Table of contents

RoadTest: Enroll to Review the Schurter Smart Connector DT31

Author: hlipka

Creation date: 29 Sep 2023

Evaluation Type: Connectors & Cable

Did you receive all parts the manufacturer stated would be included in the package?: True

What other parts do you consider comparable to this product?: A Tasmota Smart plug

What were the biggest problems encountered?: Fund a bug in the cloud dashboard

Detailed Review:

So first I would like to thank Element14 and Schurter to making this road test possible. Without this, I would never have looked at this device.

The road test asked mostly for first impressions, and how easy it is to get going with the DT31. So let's dive right into it.

The DT31 is described as a 'smart connector', and that describes it quite well. It allows you to remote control the device attached to it, as well as to measure its state. You might already have see similar devices called 'smart plugs', designed to be used in your smart home installation. In contrast, the DT31 is designed to be used in more industrial (or at least 'rough') installations. You see this already by the choice of its connector: a C14 socket is not used that often in your typical home. Fortunately for us, quite a lot of lab gear and measurements devices come with it.

First impressions

In the road test package, I found

• the DT31, nicely packaged in its box
• a small Victorinox Swiss Army knife, Schurter-branded (the are a Swiss company, after all)
• and a 'Welcome to the road test' leaflet

Unfortunately you cannot use the knife to open the package :( I think the leaflet was a great idea, for me its the first time I ever got one for a road test. It contained a welcome letter, which again outlined the asks Schurter wanted to get answered from the road test, and all contact information I might need. There were also all of the data sheets and information material for the DT31.

{gallery}Unboxing
Package contents
The DT31 box
Front of DT31
Back of DT31

The packaging of the DT31 is very good, it comes in a nice box, sturdy enough so it should not be damaged even during rough shipping (but then, you never know...). It bigger than I had expected, but then also the DT31 is bigger than you would think (its 10x10 cm). The DT31 looks rather good (especially for an industrial device..), and also sturdy enough to withstand some abuse (maybe with the exception of the sticker with the serial number and the MAC address on tha back, I think they will be gone soon). What I don't really like: the cable with the C13 plug is fixed to the DT31. While I can understand that decision (because it make accurate measurements easier), it adds some limitations:

• you are required to place the DT31 physically close to the dive you want to control (its 80cm long)
• the cable cannot be replaced should it break (it looks rather sturdy, and has nice bend protection, but at one point it will break)
• it limits you to use devices with a C14 socket

The data sheet tells you about this, so it should not be a surprise to buyers. But I would have preferred to see a socket on the DT31. I like my flexibility...

Initial setup

The initial setup was very simple. Hint for future users: go to the Schurter cloud site first and register a new account first, you will need it when using the phone app for the first time. Use a password manager for easy transfer of the login information (you can register from the app, but creating a secure password on a phone keyboard is just painful IMHO). Apart from that its rather smooth: open the app, log in, follow the instructions. For the DT31 you just plug it in and press the button. It will be found automatically by the phone, and is registered. At least as long as you are on an Apple device. For Android its a bit more painful:

• the QR code printed everywhere does not work (it has some prefix text before the actual URL, which you need to remove manually. Make sure the URL is www.schurtersmart.com, schurtersmart.com just does not work at all)
• registration then works fine (I did register my device that way)
• but any click on the second icon (which should show the device data and the graphs) the app crashed for me reproducible (this  might be just my phone, but its on a current Android version and should be capable enough to handle the app)

The app itself seems to be built for a different device originally. It shows two outputs for the DT31 (which has just one). Also, toggling the state of the outputs does nothing - you need to turn the DT31 itself on and off in the app to switch the connected device.
Once I installed the app on my iPad the DT31 appeared there as well, and there the second screen with the graphs works fine. You can zoom into the timeline of the shown data (but a zoom into the value axis is not supported, it seems).

