0
Answered

Post - 8.1 update, what's the plan?

mkolowich 7 years ago updated by James 7 years ago 18

I just saw the notification that the 8.1 firmware update messes up logging in TeslaFi. What specific areas are affected, and what's your thinking on how and when you can fix them? Is Battery Current the primary way you determine energy consumption? What alternatives are there?


Most of all, how can we in the TeslaFi community help? This is an invaluable app and we don't want to see it damaged

Last 2 days I have 5 superchargers charges on 8.1 if that helps.

Thanks. I think I have everything showing now for supercharges on 8.1 but let me know if you notice any issues.

I'm still seeing an issue with the charge graph post 8.1 - battery current in the graph is still missing. (so is charger current, but that didn't work pre 8.1 either). Thanks!

On 8.1 battery_current was removed so it won't so for any charging sessions. I can access the kW but there's no voltage measurement so I can't calculate it.


Charger current was never reported on Supercharges but I just changed it to show a calculation based on the kW reading and voltage so it should show up for Supercharges now.


Let me know if that makes sense.

Answered

I believe all of the Supercharging issues have been corrected for 8.1 users. I had to calculate the amperage using other fields. It's not quite as precise as it was, but very close.


Sleep sessions appear to be working well too. Not as well as before since some of the sleep attempts fail to actually put the car to sleep but I'll still look for other ways to get the car to sleep faster.


I'm going to close this but feel free to open ticket, bugs or reply here.

+1

James my Model X AP1 with 8.1 is now sleeping again, thanks.

Thank you for fixing charge graphs in post-8.1 readings for non-SuperCharging.


I like that you've kept the battery_current field in everything so far, since that means I can go back historically and look at old data where that variable is available.


SuperCharging doesn't show power level. In the table, there is a field called charger power, that is basically what I want; if that could be graphed, I would like that. Before I saw that, I started searching for other solutions: I have an idea how to calculate it. In case it is a problem of missing battery_current, you can look back to my records for battery_current during SuperCharges to see what that corresponds to in past charges for the data you need. But I'm not sure battery_current is what you're missing. Going from what's available:


charge energy added -- seems precise

charger voltage -- seems precise

battery level -- seems imprecise

charger power -- that's what I want; well, I guess calculating it isn't necessary if it's already there.

power_level is actually the same as the kw in the graph. I never noticed, I was doing a kw calculation to get the data but charger_power is the exact same thing. I added the charger_power to the graph and they were right on top of each other so I removed my kw calculation and replaced kw with the direct numbers from charger_power.


Let me know if that makes sense and what you're looking for.

James FYI my AP1 car on 8.1 does not sleep anymore.

Thanks. I should have new options for sleep mode up tonight.

+1

James, don't know if it is any help to you but my car an AP1 came back from the service center with 8.1. Lots of detailed notes on the changes in the notes section. For some reason seems Iike I am the only AP1 car on 8.1!

+1

in case you're still looking for 8.1 supercharger data: I've upgraded to 8.1 and did a long tour with multiple supercharges today. final charge was DC via chademo adapter, no AC charge yet.

Sorry, I was a bit shocked when I woke up to all the errors.

Thanks for those updates, James. You had some of us worried that the whole app was going to be compromised. Glad you're on top of this

I just posted this on TMC -


TeslaFi.com


So I'm trying to test out some new sleep mode ideas. I set a 8.1 vehicle to poll full data every two minutes during an idle session and the darn thing keeps going to sleep on me. I wake it up and continue to poll every two minutes and it's back to sleep within 10 minutes.

Could be a fluke but it's encouraging.

Under review

Hi.


It doesn't look like things are as bad as I originally thought. A lot of data wasn't displaying correctly originally but I was able to easily fix it. It just so happened that I used the battery_current field in many places to check for good data points. If I needed to look for the last good data point at the end of a session I would many times check to see if the battery_current field had been populated by the API. If it was empty then it would look for the next best data point but since the battery_current field was always empty it ran out of alternatives and broke. Once I changed a lot of the queries to look for 'odometer' instead, everything was back in action.


So far the areas that I've identified which will need work are sleep modes and most likely supercharge efficiency.


Supercharge kWh used is directly calculated by the battery_current and charger voltage since the charger current does not produce any readings during the charge. I can't seem to find any supercharges by any 8,1 users yet so we'll have to wait and see how it looks. Worst case is that supercharging doesn't show effeiciency stats if I can't find a workaround. Not really the end of the world.


Sleep mode is a bigger issue though. I spent months tweaking sleep mode to know the best time to try to put the car to sleep. It ended up being when the battery_current reached 0.0. Now that we don't have access to the field anymore the only other field which shows energy usage during an idle session is the power field which reports kW currently being used. The only problem is the accuracy is in whole numbers. So pretty much when you park the power readying is 0 unless you have heating/cooling on. I'm currently testing a few beta cars to see if the current sleep modes will work monitoring this field. I'm also testing out trying to allow for a set polling time when idling to try to limit vampire loss. For example poll every three minutes instead of each minute. Who knows, it could let the car sleep easier.


For now just reporting bugs will be helpful so I can try to get things that pop up fixed will be helpful. Maybe the battery_current field will reappear in the next software update, who knows. If it's still missing when AP1 cars get 8,1 then it's probably safe to assume it's gone for good.


Let me know if you have any questions.


Thanks!


James