EV 12V battery and AutoPi power consumption

My AutoPi dongle is plugged in to my Kia e-Niro 24/7. I’m using it to send charging info to external services. I want instant access to the latest info and therefore the dongle is up and running most of the time. I’m thinking of having it on 24/7 (no sleep). How will this affect the life time of the 12V battery? With my current settings it looks like the big battery kicks in and gives the 12V battery a push about every 6th hour if no driving activity in between. I don’t know much about batteries. Is this a healthy cycle?


1 Like

I don’t know, but on my Kona there is the “Aux. Battery Saver+” mode to keep the 12V battery topped up -



Presumably the e-Niro is the same.

Also, with the car off ( not driving or charging ) only a limited amount of info is available over OBD. Do you have access to what you need in this case ?

( in Receive charging alerts on your phone ( and smart watch ) I disable sleep only whilst charging - once charging has finished, normal sleep is allowed )

Thanks @plord. Yes, the e-Niro is automatically charging the Aux battery. That is what happens in the picture I included. And I know all about your nice article about receiving alerts :wink:

The question is if the graph in the picture is a wealthy cycle for the aux battery? I don’t know how a normal graph (without AutoPi) looks like. Its hard to display the cycle without the AutoPi :slight_smile:

The current wake/sleep settings are thoose in you article (including the keep alive while charging).

Personally I would be worried about never sleeping ( ie including when not charging and just sat on the driveway ), but it seems reasonable to me to disable sleep whilst plugged in and charging.

Still not answering your question I know … hopefully someone else has more knowledge about the 12v charge cycle.

1 Like

Like any other battery, a lead acid battery (doesn’t matter if it’s sealed, glass, gel, whatever) has a lifetime consisting a certain number of cycles.

Some napkin math follows…

12.8-12.9 is about 100% full at rest.
12.1-12.2 is considered critically low, perhaps 10-20% full.

You’re basically doing an almost full cycle every 6 hours or so.

Typical lifetime of an automotive lead acid battery is 350ish cycles.
That means that at this rate, your battery will last about 2 months.


That conflicts with the information we acquired when researching for the implementation of the power management code in our system, which included speaking with a mechanic.
Do you have any sources on that information?

12.1-12.2v should be around 50% and we have our critical level configured at below 12.3v.
But yes, the battery can be discharged too much, but in this case, it should not as the vehicle will charge it before it gets anywhere near a unhealthy level for the battery.

But I agree with @plord, I would probably also not configure the device to never sleep. Especially when considering that the aux battery in an electric vehicle might have less capacity than the average car battery.

Best regards

Yes , my Kona EV battery died completely after 2 month , last week, we don’t know is autopi but to use without EV custom life system , I wake up every 3 min and go sleep, We must , need urgent update to use only when rpm go or big battery charge . !!!

I can’t find my copy of the Handbook of Batteries right now, so i might be fairly off, but even so, if i’m off by 50%, we’re still talking months, not years.

In general, you can cycle lead acid batteries fairly deeply. The problem with that is that, if they’re not deep cycle batteries optimized for deep discharge, the cycle count does not account for it. Rather, it assumes normal mode of operation, which on most cars is in the range of 12.1 to 12.9-13.0. So even if you could technically go much deeper, your ultimate number of charge/discharge cycles depends on how the manufacturer defined the working range / a typical cycle.

Bear also in mind that these are resting voltages (no load). At lower SOCs, the voltage will tend to drop right under you the moment there’s some load.

On a car battery, a resting voltage of 11.9V is normally considered flat.

That said, in reality, you could keep going all the way down to 11V, or maybe even slightly past it, and if it were a deep cycle battery, you’d be fine. On a regular LA battery, however, that’s not quite considered normal working range, for a number of reasons.

This is why you can’t realistically call a discarge past 12V (which is technically perhaps about 50% SOC for that chemistry) part of a normal, manufacturer declared ”cycle”.

Thanks! Great info for a noob :slight_smile:

Yeah, I agree with that.

The device has multiple safeguards that will shut it down if the voltage goes below 12.3v, but it should not go that far down with normal usage.
And be aware that those thresholds and value are configurable by the user in advanced settings - though with a big warning that it can affect the performance.

Thank you for taking the time to add your input! :slight_smile:

Best regards


Don’t get me wrong, by the way, i wasn’t criticizing the way it’s handled by Autopi at all, i was just objecting (perhaps not very clearly) against the idea of having Autopi configured to be always on, as that will likely severely reduce the aux battery life.


Oh, I’m sorry if my reply gave you that impression :slight_smile:
I didn’t take it as criticism at all, all the input we get from you guys is incredible helpful in what directions to take this, so we really do appreciate when you spend time sharing your knowledge!

Best regards

I think, better use is only when it wake up engine, for ev cars engine and charge.

While i had the Autopi on the bench (to check what’s going on with the update and loss of connection to cloud), i also took some measurements of the current…

It seems to use up around 100mA when idling (not in sleep, just not doing much) and goes up to about 200mA when the CPU and, presumably, modem are doing more work. It’s at about 100-150mA most of the time when awake.

Just putting this here for future reference.

PS: note that this is if i trust my chinese bench supply - it’s probably not super accurate, but i’m ballparking the values anyways.

1 Like

Is the voltage measured by autopi really the same as the 12v battery ? I’m suspecting not.

Maybe also plot the 12V battery voltage found by OBD command to see how it changes over time.

I tested it on my bench power supply, it really did measure quite accurately.

I’m not sure about difference between reported Aux battery PID and actual measurement.
But for measurement, itself (by OBD chip) it’s quite accurate.

This is what I see.

The top plot is the voltage on the autopi, and the bottom reported by OBD query ( and should be the 12V battery ).


I suppose there is some power management between the 12V battery and the OBD2 port.

Isn’t this telling us that the autopi doesn’t really do much to the 12v battery itself ?

It was indeed measuring pretty correctly while i had it on the bench PSU.

Bear in mind that PSU and battery are two different beasts. PSU is a constant voltage source - it will maintain the voltage regardless of the power draw, and there is only one load connected there - the Autopi.

In a car, situation is a bit different - there are multiple loads on the 12V bus, and the voltage is not constant, but dependent on the load. I also suspect the 12V line on the Kona’s OBD port is likely also not directly connected to the battery, but goes through some kind of battery management / switching circuit, as i’ve been getting some very weird readouts on it before (from a Navdy).

Just rechecked. I have the bench PSU set to 14.5V.

power.status returns “voltage: 14.5” so spot on.
Dashboard says 14.239V, because of the averaging i guess?