Teslafi Vampire Drain

[AndrewGR]

I had Teslafi set up with default settings but no polling between 2400 and 0500. I fully charged my M3 before driving to Birmingham airport leaving it with a respectable 210 miles range showing - I thought easily enough to get me home after some ‘vampire drain’ over my 8 days parking.

As advised on this forum, I resisted the temptation to check the charge state using the Tesla app, but only for 3.5 days. When I cracked, I was shocked to discover the car had lost 46 miles or about 15% of range. It was losing range at about 15 miles per day. At this rate, getting home would be marginal.

I suspected Teslafi, but in case most of the range had been lost in the first few hours after parking, I did nothing before checking again after another 9 hours. Range had dropped by another 5 miles. Time to act. I couldn’t work out how to stop Teslafi polling completely but limited it to 2300-2400. Success! Over the following 48 hours, range has dropped by just 1 mile, far less than the 1% per day I’ve seen quoted.

On the basis of my observations, I’d say Teslafi has been ‘costing’ around 3.5 kWh per day, say 50p per day, or £200 per year at standard electricity tariffs. Seems like a pretty greedy vampire, Has anyone else had similar experiences?

 

[MrBadger]

Sounds like you might have TeslaFi incorrectly setup - I had to change something, but can't remember what.

TeslaFi should only prevent the car sleeping for an additional 30(?) minutes (45 minute total) after each drive - if you are not seeing that, you need to revisit your settings. That will not be causing the drain that you are seeing. Looking at our logs, we lost 0% range over the 12 overnight hours to yesterday morning.

I'm guessing you are not using sentry mode?

I'm not actually convinced mine are optimal as I occasionally miss the start of a drive as its still asleep - never bothered to investigate.

My vehicle is not sleeping! / Knowledge base / TeslaFi

 

[AndrewGR]

Vanilla air
Certainly not in Sentry mode. My Teslafi settings are the same as yours apart from the deep sleep times. Only change resulting in negligible drain now is the Deep Sleep times that only allow polling one hour each day.

 

[MrBadger]

What do the logs show starting approx 30 mins after the car returns from a drive or, as in case below, being woken (at 06:30) via app/teslafi etc?

 

[AndrewGR]

New to Teslafi. How do I see this report?

 

[AndrewGR]

Found it. Here’s a bit of the report from after I parked. Looks like continuous polling.

 

[AndrewGR]

 

[MrBadger]

Presumably still polling after idle time (second last column) reaches 30? Assuming that you would have noticed, then something is stopping TeslaFi from doing what it should. I posted a FAQ above which will give a few things to try. Also the links on the settings/sleepmodes will have ore info.

Without TeslaFi polling, the car should normally* sleep in around 15 minutes.

With TeslaFi using the settings that you said were same as mine, TeslaFi will continue to poll for 30 minutes after drive ended - graph above only shows first 27 polls so it doesn't actually cover the critical point. At 30 minutes, TeslaFi should stop polling. Once the car is not being polled, the car should then naturally sleep around 15* minutes later - so 45 minutes from end of drive to sleep.

* 15 minutes is not a hard and fast rule. The car does talk to mother on a regular (potentially daily) basis. On these occasions, the natural sleep duration may be several hours depending on how much it needs to communicate. However, if delayed, you should see TeslaFi trying to sleep around 30 minutes from the start and at intervals after.

 

[gcsmoore1]

@AndrewGR
As a member in waiting (expecting delivery of a Model 3 to Fairbanks within the week) I am so looking forward to using TeslaFI that I am resistant to your attribution based on a low data count natural history study.

I have the patch kit and Bluetooth transceiver sitting on my desk.

First: How can an sensing app be causing so much power loss unless it's keeping the heater on, which I can see it is not from your report. Could the Bluetooth be drawing some of that?

Second: Might not temperature variations be responsible for the losses? My understanding is that cold batteries report diminished estimated range, estimated being an operative term.

If your losses are so high I would think that you could reproduce them on a shorter time scale.
Please do, as I very much want to use TeslaFI without fear of vampires.

 

[NorfolkMustard]

FYI teslafi uses the Tesla api which is accessed over the internet, you don’t need a Bluetooth dongle for it.

the patch cable and Bluetooth OBDii dongle is for ScanMyTesla

 

[gcsmoore1]

FYI teslafi uses the Tesla api which is accessed over the internet, you don’t need a Bluetooth dongle for it.

the patch cable and Bluetooth OBDii dongle is for ScanMyTesla

Thanks for the clarification/ reminder.
Still looking for vampires.

 

[AndrewGR]

Vanilla Air
Clearly there is more to investigate!
The table continues polling up to 4792 minutes before reporting TeslaFi sleep attempts - presumably when I changed the deep sleep setting. It occurs to me that my nighttime journey to the airport shouldn’t have been logged anyway. It was during what I had set for deep sleep. This is the end of the table.

 

[AndrewGR]

Car is now reporting 156 miles - so without TeslaFi polling the car is losing range at only 1 mile per day. I can live with that. Clearly I need to find out why TeslaFi was previously keeping it awake or abandon its use.

 

[Jason71]

Vanilla Air
Clearly there is more to investigate!
The table continues polling up to 4792 minutes before reporting TeslaFi sleep attempts - presumably when I changed the deep sleep setting. It occurs to me that my nighttime journey to the airport shouldn’t have been logged anyway. It was during what I had set for deep sleep. This is the end of the table.
View attachment 518647

I am not sure about Teslafi, I don't use it, but I use a similar setup. in mine even if a time is set to deep sleep. if the system knows the car is being driven it will still monitor it. you can tell if the car is awake or asleep without waking it so if it's asleep and then wakes up Teslafi may ping it and if it finds its being driven then it may continue to monitor it for the remainder of that journey. That's what I would do. Teslafi may be similar

 

[MrBadger]

Car is now reporting 156 miles - so without TeslaFi polling the car is losing range at only 1 mile per day. I can live with that. Clearly I need to find out why TeslaFi was previously keeping it awake or abandon its use.

One thought that did occur is, as you say you have same settings as me except for iirc scheduled sleep, whether it may be worth turning off the scheduled sleep?

Counter intuitive I know, but wonder if there is a bug/undocumented side effect of using these settings?

Another tack might be to approach TeslaFi community at General / TeslaFi