|Product Performed to Expectations:||8|
|Specifications were sufficient to design with:||9|
|Demo Software was of good quality:||4|
|Product was easy to use:||6|
|Support materials were available:||6|
|The price to performance ratio was good:||10|
|TotalScore:||43 / 60|
Avnet's IDT ZMOD4410 Indoor Air Quality Raspberry Pi HAT
As climate change and airborne illness dominated the news in 2020, many of us became increasingly concerned about the quality of air we breathe. The average human nose makes a fine air sensor, however, smell can be subjective, dynamic, and many airborne compounds can escape detection.
Of particular interest are Volatile Organic Compounds (VOC). My layman’s understanding of VOCs is that they are organic chemicals emitted as gasses often from solids or liquids such as cleaning products, paint, fuel, alcohol or even clothing, and furniture. Long term exposure to high levels of certain VOCs can impact health. Indoor Air Quality sensors can provide data to support how safe or comfortable the air is within our living or working spaces. There is a large range of air quality sensors available, including scientific, industrial, and consumer sensors.
Bringing air quality assessment capabilities to the widely popular Raspberry Pi computer, Avnet's IDT/Renesas ZMOD4410 Indoor Air Quality HAT for the Raspberry Pi positions itself as an evaluation, development and prototype board, capable of providing control signals to external equipment such as ventilation fans. Not only can the pre-calibrated Renasas ZMOD4410 sensor provide Indoor Air Quality (IAQ) measurements, it can also detect odors like sulfur which may be useful for a fan control application near a toilet.
Avnet's Renesas Indoor Air Quality HAT includes a ZMOD4410 sensor which monitors Total Volatile Organic Compounds (TVOC), measures Indoor Air Quality, and estimates carbon dioxide (eCO2) levels using a Metal Oxide sensor (MOX) - useful to determine if IAQ is good or bad. Additionally, Avnet's Renesas HAT incorporates a HS3001 Relative Humidity (RH) and temperature sensor to provide additional useful data. For those developing low power applications, a provided jumper can be soldered to the board to enable a user-adjustable power supply.
Renesas provides an Indoor Air Quality table which makes it easy to interpret the numeric IAQ results.
My approach to review and evaluate the Avnet Renesas ZMOD4410 HAT made use of the available demonstration software and covered several locations in my family's home and offices.
Unboxing, setup, and additional information to accompany this review can be found on my blog, Getting started with Avnet's Renesas ZMOD4410 Indoor Air Quality HAT for the Raspberry Pi
Impressions and findings - Hardware
Installing Avnet's Renesas ZMOD4410 Indoor Air Quality HAT onto the Raspberry Pi 3 was straight forward. The included male-female 40 pin header makes the connection easy. Be careful not to stab your thumbs when pushing the pins through the HAT's header, like I did. It hurts! My blog covers the steps needed to get the system up and running.
The HAT performs very well and quickly adjusts to changes in the environment. I was impressed at how quickly a breath, odor or nearby use of cleaning chemicals would trigger a change in IAQ readings – often by the next sample (3 seconds) with a peak reading in 30-60 seconds.
The MOX sensor measures TVOCs which are important measurements to determine air quality. Another important contributor to air quality is fine Particulate Matter (PM) such as smoke particles, which do not seem to be measured directly by the sensor. However, PM should be more applicable for an outdoor sensor. Still, many particles emit VOCs which should then be detectable by the sensor. Carbon Dioxide measurements are estimated (eCO2) and not measured directly. Additional sensors may be required to get a complete air quality picture, however the MOX measurement of TVOCs should address most indoor applications.
Fitting closely to the Pi, the HAT traps heat from the single board computer. This heat influences the board's HS3001 temperature sensor readings which consistently record higher temperatures than ambient air. When using the Pi and HAT installed in the case in a room at 22 degrees Celsius, the HAT's HS3001 MEMS temperature sensor was registering 41 degrees. Without the case, the temperature sensor would still reach 33 degrees. Installing the HAT further away from the Pi would help alleviate this but would create a more cumbersome configuration.
The Pi will draw additional power when the HAT is powered, however, the total draw remains comfortably within the Pi's specifications. When idle, the Pi with HAT installed draws around 320mA. This jumped to 400-550mA when the HAT was actively sampling. Running a basic application on the Pi increased the current draw to 650-800mA and likely higher when I was not looking. It is important to use a quality power cable and power supply to avoid low voltage warnings and spontaneous system reboots as I experienced using a 1A Apple power cube. The recommended power supply for a Raspberry Pi 3 B is 2.5A at 5.1V. A 2A supply worked just fine for my application.
