I'm not 100% sure if the behavior is the way I mean it. Let's assume I set the idle threshold at 15 minutes and about 15 minutes after I stopped using VT I take the car for a drive. Will VT continue to monitor and poll the car at high rate as the odometer is increasing or will it consider that maybe I'll let the car sleep and therefore ignore for 30 minutes the car driving? That was my main point that I'd like the app to understand that the car has moved since last check and therefore it's not likely it'll go to sleep soon so poll it at high rate to get good graph data. Once the car sits idle for a while (say 10 minutes with no change in odometer and car off) it'll allow sleeping, but check every few minutes if the car is really sleeping. The moment it detects the car isn't sleeping anymore it'll start to record data again.
I'm continuing to tweak how this works, but the basic idea is that while the app is in "Allow Sleeping" mode:
- If the car is in motion, queries will happen at the same rate as if the app were Awake (2 times per minute)
- If the car is charging, queries will happen at the same default rate as if the app were Awake (once every two minutes)
- If the car is awake, but is motionless and not charging, it will not be queried for 30 minutes. HOWEVER, the app will test whether the car is still asleep every 5 minutes.
It will never be fool-proof because I don't want to be hammering Tesla's servers by constantly polling and I do want to let the car go to sleep. I'll be testing my latest tweaks over the next couple of days and then releasing an updated version.
With regard to exporting I'd like a customized start and end dates as I might want only last 1-3 days or what not. But you might have buttons for default ranges that just set the dates then in the start and end fields and still allow customization.
OK, this is on the list of enhancement requests on github.
With regard to the empty lines here's an example:
| 1387967872 | 0 | -57,8 | 284,67 | 0 | 94 | 0 | 22 | 56 | 12.25.13 12:37:52 |
| 1387967894 | 0 | -50,4 | 284,28 | 0 | 93 | 0 |
|
| 12.25.13 12:38:14 |
| 1387967904 | 0 | -57,1 | 283,89 | 0 | 93 | 0 | 22 | 56 | 12.25.13 12:38:24 |
| 1387967924 | 0 | -56,7 | 283,5 | 0 | 93 | 0 |
|
| 12.25.13 12:38:44 |
| 1387967937 | 0 | -50,9 | 283,11 | 0 | 93 | 0 | 20 | 56 | 12.25.13 12:38:57 |
| 1387967956 | 0 | -73,2 | 282,72 | 0 | 93 | 0 |
|
| 12.25.13 12:39:16 |
| 1387967969 | 0 | -51,8 | 282,34 | 0 | 93 | 0 | 21 | 56 | 12.25.13 12:39:29 |
| 1387967985 | 0 | -32,8 | 281,95 | 0 | 93 | 0 |
|
| 12.25.13 12:39:45 |
| 1387968004 | 0 | -30 | 281,56 | 0 | 93 | 0 | 11 | 56 | 12.25.13 12:40:04 |
| 1387968015 | 0 | -52,8 | 281,56 | 0 | 93 | 0 |
|
| 12.25.13 12:40:15 |
So the S_PWR and S_SPD fields are empty while others have readout values. If the cars speed is 0, then it's 0 and not an empty field and those lines are ok, but the ones with empty lines are odd.
Thanks for the example. Not every piece of sample data that I get contains every field. When you see a blank field, it means I don't have data for that field at that point in time, it does not mean the value is zero (as you know). You could interpolate the values, duplicate the nearest neighbor, or ignore those samples completely. I won't unconditionally remove them because they contain valid data. I'll consider other options on export.
- - - Updated - - -
I've been playing around with the notification settings as well. A couple of days ago I got a notification at 10 PM that the MS was unplugged (which it was not.) When I looked on the overview tab (which is where I leave VT by default) the charge cable was not drawn. After I refreshed it, the cable appeared. I've noticed that on other occasions that it looks unplugged, and then updates correctly once refreshed. Something doesn't seem to work consistently in determining the plugged-in status.
I know it sounds vague, and next time I see it, I'll make sure to save the log. I may have the log from the time of the erroneous notification in Time Machine. I'll look when I get a chance.
To deal with the fact that I don't always get consistent readings from Tesla when I look at the Pilot Current, I'm querying it multiple times. I used to do this in rapid succession but am now doing the following: Query. If plugged in, done. If not, wait 20 seconds then query again. If it still looks unplugged, send a notification.
Hi guys. Just a quick question; do you think the 12V battery suffers when VisibleTesla (or phone App for that matter) is running? Could it also be a reason for the 12V problems that some people are experiencing? Anyone know how much or which battery it uses for these features?
The 12V battery problems surfaced on the forums before VisibleTesla existed so I don't think it's cause and effect. Having said that, I have no data to share on the impact of VT on the 12V battery. It would be interesting if someone wanted to do a study of this, but I don't know anyone who has. Perhaps someone using teslams has some insight on this.
Well last night VT correctly queried my car to see if it was unplugged, as well as reported my charge start/stop times and SOC state threshold.
Not sure why it didn't the previous night. I changed no settings on it from the day before.
(Oh and Joe, I responded to your PM)
As to the plugged/unplugged - that's hit and miss as described above. Hopefully the tweaks described above will make it a little more reliable. Not sure what's going on with start/stop and SOC. I'm getting those regularly, but I am on a slightly newer version of the code than what is released.
The feature I wish they would eventually implement (probably not something that visible tesla can accomplish ) is the ability to stretch charging out over a longer period of time and reduce the amperage to provide a smooth charge between the time that the plug-in and the time that you need your car.
for example, it would be great if when I pull into my garage my car says "when do you need to use your car next?" And then I would and put in a time and it would automatically calibrate the start/stop time and/or amperage to provide full charging right up to the specified departure time.
Hi dcopperfield, in the case of VisibleTesla, "they" is me. As you surmised, this is not something that I can control through the API.
