Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

What are those 12V voltage spikes when idling?

This site may earn commission on affiliate links.
I performed some idling consumption tests before by Scan my Tesla but the CAN bus always shut off after about 1 or 2 hours making it impossible for long duration test.

Several days ago, I started the test again but this time I measure the voltage of the 12V battery directly and here are some results:
V12-on-on-e.png

I set Energy saving off, Always connect on, the car was parked in 5C temperature (underground parking lot) for 4 hours before the test.
When I locked the door and left, the DC-DC was immediately off, the 12V battery last for less than 5 hours before the DC-DC kicked in, and remained charging/supplying for the rest of the test. The DC-DC current is measured by monitoring the voltage drop of the DC-DC fuse (F1, 250A) which is calibrated against the value from the CAN bus.

At the end, the battery dropped by 1.77kWh in 16 hours, or 2.8% per day, equivalent to 110W of power.

Another test result. This time, I set Energy saving on and always connected off.
Other conditions are the same as the last, which are:
Temperature 5 degree C
Firmware 2017.50.3
Test car X 100D build July 2017, mileage 4700km/2940mi
Test equip: Graphtec GL220 data logger
Sampling interval: 2s
CH1 range 20V
CH2 range 20mV
X12V-GL220.jpg

V12-off-off-e.png

Test results:
DC-DC not started for the whole duration of the test.
Rated range dropped only 1.61km (562km when charged full), which is 1.61/562*94.1 = 0.27kWh, a drop of only 0.27% per day.
(It was 532.69km when the test start and 531.08 at the end, both recorded by Scan my Tesla)

However, there are voltage spikes of about 0.13V, and it happens every hour.
Here is one of the detailed spike. It seems to me that the 12 battery is being charged for about 15 seconds every time, but I measured *no* DC-DC current which normally is the source of charging. Any ideas why?
V12-off-off-d.png
 
Last edited:
My guess is that whatever module is turning on every hour has its own 12V charge storage device (e.g. battery or supercap). From your graphs, the spikes are only there when the voltage is low... and they get smaller over time.

This is consistent with a module getting recharged every hour. When the 12V inverter kicks up to ~14V, the module gets charged. When the main 12V battery droops, the module briefly imposes it's voltage on the 12V main to check to see if it can get charge from the main 12V battery. Each time it does this (and keeps the module powered) it loses a little charge and the spike gets smaller.

And, if my guess is right, then what module is doing this? Something that needs to keep working if 12V power goes away?
 
  • Helpful
Reactions: lymex2018
It just come to me: could that spikes be the temporary interruption of the power consumption?
I've no access to the positive terminal of the 12V battery nor the battery wire otherwise I can measure the current (in and out).
 
It just come to me: could that spikes be the temporary interruption of the power consumption?
I've no access to the positive terminal of the 12V battery nor the battery wire otherwise I can measure the current (in and out).

That was my first thought. Something that normally consumes power is shutting off for a few seconds every hour. Can't think what might do that.
 
One more test of 16 hours before the long vacation.
This time, I set Energy save on and Always connected on, and wake up the car 6 times during the test indicated by small triangle.
V12-off-on-e.png

The small positive going spikes persist when DC-DC is not working, but there are three large negative going spikes that correspond to the time when I use the app. The discharge of the 12V battery seems goes to the DC-DC converter because there are negative DC-DC current.
 
More test, more spikes.
56h-on-on.png

This time, I tested for 56 hours when Energy saving on and always connected off. The hourly spikes remains the same when DC-DC is off, but there are three more types of spike:

1. DC-DC current spikes.
These spikes seen previously but this time become regular at precisely every 6 hours. Looks like some device wake up for 1.35 minutes periodically.

2. DC-DC negative current spikes when off.
It's very bizarre because it means the battery gives out current back to DC-DC, and it lasts 1 minute every time.
56h-on-on2.png 56h-on-on3.png

3. small negative spikes every 10 minutes when battery voltage is around 12.6V.

The DC-DC also turns off every day at about 23:00 to 5:00.
 
  • Informative
Reactions: hiroshiy
First of all: thank you, @lymex2018, for testing this and sharing your results with us. :)

My guess is that whatever module is turning on every hour has its own 12V charge storage device (e.g. battery or supercap). From your graphs, the spikes are only there when the voltage is low... and they get smaller over time.

This is consistent with a module getting recharged every hour. When the 12V inverter kicks up to ~14V, the module gets charged. When the main 12V battery droops, the module briefly imposes it's voltage on the 12V main to check to see if it can get charge from the main 12V battery. Each time it does this (and keeps the module powered) it loses a little charge and the spike gets smaller.

