[Ovmsdev] 12v battery alert threshold?

Mark Webb-Johnson mark at webb-johnson.net
Mon May 21 10:44:05 HKT 2018


Greg,

The roadster has:

APS: Auxiliary Power Supply. Generated by a DC-DC converter in the ESS and producing about 13.5v. There are two APS outputs - one for the PEM and one for the rest of the car.

BPS: Backup Power Supply. 1.x cars have this from the first two sheets. 2.x cars have this from the little 12v battery in the front wheel arch. This is used to power emergency systems (hazard lights, etc) should the APS fail. In normal operation, this is not used at all.

I’ve never really bothered to look much further than that. From my understanding, the BPS is irrelevant to OVMS - we never see it and it is certainly not what is powering that 12v line we are on. I think the 13.5v vs 11.4v is different states of the APS.

Regards, Mark.

> On 21 May 2018, at 9:16 AM, Greg D. <gregd2350 at gmail.com> wrote:
> 
> Hi Mark,
> 
> So I'm really confused how the Roadster is wired.  My own car (a 2.0
> model) had its 12v battery replaced about 2 years ago.  When we did, I
> brought along some wires and a PowerPole connector, and asked the Tesla
> tech if they could wire that in at the battery, so that I could "jump"
> the car if necessary.  As a side effect, that's given me a peek into the
> battery's voltage without having to take half the car apart. 
> 
> The OVMS module is attached to the diagnostic connector under the dash,
> as usual.  It tells me that the 12v supply is at about 13v when the car
> is awake, and 11.4 when asleep.  Uncalibrated, that sounds about right. 
> But the battery (via the jumper cable) stays at 13.7, regardless of the
> car's sleep state.
> 
> At the last service I challenged the service folks to explain what was
> going on.  They couldn't.  We thought that when the car was asleep, the
> only systems alive were being powered by the battery alone, and that the
> DC-DC converter was only active when it was awake.  That would explain
> the OVMS readings, but not that of the 12v battery.  If the DC-DC was
> always active, that would explain the battery readings, but not the OVMS.
> 
> How can they both be true?
> 
> Greg
> 
> 
> Mark Webb-Johnson wrote:
>> Roadster doesn’t have a 12V battery. Ok, it does, but it is only used to power emergency systems and not OVMS or the majority of the car.
>> 
>> The 12V power for OVMS comes from the main first sheets in 1.5 car, or a dc-dc converter in 2.x cars. I haven’t worked out how to get a reliable 12v reference for that, so disabled the alerts for the moment. If we do manage to read the actual 12v voltage, it will be from a can bus message and will have to be handled specifically for Tesla Roadster.
>> 
>> Regards, Mark
>> 
>>> On 20 May 2018, at 2:44 PM, Michael Balzer <dexter at expeedo.de> wrote:
>>> 
>>> Greg,
>>> 
>>> if a car cannot provide the 12v charging flag it has two options:
>>> 
>>> a) disable the 12v alert system by overriding the methods (as Mark did)
>>> b) provide a suitable reference value at startup.
>>> 
>>> Option b can be done by yourself / script:
>>> 
>>> metric set v.b.12v.voltage.ref 12.65
>>> 
>>> …or you can raise the threshold from 1.6 to whatever fits your readings.
>>> 
>>> 
>>> Maybe I'll add a config for the initial reference voltage on startup, as that can also be wrong due to 12v noise.
>>> 
>>> Regards,
>>> Michael
>>> 
>>> 
>>> 
>>>> Am 20.05.2018 um 01:10 schrieb Greg D.:
>>>> Thanks for the overview, Michael.  I have not experienced the v2 module,
>>>> so didn't know how this worked.
>>>> 
>>>> If the reference is taken at boot, then plugging the module in with the
>>>> car awake would give a too-high value vs when it's asleep.  I guess the
>>>> idea is that it should reset itself to a more reasonable value after the
>>>> first charging session, yet I'm not seeing that.  It seems to be stuck
>>>> at 13v, which presumably was what it recorded at the reboot when I was
>>>> playing with the SD card (car was awake and apparently annoyed at the
