Last Updated on April 4, 2023 by pg@petergamma.org
- We are very happy about this instruction how to build a 16 bit A/D converter, which can be attached for instance to a ESP 32 modules:
- This instruction is a backbone to built up new devices with ECG, EEG and MEG and InfluxDB
- InfluxDB is currently only rarely used for low-cost high quality devices, because it is challenging for beginners
- But if you read our journal there are also easy examples to start working with InfluxDB
- But building new devices from single components and InfluxDB is a lot of work. These components have to be soldered together. To make a view tests an a breadboard is easy, but if we want to build a device, this takes time.
- It seems also the components to build such devices are are not very highly integrated. Are there for instance no ESP32 modules available with integrated 16 bit A/D converters? This would save a lot of time
- the prices for single EEG modules remained constant over the last several years
- the prices for single ECG modules have dropped over the last several years from around 20 – 30 USD dollars to only a view dollars
- Do we therefore have to expect an increase in prices of single EEG modules if more pieces are sold? Yes, but there are more sellers who sell single EEG modules than for instance OpenBCI modules. Therefore we have a market for single EEG modules, and we think it is difficult to increase prices for single EEG modules.
- Arduinos only offer 10 bit A/D converters built in
- The ESP 8266 has 12 bit A/D converters built in
- the recommendation for research grade applications is 16 bit A/D
- At least 16 bit A/D conversion is recommended for research applications, but devices as for instance Holter ECG device, Adinstruments Equivital have a lower sampling rate than 16 bit, and are not of the highest quality.
- Are these devices built from the most affordable components on the market, and not of the highest quality?
- We could do the same and be satisfied with ESP 8266 and 12 bit A/D to start with.
- We think it worth it to start do some experiments with these components, Home Assistant and InfluxDB.
- There we can make some tests, for instance if we can combine these sensor data with sensor data we already have in Home Assistant, and try to do data analysis in Python.
- But to build a new device with InfluxDB is challenging. Aren t the newest developments on the marked built from these components? But aren t these also the most expensive ones on the market, since they are built from highly qualified persons?