentages from reality (I.e. side by side testing with a voltmeter and what you get with the app)?

EDIT - the draw is not going to eat into your car if it sits whilst not being charged. besides, the difference here is you have a huge traction battery that would keep the battery charged anyway, so id not worry about the teeny tiny barely virtually not there draw from this gadget. again i've been using this particular model for some time across both cars.

 

[scottf200]

Low cost way to monitor your 12V Battery

no. i used one of these on my old combustion vehicle and the data from the app measured correctly with the dashboard. on the tesla, i get reliable results as well. ive been using this particular model for over a year on my old car and a few months thus far on model 3. i recommend this one.

I use that brand for an IR battery tester (I think it is rebranded tho -- model BA101).

FYI -- HOWEVER that device is certainly just a rebranded one from the same company. It is $10 more and the software looks the exact same. The hardware looks the exact same as well except for the label. https://smile.amazon.com/gp/product/B01MT4583U/ref=ppx_yo_dt_b_asin_title_o04_s00?ie=UTF8&psc=1 .

 

[lUtriaNt]

yes they are mostly rebranded but ive had this one for over a year in experience and its been rock solid.

photo to show the module is pretty spot on:

i have an ohmmu battery.

 

[Phlier]

In Model 3 (other vehicles may be different!) we don’t know this to be true, AFAIK. The battery manufacturer’s document suggests it is not AGM (see above). MF type.

Circumstantial evidence: Tesla charged someone only $120 to replace the battery out of warranty (see above). That is too cheap for a small AGM battery as far as I can tell.

Anyway, one of these days we will find out whether or not it is AGM, with some sort of proof. My current bet is it is not AGM in Model 3.

Ah, good to know. I think me thinking that it was AGM was a case of seeing it written so many times that I assumed it to be true. One of the true pitfalls of reading stuff on the internet that I try to be careful to avoid.

 

[smatthew]

I've been trying to figure out what Tesla could do to predict battery failures.
I figure, first, you have to look at what the failure modes of the battery are, and how you can detect them.
Top 5 battery failure modes:

1. 31.8% Plate / Grid Failure.
   This is a normal function of aging of a battery, but can be accelerated by overcharging. This failure means your battery has reached it's normal lifetime. Often detectable as rising internal resistance
2. 23% Short Circuit. Heat increases frequency of short circuit to about 33%. Can sometimes be predicted by lowering internal resistance.
3. 21.5%Worn Out/Abused
4. 18.1% Serviceable
5. 3.6%Open Circuit. Often detectable as rising internal resistance.

But how do you measure internal resistance? You put a load on the battery and measure the voltage. Gee, guess what the car can do? Control it's 12V loads. The car could very easily be programmed to do a load test on the battery. It may already do it.

I still bet, based on Tesla's prior behavior, that they're trying to get a neural net to detect 12v battery failures. But the charging system and battery is different on the 3 than the S or X, so they have had to wait for data to be collected from cars with failed batteries.

Either that or they're going to upgrade VCFront so it can do Electrical Impedance Spectroscopy. My money is on a neural net. Just think of how long the rain detector net was abysmal.....

 

[AlanSubie4Life]

Ah, good to know. I think me thinking that it was AGM was a case of seeing it written so many times that I assumed it to be true. One of the true pitfalls of reading stuff on the internet that I try to be careful to avoid.

It definitely would be nice for someone to disassemble the top of a battery as suggested above, so the whole world could know for sure.