And, if my guess is right, then what module is doing this? Something that needs to keep working if 12V power goes away?

Really going on a limb here : could it be the security package ? It was for some time not available in the US but available in the EU (no idea about China). Maybe it is now in included in every car ? That package included some sort of intrusion sensors and a siren, but more importantly, it had its own backup battery.

Some links : "Security Package?"
Security Package option in BG | Tesla
Security package now included in EU-models
 
  • Informative
Reactions: lymex2018
Finally, I got firmware update to 2018.6.1 (was 2017.50.3), and I was able to test the battery current directly by using a clamp meter (Wavetek CT238) attached to the negative wire of the battery.
Again, set Energy saving on, always connect off, and apply a 10A constant current load to the battery.
20180307-off.png


Those large negative spikes are apparently due to that DC-DC was trying to turn off after it on for about an hour, and turn on again by the low voltage of the battery (12.1V to 12.2V). However, there was one very large negative part after 12:00.
20180307-off-d.png
As can be seen from the graph in magnified scale, some device turned on and off several times, the combined current reached 33A and the voltage dropped to 11.5V without triggering the DC-DC.
 
  • Informative
Reactions: hiroshiy
12V Battery current measurement when Energy saving on and Always connected off at no load.
20180309.png

1. The battery makes a "cycle" in about every 40 hours. Consumption and charging in this period is about 4Ah both, which counted for only 7.3% of a full cycle (1 cycle = 33Ah) per day. In another word, it takes two weeks to complete one full cycle. The average current of the 12V is only about 0.11A when DC-DC is off.

2. In the whole test period, the DC-DC turns on for less than 4 hours and charges 56Wh to the 12V battery(= 4.0Ah * 13.9V), and there is also a mysteries 5.3A current consumption (73 Watts, 271Wh). The total consumption of the main pack by the DC-DC is about (56+271)/75%=436Wh, less than 800Wh of the actual value that obtained from Scan my Tesla data at the beginning and the end of the test. Therefore, there must be other consumption devices apart from the DC-DC converter when the car is parking.

3. The consumption of 5.3A current is only on when DC-DC is on, but there is a delayed off that makes the negative spike of the 12V battery current curve. Also, this consumption is on before the DC-DC is making another negative spike.

4. Each hourly spike (mentioned at the beginning of the thread, when the DC-DC is off) is battery being charged at 0.5A for 18 seconds.
20180309a.png

Could the source of the charge be the small DC-DC converter in the large pack?

5. The whole test lasted for 58 hours (actually ended at 8:30, the recorder ended earlier at 7:06 because the battery was out). The main pack dropped by 0.8kWh, equivalent to 14W of average power.
 
  • Informative
Reactions: hiroshiy
I tested again several times when Energy saving off and Always connected on, and I think I found the reason of the DC-DC current spikes of every six hours: it is the remote access.
Not done by me, but and app call has the same spikes either when DC-DC is on or off.
20180320.png

Four remote accesses a day:
23:00, immediately after the DC-DC is off, there is a remote access
05:00, this remote seems consume more power, but the DC-DC is not on, even the battery voltage drops to 12.2V
11:00, only a spike when the DC-DC is on
17:00, same as above
 
  • Informative
Reactions: hiroshiy
Spikes when trickle charging.
trickle-charge.png

The trickle charger is a constant current constant voltage source set to 13.5V and 0.8A.
Setting energy saving on, I run the test for 80 hours and there is no turning on of the main DC-DC at all. However, because the 12 battery is in near-full state, the hourly spike become larger and larger. At the end, the total consumption is only 0.9kWh which equivalent to 11W of average power. The output current of the CC-CV is around 0.05A which include the trickle charging and the load consumption. Therefore, the load is less than 1 watt and some source (probably the BMS) consumes about 10 watts.

Here is the charger circuit diagram:
trickle.png
 
Spikes when trickle charging.
View attachment 289916
The trickle charger is a constant current constant voltage source set to 13.5V and 0.8A.
Setting energy saving on, I run the test for 80 hours and there is no turning on of the main DC-DC at all. However, because the 12 battery is in near-full state, the hourly spike become larger and larger. At the end, the total consumption is only 0.9kWh which equivalent to 11W of average power. The output current of the CC-CV is around 0.05A which include the trickle charging and the load consumption. Therefore, the load is less than 1 watt and some source (probably the BMS) consumes about 10 watts.

Here is the charger circuit diagram:
View attachment 289918

With no DC-DC, is that a load's kickback making the voltage spikes? Very strange.