>>>> time).  I'm on the .005 code, which apparently pre-dates Mark's commits,
>>>> though the commit was 30-April and the .005 build is dated 1-May...  {shrug}
>>>> 
>>>> The algorithm sounds like it is intended to detect a weak battery by
>>>> observing its behavior after being charged, but is this sensing
>>>> available on all cars?  I'm guessing not, and that the Roadster not one
>>>> of them.  For those that aren't, couldn't we just have a configured /
>>>> configurable threshold that would trigger the alert?  Regardless of
>>>> vehicle type, 12v battery health is an important metric to watch.
>>>> 
>>>> Since I'm currently waiting for the modem issue to repeat, I'll just
>>>> live with the alerts for now (vs resetting the module to capture a lower
>>>> threshold value), and see what happens with the .006 software.  The car
>>>> just completed its periodic top-off, generating another pair of events
>>>> (12 restored, followed by 12v alert).
>>>> 
>>>> Thanks again for the explanation,
>>>> 
>>>> Greg
>>>> 
>>>> 
>>>> 
>>>> 
>>>> Michael Balzer wrote:
>>>>> Greg,
>>>>> 
>>>>> Mark disabled the 12V alerts on the roadster firmware some commits ago.
>>>>> 
>>>>> Regarding the alert configuration:
>>>>> 
>>>>> commit 9c1a991a9a27de8afdc9fd408262a2fe0be7aef4
>>>>> Author: Michael Balzer <balzer at expeedo.de>
>>>>> Date:   Mon Apr 30 22:39:05 2018 +0200
>>>>> 
>>>>>   Vehicle: 12V battery monitoring & alert
>>>>> 
>>>>>   New configs:
>>>>>   - vehicle [12v.alert] = 1.6       Voltage drop alert threshold in V vs. reference
>>>>> 
>>>>>   New metrics:
>>>>>   - v.b.12v.voltage.ref             12V reference voltage [V]
>>>>>   - v.b.12v.voltage.alert           Alert status (bool)
>>>>> 
>>>>> 
>>>>> This works like the old V2 12V monitoring, just a bit improved. The reference voltage is taken after max 15 minutes calmdown time after 12V charging stops. The
>>>>> initial reference is the first voltage measured at boot.
>>>>> 
>>>>> Also, the AD conversion needs to be calibrated for each module, to compensate component tolerances. If you haven't done that for your modules yet, their 12V
>>>>> measurements will differ. To calibrate, measure your real 12V level and change config system.adc factor12v accordingly for each module.
>>>>> 
>>>>> Regards,
>>>>> Michael
>>>>> 
>>>>> 
>>>>>> Am 19.05.2018 um 02:26 schrieb Greg D.:
>>>>>> Hi folks,
>>>>>> 
>>>>>> I'm getting alerts from OVMS that my 12v battery is failing, as it's
>>>>>> dropped below the threshold, which is set to 13v for some reason.  So I
>>>>>> get a notification every time the car transitions from awake to asleep. 
>>>>>> 
>>>>>> I don't see a configuration item, nor anything in the OVMS v2 app. 
>>>>>> Oddly, the threshold is 13.2 for my v3.0 proto module, but I can't find
>>>>>> a config item there either.
>>>>>> 
>>>>>> How do I change the threshold? 
>>>>>> 
>>>>>> Greg
>>>>>> 
>>>>>> _______________________________________________
>>>>>> OvmsDev mailing list
>>>>>> OvmsDev at lists.openvehicles.com
>>>>>> http://lists.openvehicles.com/mailman/listinfo/ovmsdev
>>>> _______________________________________________
>>>> OvmsDev mailing list
>>>> OvmsDev at lists.openvehicles.com
>>>> http://lists.openvehicles.com/mailman/listinfo/ovmsdev
>>> -- 
>>> Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal
>>> Fon 02333 / 833 5735 * Handy 0176 / 206 989 26
>>> 
>>> 
>>> _______________________________________________
>>> OvmsDev mailing list
>>> OvmsDev at lists.openvehicles.com
>>> http://lists.openvehicles.com/mailman/listinfo/ovmsdev
>> _______________________________________________
>> OvmsDev mailing list
>> OvmsDev at lists.openvehicles.com
>> http://lists.openvehicles.com/mailman/listinfo/ovmsdev
> 
> _______________________________________________
> OvmsDev mailing list
> OvmsDev at lists.openvehicles.com
> http://lists.openvehicles.com/mailman/listinfo/ovmsdev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openvehicles.com/pipermail/ovmsdev/attachments/20180521/512fdffd/attachment.htm>


More information about the OvmsDev mailing list