The web app (under http://dashboard.schurtersmart.com) provides the same functionality. With the difference that it thinks the DT31 has now three outputs (but they also don't do anything). While the graph here is a big larger, one cannot zoom into it. The graph can be switched between voltage, current, power and energy, and between showing the last hour, 12 hours or 3 days (but not 24 hours). You also get to see the current values for power, voltage and consumed energy (accumulated):

In contrast, the phone app shows the last day and allows to zoom in (and then to scroll in the data). It also has graphs for power, voltage and current (but omits the consumed energy). The current values for current, voltage and power are not shown beneath the graph, but on the page where the device itself is controlled, and only one value is shown at a time:

Speaking of controlling: both the mobile app and the web dashboard allow to turn the DT31 on and off. It's not instant though, I think the DT31 polls the target state with each transferred measurement, so it can take several seconds.

Fun fact: the web dashboard shows the DT31 as 'DS11 without Device Interface' in the device information page. I have no idea what that means (apart from both devices being very similar).

Usage and Measurements

I had two ideas for how to use the DT31:

• measure how much energy my 3D printer uses when its working on a print
• repeat the measurements I did for a previous road test, where I was measuring the power consumption of my GT90 soldering station during heating up

Measuring a 3D printer

Connecting the DT31 to my printer (a Prusa MK4) was simple, because it already comes with the proper socket. When the printer is turned on, the dashboard reports the current power consumption after about 10 seconds, altough the graph itself takes a bit longer to show this. It seems that the graph only reports in 1-minute steps, whereas the DT31 itself reports data faster (more on that below). I can also control the printer with the DT31 and turn it on and off remotely.

Unfortunately for me this uncovered an issue with the firmware I was using (the MK4 has power loss detection, and under certain circumstances after it gets power back it wants to start a new print even though none was running before. This was fixed in the newest firmware, though).

I can also see how the power consumption changes throughout a print:

When looking at the energy consumption shown in the dashboard, I was confused: the value there changes faster than I would expect. Throughout a print, the MK4 needs about somewhere between 120W and 200W (depending on the filament, which determines the target temperatures for heated bed and hotend). So for a one-hour print, this should be about 120 to 200Wh of energy - but the dashboard reported three times this value. I double-checked this be letting the printer in standby (with about 35W power consumption), and let it sit there for multiple hours, and the result was the same - the accumulated energy consumption was about three times of what it should be. To be really sure, I also connected the whole setup to a Tasmota-based Smart Plug. It reports power consumption itself a bit higher (at around 40W, and I trust the DT31 to be more exact here - the data sheets states an error of just 0.2% over the full range), but the accumulated energy consumption was where I would expect it:

(For both devices I started with accumulated energy being reset, and I took the screenshots approx. 3 seconds apart)

I'm not sure whats happening here (maybe each of the three supposed outputs that the cloud thinks my DT31 has get accounted for?), but it needs fixing.

Measuring the GT90 soldering station

So the idea was to get a graph of the power consumption when the station heats up, to verify its claims of how much power it does put into its tip. Back then when I did the road test my power meter reported at most 52W, which I still think is a bit low.

To do that, I would need a fairly fast reading since the heat-up time is just 20 seconds. The data sheet of the DT31 claims a 'update rate up to 60 sec'. Which is a bit confusing in itself, because a 'rate' would be 'x per second', and just giving a time here would be the time between measurements. So it was not clear whether these 60 seconds where the shortest or the longest time between measurements. As it turned out, it seems to be both. The DT31 seems to send data at a faster rate to the cloud (as one can see with the live power draw in the dashboard), but there it then gets aggregated into 1 minute granularity. So there was no chance to get proper readings for what I needed, therefore I abandoned this project.

Some API experiments

Schurter not only provides a dashboard for its cloud, but also API access. Although officially this is not part of the free plan I registered for, but maybe Schurter activated this for the road testers. Speaking of plans: the pricing seems quite reasonable. The most expensive plan is 4CHF per month (I assume this is also per device, even though its not stated explicitly). For that you get 1 year of data retention, API access and the full feature set (I have not been able to find out what that is and how it differs from the 'basic' feature set). For a company, that would be about 50$ (or EUR) per device per year.

The API is rather easy to use:

• call an authentication endpoint with username and password, and get a token back
• pass this token into all further calls

There are API calls for getting your devices, and to get current status and data history. So you can build your own dashboards if you would like. I did createa simple test script utilizing httpie for making the calls, and jq for parsing the data:

TOKEN=$(https --print=b -a $USER:$PASSWORD POST https://api.schurtersmart.com/v1/authenticate | jq -r '. | .token')
echo token=$TOKEN
https -A bearer -a $TOKEN GET https://api.schurtersmart.com/v1/$DEVICEID/measurements startDate:2023-09-26T20:09:00.00

The first call is the authorization, and extracts the token from the JSON response. The second call then retrieves some measurements starting at a specific time.

