Kasa integration stopped working

I’ve been really liking my Sense+Solar, and had added a few of the supposed Kasa KP-115 and HS300 smart plugs/power strips. It was all working fine, but recently, Sense stopped reporting data from the plugs (I verified this against what the Kasa app reports, and also Sense became unable to control the outlet on/off function).

I can use the plugs just fine with the Kasa app, have disabled and re-enabled the Sense integration and have power cycled the Sense monitor more than once. I’ve also rebooted my WiFi more than once.

Any ideas of what I can try?

How many is “a few” KP115s and HS300s ?

Currently three HS300 power strips (a strip counts as one device) and two KP-115 plugs. Well below what should be a limit as I read it.

Actually right at the limit - 3 x 6 plus 2 = 20 outlets… support@sense.com might be able to help to see if that is the issue.

1 Like

It shouldn’t be. The Sense documentation specifically states that the power strip (HS300) counts as one device regardless of the number of things plugged in.

The definition of the limit has changed recently to clarify HS300s. You still should be OK per the guidelines, but that’s a little network dependent too. Just trying to think through possible issues.

Is there a limit to the number of smart plugs I can connect to Sense?
The number of smart plugs is limited primarily by Wi-Fi network bandwidth. Typical Sense users can use about 20 smart plugs without issue. Note that each plug of the TP-Link HS300 smart strip counts as 1 smart plug, regardless of status.

1 Like

The level of strain the smart plugs put on a Wi-Fi network has been a curiosity of mine. My internet performance can be poor at times, but I have no idea if this is a problem with bandwidth on my local network or with my ISP (Cox). Have you seen anyone experiment with this or shown there to be an appreciable effect on how other devices (laptops, phones, etc.) perform?

This is interesting, as I always assumed the 20 plug limit was a hard limit set by Sense. Would setting up a secondary Wi-Fi network dedicated to smart home devices be a good way to both ensure the plugs don’t hinder Wi-Fi performance on other devices, and increase the max plug limit? Does this soft limit mean the max # of plugs is theoretically infinite if the network had enough bandwidth?

Regarding the potential bottlenecks… Is data from the plugs routed directly to the cloud, and then from the cloud to the Sense apps (Wi-Fi, router, and ISP bandwidth potential bottle necks)? Or does the plug data first have to be sent to the Sense unit in the utility panel (making the radio of the Sense unit another potential bottleneck)?

How does one know when they’ve maxed out the # of smart plugs they can use? i.e., when the Sense FAQ post mentions “Typical Sense users can use about 20 smart plugs without issue”, what are the issues one would expect to see? Degraded performance on all network devices, or just on how the plugs perform? In other words, is the amount of total bandwidth the smart plugs can consume capped or denoted as a low priority to prevent disturbing regular network usage?

I do not think a limitation is a result of a local WiFi network. My HS300 uses about 10MB per hour of local WiFi traffic; this also includes Home Assistant querying it. I have a handful of HS105s (which do not include power information, and are not included in the Sense app), and they only use about 1.25MB/h with Home Assistant status updates.

My guess is that the local Sense device itself only has so much CPU power to process plug status while retaining an acceptable level of performance for what it normally does.

1 Like

It sounds like your belief is that the power consumption data from the smart plugs is routed to the Sense device, not to a cloud server (managed by either Kasa or Sense). I assume the numbers you provided for local Wi-Fi traffic are for data transferred only between the smart plug and the router/AP. Do you happen to see Wi-Fi traffic between router and the Sense device that correlates to the traffic from the plugs (i.e., 10MB received from the plug results in a 10MB transmission to the Sense device shortly after)?

Do these data numbers give an accurate representation of the total demand the plugs placed on the Wi-Fi network, or does it omit the “behind the scenes” operations like handshaking that are required to transfer this 10MB of data? In other words, could these plug still be a significant strain on the Wi-Fi network even though the amount of actual data transferred is small. Your findings seem to contradict the statement of “The number of smart plugs is limited primarily by Wi-Fi network bandwidth” in the Sense FAQ shared by Kevin above. Once again, I’m a novice in networking so I apologize for any lack of understanding (not questioning your findings, just trying to better understand what they mean)(for reference, I’m a design EE but work mostly in the realm of physical hardware / Hi-Rel electronic packaging).

(Lower priority) Out of curiosity, do you see any change in the HS300’s data transmitted per hour vs # of outlets on strip utilized, how much power device(s) are using / how frequently measured power use is changing, whether you have the power meter for one of the devices open for the hour, etc.? Or does this 10MB remain fairly constant?
Do you have any HS110s or KP115s? It would be interesting to see if the data rate for a single plug is 1/6th the rate seen on an HS300 fully loaded with 6 devices plugged in.

You’re right, this is purely traffic to/from the AP. Based on the other devices on my network, though (e.g., Nest cameras putting out 1GB per day, each, to the internet), I don’t think local WiFi is the bottleneck for smart plugs, at least based on throughput alone. It’s possible some APs are unable to handle a large number of clients routinely transmitting data.

These numbers are coming from my UniFi AP. I’m not sure what layers of the network stack it’s including.

I haven’t looked into the HS300 data in that level of detail, yet. No HS110s or KP115s, either.

1 Like

A few additional thoughts.

  • I agree that the main issue is Sense monitor processing power. The “too many plugs” failure mechanism I observed was the plugs dropping out, then the monitor itself dropping data from the main CTs until it eventually went offline from task overflow.
  • Another possible issue might be peak traffic. The Sense monitor sends out a broadcast “emeter real-time” request on the local network every 2 seconds. All the appropriate listening Kasa devices answer back, all probably in a fairly tight timeframe, so there’s lots of Sense smartplug traffic during that fairly short period.
  • I don’t think there is a difference if the HS300 plug is unused. The data is still sent back.
  • There’s some code here that lets you play with pulling power data off a Kasa device. If you turn on the debug options you can actually see the messages being sent back by the smart plug.