Direct Circuit Monitoring vs. Sense Detection

About one month ago, I configured my second Sense to run DCM on 2 of the biggest 240V power users in my house, the high power wall charger (connector) for my Model 3 and the subpanel for our 5 floor heating loops in our downstairs. To some degree, Sense had already detected Model 3 charging and some of the floor heaters associated with the heater subpanel, but I was suspect of both sets of detections, so I decided to use DCM to double check. Below are 1 week comparison plots for each for the interesting weeks (I excluded 2 weeks when I never had my Tesla plugged in). The DCM data is the black line, Sense detections are the stacked bars (all from Sense exported data).

Just doing the comparison, it looks like there are times when the Dryer detection is really the floor heating (tracking the DCM measurement well), but other places where the Dryer is really the dryer.

**Tesla Charging **
I only have one native EV detection currently, though I had two earlier in the year. I deleted my Model S native detection because it was coming up too low on power and it hasn’t returned yet. The Model 3 detection appeared in March and has been OK but sometimes was falsely triggered. I’m not sure if Sense has tapped into DCM, but the native Model 3 detection in orange, seems to be almost 100% spot on except for the short charging ramps. Nice job Sense - I’ll be watching you.

Floor Heating Subpanel
This one is messier. I have a lot of native detections associated with heating, not all of which were initially associated with my floor heaters - things like Dryer detection that seem to be conflated sometimes. And to make things messier still, I have 4 floor loops, a couple pairs of which use about the same amount of energy/power, plus one that is only 120V so it might be part of the DCM measurement, or not. Once again, the black is the DCM measurement and all of the stacked bars are native detections that may or may not be part of the floor subpanel.

1 Like

Hm, your native detection should not have been impacted by the DCM installation, but this really interesting to see. This could be for several reasons, including a better model for the Model 3 being released. I’ll ask some questions around here.

1 Like

I spoke too soon… My Sense missed native detection of Dec 14th charging (in red), probably because the the on-ramp overlapped with Model S charging. DCM measurement for Model 3 in blue line, Total Usage for the house in the orange line (both measured directly). Sense Model 3 detection in orange bars. You can see the Model S charging in the peaks in the orange line. Sense nailed the detection on the non-overlapped charged later in the week.

Looking in little greater detail, here is the DCM measurement of that last charge vs. the Sense detection for the Model 3. Remarkably similar, including drop-off in the middle. Looks like the Sense detection misses the small increase in power for the second ramp, but that’s OK by me…