Testing the Nordic nPM1100 with ECG // Part 1: Hardware

Welcome to this new three-part series, which aims to take certain developer kit components — namely the excellent nPM1100 Evaluation Kit — to manage power, an nRF dev kit and some other miscellaneous components, and build a complete IoT device. In this case, we will be creating an electrocardiograph based on a past project, but we will take it as far as we can. In the final installment, we will be testing several other use cases and coming to a consensus to which kind of usage is better.

The series will be divided into:

Part 1 - Hardware for the ECG: Dev kit integration, system's architecture, and component development. We will also test it with Nordic Semiconductor's Power Profiler Kit to check on its draw.

Part 2 - Software for the ECG: Mobile app development along with some AI and cloud integration.

Part 3 - Multiple use cases and which is best: We will go ahead and test several other common, mainly IoT use cases, from small solar panels to other gadgets.

Hopefully part three of this series will coincide with the market release of the nPM1100, and we can go ahead and possibly proceed with this prototype. But for the moment, let's get to the first part.

The ECG: A little bit on the problem and why it is important to quantify health markers

Many factors of daily life can permanently affect cardiac function.

Quite a lot of people have to undergo cardiac tests frequently in expensive hospitals with gigantic measuring devices. We are in an era where open health is stronger than ever and it is time to make the patient the point of care.

The market for electrocardiography is enormous, as it has become the standard for patients with heart risks.

What we can see in the graphic above is that most of the electrocardiographs are those large machines (as a biomedical engineer, I can attest that most are quite old). In addition to this, most in the "Holter" category are not really wearables but smaller ones that can be carried — despite that a wearable version can be used at home to provide invaluable information about the patient's heart.

One thing to note is that the first wearables have already come out in the market and the results are not that great. The main issue that doctors have put forth is that there is too much information. Think of the internet before data aggregators — it has no value if it cannot be interpreted correctly and that is something which has to be taken into consideration. A solution should aggregate all that data and provide carers with useful information.

Second problem

This brings us to the second issue, which is quite basic: most EKG machines, whether they are Holters or rest EKGs, use gel-based electrodes. These are completely unusable in an athletic environment, e.g. athletes trying to measure themselves during activity. For these reasons, while designing the IoT device, we will try to develop dry electrodes at the same time.

Design and choice of parts

This is the connection diagram of the components, separate from the circuit, if you want to do it in a modular way with these pieces,

• Communication with the ECG is done through the SPI protocol, unlike other more economical modules such as the AD8232, which are handled with analog readings and occupy an ADC.
• The power that we are supplying with the nPM1100 EK is 3V, which is the maximum voltage that the EK supplies.

The ECG 3 click by MikroE was chosen as it is a complete solution for ECG and HR application development, utilizing a specialized IC with a clinical-grade analog front-end (AFE). The ECG 3 click uses the MAX30003 IC, an ultra-low power, single-channel bio-sensor, which features a wide range of different options, making it an ideal solution for development of heart rate and ECG monitoring applications, fitness applications, ECG bio-authentication, and similar applications related to heart monitoring. This board is also equipped with the 3.5mm electrodes connector, allowing it for out-of-the-box use.

So with this sensor we cover a lot of ground developing this solution.

The XIAO nRF52840 was selected due to the compatibility with the nPM1100, as they work best when combined. It also functions as our Bluetooth Low Energy gateway for our mobile application.

Basically, through the nRF gateway and cloud we get the signals from our microcontroller. We then clean it a bit and send it to our dashboard.

The EK configuration is as follows. The configuration elected is the best one we have to work with 3V. More details can be found in the official documentation.

Spotlight on the nPM1100

Of course, all is useless without proper power management. One of the biggest problems we have with these types of applications is that power management is very poor for these kinds of devices. And this is where the nPM1100 EK is actually a superior solution for this kind of applications. My team and I have developed in the past several products with other chips and power management features, but the nPM1100 seems to be a great choice because of its features.

