We continue with our Smart IoT Power Design video series available at our IoT resource center.
In this chapter Vanja Samuelsson, Founder CEO at Qoitech and David Martens, Application Engineer at Thales DIS, explain how to set up the environment to run power consumption measurements by combining a Thales IoT module and the Otii tool by Qoitech.
Video Transcription
Qoitech: So let's get right into this. David is a Thales expert on automotive and low-power modules.
So David, the IoT devices with connectivity via cellular network use CAT M or Narrow Band IoT (NB-IoT) technologies for energy efficiency and long battery lifetime. Both leverage so-called power saving modes which are specified by 3GPP.
Can you tell us about these modes? Are there any differences in the modes for CAT M and NB-IoT?
What do you need to take care of when you're working with this type of technology?
LTE Cat.M vs NB-IoT
Thales: CAT M and NB-IoT are not competing technologies. Both as you mentioned were introduced in 3GPP Rel. 13.
There are a few things that are really crucial to understand why both technologies have different power consumption profiles, even when they have the same eDRX (extended Discontinuous Reception) and PSM (Power Saving Mode) power-saving modes.
So basically let me give you a quick overview of this.
CAT M has, for example, even some additional features like audio and mobility.
It has a higher throughput, higher bandwidth but all of this comes together with higher current consumption of the module itself.
Both have, from a power-saving perspective, the same PSM and eDRX.
There is also one crucial aspect to understand that CAT M and NB-IoT are not available everywhere.
It is a relatively new technology even if it is already rolled out mainly in Europe and the US, for example.
The US focuses more on CAT M, and by contrast, Europe focuses more on NB-IoT. This will change over time.
We see in Europe now the first CAT M networks because it always depends on the application use case.
Both have benefits, and disadvantages and you cannot say one or the other is a better technology.
eDRX and PSM
Qoitech: Right, understood. So this really does depend on the country where you're deploying, the network operator and the use case.
So you know I'm interested in details of the power modes.
Can you explain more about eDRX and PSM and how do they really work?
We can take a little higher level, to begin with.
Thales: What we see here is the power consumption over time, and this represents a CAT M and NB-IoT power cycle because as mentioned, these power-saving modes are in both technologies.
At the beginning (grey: A, B, C, D, E) we see the power-up, startup, registering to the network.
Followed by the timer T3412 TAU (black) and within the active time timer T3324 (green).
These we are typing actively in, or basically the application microcontroller is transmitting these timers to the module itself after that module is negotiating them with the network.
You can imagine when you have 100,000 devices on the same cell, and all devices will transmit the same timer at the same time, of course, then this cannot really work out.
So the network is dynamically scheduling and answering you: "okay you can go now to sleep for this and this period".
To go a little bit deeper into this with our T3412 TAU which is like the overall which include the active time T3324 and the PSM.
The PSM is a new mode that got introduced, and with that mode, the module really reaches the lowest power that you can achieve.
In this mode, the module cannot be reached by the network so almost everything within the module is shutting down and the module falls into this PSM mode from the protocol perspective and waits until the timer T3412 TAU (black) that got negotiated with network expires.
On module level, this mode is called suspend.
Of course, there is the case where you can shut down the module and of course, restart the module and register again but with PSM.
The cool thing is that you are instantly again on the network so you do not need the starting phase, which can take a while with network scan and registration.
In terms of current consumption in PSM, we are talking here from a 2-3 uA range.
With the second timer T3324 (green) we are setting up the active time.
Usually, in LTE-Cat 1 and higher categories, we have a periodic DRX which means on a module level, the receiver is turned on and listening periodically to the network in a specific time.
Now with the extended DRX, we do not need to continuously listen and react really quickly. This process helps power consumption savings because with CAT M and NB-IoT which is rather for data only, e.g. for some sensors, we are listening in so-called paging time windows.
If you would now take the average over a paging time window and eDRX floor, of course, the current consumption is much lower.
You can almost freely configure the active time (the period how long you are listening and in which distances of the PTW).
Qoitech: It would be cool to see it in reality.
Can you guide us through a quick setup and show how these modes work from the power perspective?
How to run power consumption measurements
Thales: Of course!
Here we have our LGA DevKit the first thing what I do is putting in the frame to fit the ENS22 inside our socket.
The ENS22 is our narrowband IoT module, and so I'm placing the module and the lit of the socket pressing it a little bit down, and now the module is connected.