These measurements then look like this:

[
    {
        "current": 0.2565,
        "dateCreatedUtc": "2023-09-26T20:09:27Z",
        "deviceId": DEVICEID,
        "energyAccumulated": 5384.40625,
        "frequency": 49.98,
        "power": 32.5,
        "powerFactor": -0.56165499137342,
        "voltage": 236.96
    },
    {
        "current": 0.25725,
        "dateCreatedUtc": "2023-09-26T20:09:48Z",
        "deviceId": DEVICEID,
        "energyAccumulated": 5385.125,
        "frequency": 49.99,
        "power": 32.75,
        "powerFactor": -0.561689897373319,
        "voltage": 236.81
    },
    {
        "current": 0.2575,
        "dateCreatedUtc": "2023-09-26T20:10:19Z",
        "deviceId": DEVICEID,
        "energyAccumulated": 5386.21875,
        "frequency": 49.99,
        "power": 32.75,
        "powerFactor": -0.561724803373218,
        "voltage": 237.17
    },
    {
        "current": 0.25775,
        "dateCreatedUtc": "2023-09-26T20:10:49Z",
        "deviceId": DEVICEID,
        "energyAccumulated": 5387.28125,
        "frequency": 50.0,
        "power": 32.75,
        "powerFactor": -0.56165499137342,
        "voltage": 237.28
    },
    {
        "current": 0.2585,
        "dateCreatedUtc": "2023-09-26T20:11:00Z",
        "deviceId": DEVICEID,
        "energyAccumulated": 5387.65625,
        "frequency": 50.0,
        "power": 33.0,
        "powerFactor": -0.56163753837347,
        "voltage": 237.38
    },
    {
        "current": 0.2575,
        "dateCreatedUtc": "2023-09-26T20:11:10Z",
        "deviceId": DEVICEID,
        "energyAccumulated": 5388.03125,
        "frequency": 50.0,
        "power": 32.75,
        "powerFactor": -0.56165499137342,
        "voltage": 237.42
    },
    {
        "current": 0.2585,
        "dateCreatedUtc": "2023-09-26T20:11:21Z",
        "deviceId": DEVICEID,
        "energyAccumulated": 5388.375,
        "frequency": 49.99,
        "power": 33.0,
        "powerFactor": -0.56165499137342,
        "voltage": 237.29
    }
]

I have yet to find out how the energy consumption gets reset. Most likely explanation for my observations is that anything before your data retention period (30 days for the free plan) gets discarded.

In the above example one can see that the DT31 seems to report every 10 seconds when the power consumption changes, and falls back to a larger time span when it doesn't.

There seems to be no way to push the data from the cloud to somewhere else (e.g. into MQTT, or your own time-series database) for analysis. In the marketing material Schurter states that the cloud allows customers to customize the dashboards to their needs, but I have not found more information about it. Probably they will tell you when you ask about buying many of these devices.

Final thoughts

So is the DT31 the device I thought it would be, and would I buy one? Even after using it for a while, its a tough call. Its a nicely built device, its very accurate, and its easy to use. But on the other hand, its not a measurement device (like a DMM would be). To be fair: its not intended as such. The announcement talk abutnthe DT31 and its DS11 sibling as devices to record usage of devices, check their power-on hours and activity. For that I think its a near-perfect offer (both the device and the cloud dashboard). Customer needing that can easily pull the data they need from the Schurter cloud, and you don't need real-time measurements. For a hobbyist or smaller companies it might be better to use some smart plugs intended for smart homes (e.g. the already mentioned Tasmota plugs or some Shelly devices), and to spend some hours setting up a HomeAssistant instance instead. This would be more flexible and cheaper in the long run.

For my needs, what I would like to see:

• be able to skip the Schurter cloud and use it locally (e.g. push data into a MQTT server, or at least read data from a local server)
• get faster upate rates
• for the existing cloud dashboard I would wish it would be more like a Grafana instance, which provides better navigation in the data