[Ovmsdev] deep sleep

[Patrick Stein]

I my case I have no problem with the modem working on USB. But to make it clear. The test case I had was the following. I powered the module on 14V OBD and the USB cable connected and not drawing any power over USB. Then I set the voltage to 9V so I would sleep and it told me in the console that the modem was beeing powered down - even though I could see messages send to the modem just before sleep.

When the module was in deep sleep ( going from over 1 Watt down to 490 mW ) and waited a few 1 minute cycles in deep sleep cycles I switched off the powersupply while in deep sleep between cycles and saw on the USB side the power beeing used was 290mW . Turned back the lab power supply for the next wakeup check and did a few cycle measurements while in deep sleep switching the power supply.

Isn’t the modem fully powered down for deep sleep ? I looked at the deep sleep code and it tells me that the modem will be powered down. Might it be feasable to pysically cut the modem power supply and replace that part with a switch that can be switched from the esp32 ? I’m into software not hardware so excuse my simple thinking here. I do understand that low power was not a goal for the board designers. But my use case is different with the car being solely beeing powered by it’s little 12v battery.

Regards , Patrick


> On 21. Nov 2022, at 07:53, Mark Webb-Johnson <mark at webb-johnson.net> wrote:
> 
> It should be the other way around (particular at low power sleep). The usb serial converter chip draws current but only from usb, so when powered via 12v that should be off and not consuming anything. Without a modem and in deep sleep that made a significant difference.
> 
> I guess the difference you are seeing is most likely the modem. That consumes the most power, but in some cases can’t be powered by usb due to insufficient power availability (particularly over low powered usb ports on laptops, etc).
> 
> With the modem powered on, there are significant differences in power consumption depending on whether it is locked onto a cell tower, or searching, and the frequency / technology negotiated. 4G is pretty power hungry.
> 
> It is hard to get much benefit from deep sleep with the modem powered on. Some other systems get around this by putting the modem in sleep mode as well; awoken by a sms message.
> 
> Regards, Mark
> 
>> On 21 Nov 2022, at 2:01 PM, Patrick Stein <4seasons-ovms at jinx.de> wrote:
>> 
>> Hello Michael,
>> 
>> 
>>> On 14. Nov 2022, at 17:26, Michael Balzer <dexter at expeedo.de> wrote:
>>> 
>>> Am 14.11.22 um 10:25 schrieb Patrick Stein:
>>>>> As written in the issue discussion, there's a minimum load from the unswitchable components. In deep sleep mode, that's around 155 mW with the GPS antenna attached, and ~ 100 mW with GPS antenna detached. An option to control the GPS antenna could be using a simple 12V relay, either on SW_12V or on the ignition 12V line, to switch the antenna's GND connection.
>>>>> 
>>>> I’ve never seen such a low value - lowest I’ve seen is 250 mW in deep sleep with all my four modules. 
>>> I did the measurements using a USB power meter, just verified with a multimeter for 12V (11.6V Lipo), which gave slightly higher results:
>>>   • base load in Wifi mode, GPS antenna attached ~ 60 mA / 700 mW
>>>   • sleep, GPS antenna attached ~ 16 mA / 180 mW
>>>   • sleep, GPS antenna detached ~ 11 mA / 130 mW
>>> 
>>> So nowhere near 250 mW. Mark, any idea about this?
>> 
>> I now know why you have different measurements. I bought a USB power meter and it just arrived and it’s now clear. I have just tried one module (v.3.3 with the current edge software) asleep at 9 Volts and it draws around 490 mW (54-55mA) , when I switch my lab power supply off it draws 290mW at 5.1Volts (55mA) over USB. 
>> 
>> So it seems that the module requires a lot less over USB than over OBD.
>> 
>> Best regards, Patrick
>> _______________________________________________
>> 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