As a second step, I need to plug in a narrowband capable SIM card and of course an antenna to the “main” SMA connector.
Now I am connecting the USB cable to ASC0 which is powering the LGA DevKit and, with that, we also have a serial connection over an FTDI VCP.
Now I need to connect the Otii banana cable (one ground wire and one 3.75 V wire) to power only the module itself.
So this is how the setup looks like. Now I can simply press the on button, and I can see also that the LED on the Otii is flashing and I am ready to operate.
We can immediately see that the module is starting with a “sysstart” over the terminal and what we also have seen in the background based on the current is that the
module is now starting and registering.
Now we can set up with AT+CPSM for the TAU and active time.
We discuss in a separate video on how to calculate these timers.
What we now nicely see is that the module is already registered and already in the low power mode.
We see the paging time windows, and we see in the beginning a little bit higher peaks, so the module is registering here to the network and negotiating the timers.
When we turn off RTS, and we see that we drop even a little bit more with the current consumption so from the mA to the uA range.
Okay as you see we can verify that this is working and not only theory and based on that we can redo any calculation so basically an estimation of the battery lifetime based on different timings.
In the end, we can now see that module is falling into PSM where we are reaching really low current.
Qoitech: We were talking about the impact of the device and what you can change at the device side.
We also understand that there is a huge impact on the network side and different operators have different settings.
So, this kind of default current curve that you would see in Otii when you measure will look very different in different networks.
Can you tell us a little bit more about that?
Thales: As mentioned, it is a regional thing and of course the different network operators have different strategies of the deployment itself, and therefore the timer ranges are even sometimes a little bit different.
With narrowband, you have longer power period compared to CAT M, but from the network perspective, of course, it's not from one to the other day that everything everywhere is deployed and it works perfectly.
Some network operators have right now fixed timers or some features as eDRX are not working.
When and where to measure?
Qoitech: So you need to understand in what network you are deploying and what you can measure during these different measurements that you're going to be doing during the development project.
With that said, what would be your recommendation when and where to actually measure power during the project development?
We see quite often the people might not do it often enough right so how often should your customers measure and how should they start measuring and where?
Thales: My experience is based on what our company and also what my customers are doing.
They need at first the right tooling.
For sure, so the Otii is a perfect device for that. It is easy-to-use also in combination with our LGA DevKit. We prepared it for current measurement, so you only need two cables to connect it, and you can start instantly measuring.
Of course, there are RnD measurement, verification measurements for this are like the standard procedure, but what is a crucial point with that type of new technology is the measurement on the application level.
Whenever a firmware change is happening on a module or on a host controller of the customer application, you need really to be sure before you are deploying this firmware update that nothing changes on the current consumption usage.
An important aspect is the verification step, not only in the lab but also in the field.
If you use CAT M within a track and trace device, you really need to make sure that in the field you have the coverage with CAT M, but the reality might be different.
So you will have based on the measurement a result where you can decide what strategy you need to apply with your software.
There is not really a case where you can skip somehow the measurements, so you need to do it from the verification phase, through the development, through the deployment and even further for the whole product lifecycle.
Best setup for proof of concept
Qoitech: Finally, if a customer wants to quickly do some concept or proof of concept, what sort of setup would you recommend for them?
Thales: So, of course, they can easily purchase our LGA
DevKit it is prepared and ready to use but for proof of concepts. Sometimes you want to connect different things with each other that means a host application or a host processor and some sensors also. For that, we have with our partner MikroE the Click Boards.
There we have, for example, the ENS22 which is an NB-IoT module.
We will also have quite soon our EXS82 Click Boards with CAT M, NB-IoT and 2G with that you can simply plug together your preferred host application with some other sensors.
You can also measure current as we prepared it especially to connect it as easy as possible which makes verification and measurement even on that small proof of concept possible.
Qoitech: And of course having an Otii to measure all of these very important, right?
Thales: True!
Qoitech: Alright thank you, David, for these insights.
Thank you for listening.
Stay tuned to more of these videos where we are going to be digging much deeper into these different low-power modes of cellular IoT.
Thales has teamed up with Qoitech to help developers improve their power design and conserve energy throughout entire IoT applications. The Thales Cinterion LGA DevKit and Otii measurement package closely examines power use at every step and makes design recommendations to achieve the right balance of functionality, device size and battery lifetime. Learn more in the videos how to optimise your power management design and save on costs.