The brief on the nPM states: "The nPM1100 is a dedicated power management IC (PMIC) with a highly efficient dual-mode configurable buck regulator and integrated battery charger. It is designed as a complementary component to Nordic’s nRF52® Series and nRF53® Series System-on-Chips (SoCs) to ensure reliable power delivery and stable operation, whilst maximizing battery life through high efficiency and low quiescent currents. It can also be used as a generic PMIC device for any other suitable application.Its extremely compact form factor makes it ideal for advanced wearables, connected medical devices, and other size constrained applications."

This makes it well-suited for the application that we are trying to develop here.

More features:

• Battery thermal protection and automatic selection of three charging modes: automatic trickle, constant current and constant voltage.
• Maximum charge current is resistor selectable from 20 mA up to 400 mA.
• The charger features a discharge current limitation and is JEITA compliant as well.
• The PMIC is USB-compatible and allows for charging batteries through USB, with automatic port detection.

I also have the EK i hand we will be testing it with the Power Profiler Kit.
Documentation is available here.

Connection diagram

To use the device, an ECG cable connected to the arms was attached to the ECG module, which has a jack input. Later through this, data is sent at a frequency of 100Hz to the app over so it can be visualized as a coherent graph.

Due to the fact this is a device that we are going to be using for long periods of time and that must be used every day, we can understand that the use of disposable electrodes is not feasible. This is why we decided to make our own dry electrodes.

Materials:

• Copper plate
• Silver conductive ink
• Electrode external snap

Electrode arrangement

In order to read the EKG and make the device as comfortable as possible, we take into consideration the arrangement of electrodes on the Apple Watch.

First we place two electrodes on the right hand and one on the left hand as follows.

With this arrangement of electrodes, we can obtain an ECG signal that while not perfect, we can fix with a little processing.

BLE service

BLE stands for Bluetooth Low Energy. As aforementioned, the XIAO nRF52840 was picked for its compatibility with the nPM1100 as they work best when combined and it functions as our Bluetooth Low Energy gateway for our mobile application.

All ECG information is sent at a frequency of 100Hz to the device connected to it.

BLE service:

• Name: ECG
• UUID: 0x2a37

BLE characteristic:

• UUID: 0x2a37

Properties:

• Read
• Notify
• Properties: read, notify

Value:

• 8-bit unsigned int.
• Value: 8-bit unsigned int.
• BLE characteristic: UUID: 0x2a37 Properties: Read Notify Value: 8-bit unsigned int.
• BLE service: Name: ECG UUID: 0x2a37 BLE Characteristic: UUID: 0x2a37 Properties: Read Notify Value: 8-bit unsigned int.

Demo

Here is a video of our first attempt to get a signal from this initial design.

You can get the simple React Native app to get that feed from here.

We will go more in depth on its development in part two of this series.

Power consumption

Within the official documentation, they indicate how to read the current consumption of the nPM with the PPKII.

See Nordic Semiconductor Official Documentation.

Now, let's showcase the power consumption of the nPM1100!

The power consumption of our entire design is less than 10mA, as can be seen in our measurements with the PPKII.

• Connection of the POC to the PPKII in Ampere Meter mode.

• Device measurements: we have peaks of 20mA due to the reading of the ECG data every 10ms.

• Here is a two-minute consumption test:

Proof of concept

Finally, we put everything inside a 3D-printed case for the moment in order to test it.

End of part 1 and next steps

As a three-part series this is just the initial design, covering component selection, a showcase of the nPM1100, and how to correctly integrate these parts. In addition, we have the sensor and device selection in order to reach our final destination, which will be a hardware prototype that we can develop an IoT cloud-based application. For that, we will use React Native, AWS, and Nordic's own proprietary cloud to consolidate everything together.

______________________

If you're new to Nordic Semiconductor's nPM1100, explore this Getting Started project tutorial that walks you through initial set-up.

For a high level overview, check out Hackster's Spotlight video on all things possible with nPM1100.