474.9mA, 5.077V during a sampling cycle
One of the results of the power requirement is that the setup becomes largely stationary when in use. I had occasions where I wanted to move the setup without shutting down the application, however my portable USB battery did not provide enough power. After several failed attempts I resorted to this “mini pocket size” setup:
Impressions and findings – Software and documentation
My evaluation included demonstration software, version 7.0.1 for use with Avnet's Renesas Indoor Air Quality HAT. This software is great to demonstrate some capabilities of the HAT, but is not suitable for consumer or general use, despite its reasonably polished appearance. This may be ok as the software and HAT are positioned towards evaluation and demonstration purposes.
The HAT included a getting started guide which points users to the location of the software. I was not off to a good start as the guide did not cover the software use and installation. Searching on the Avnet and Renesas sites turned up a ton of useful information on air quality, the sensor's performance and technical data. Unfortunately, information to get the HAT up and running on the Pi was not found there.
The installation was mostly intuitive but there was one issue: there are additional steps (explained in my blog) required to get the software running. These steps are not described on the product support sites. I took the extra step of reaching out to Renesas technical support. The response was extremely helpful and listed all the steps necessary however, they explained that further support was for corporate customers only. Set up instructions should be added to the getting started guides.
The software logs 35 variables for each sample in a CSV file and shows a few of key interest within the application - five of which are presented using nicely displayed and informative graphs. With one sample every three seconds, the log files get large quickly – very large. I could not find information on the support websites to explain what each of those 35 variables mean however, I suspect there may be some references available for corporate developers. Unfortunately, while seeking information about these variables, Renesas support became no longer available for this non-corporate customer. Is a reading of 86531.54710458081 for MOX 400C good? Or bad? What is "MOX DEC 350C"? More information would be helpful to isolate the cause of a poor IAQ reading.
I experienced several crashes of the software while displaying a duration of results longer than the last hour. When displaying a week, or even a day's worth of samples, the Pi would instantly go to 100% utilization on one of it's four cores. At that point, the software becomes barely responsive or crashes completely. The software leans heavily on Python and the performance could be improved by using a different solution to generate graphs on the Pi. Perhaps the software can leverage compiled code with capability for load balancing across the processor cores or by summarizing the data points.
For this evaluation, crashes were frequent and resulted in several corrupted data files and many re-tests. I found software problems to be very frustrating! I lost some interesting data as a result of crashes. The software needs work.
My evaluation involved our whole family. Everyone was excited and interested in participating including documenting their bathroom visits. Well, maybe they were less excited about recording their visits and slightly concerned I may publish their individual results! They did enjoy watching me run around the house, setting up the sensor in new places.
While many Canadians are working at home, my department works mainly from the office because we are a designated critical service of the government. Questions surrounding air quality are amplified by the current pandemic. While I have long suspected the air quality in my office was poor, the evidence suggests that the air quality is not too bad – that is until I show up in the office, then it goes down from “clean” air to “good” air.
The peak and continued IAQ levels between 2-3 occur when I arrive and stay at my desk. Am I the cause of reduced air quality? I do shower everyday! Thankfully, I was breathing which is the likely culprit. IAQ remained within good levels.
2) Home – Common Areas
With pets, kids and many activities, quality of air is a key focus area in our home.
a) Hallway - My wife was planning to do some painting in a hallway, which provided a great opportunity to provide a baseline reading along with data for painting. The paint used was Behr Premium Plus, which advertises as a low odor paint. Thanks to 14rhb for the suggestion! Unfortunately, there was too much data for the software to handle and the system crashed before I could determine the duration of the paint's impact on IAQ. I did not dare ask my wife to repaint the hallway so I could collect new data! The available data remained within good parameters.
Painting the hallway
b) Kitten. Measurements near the kitty litter box displayed some predicable results. I can readily see where the IAQ reading jumped and box was likely in use. The IAQ remained between very good and good.
Near Hawthorne's litter box
Litter box Estimated Carbon Dioxide (eCo2)
c) Furnace Room. We live in Alberta, Canada. It can get cold here in the winter, sometimes very cold, and a high-quality furnace is a necessity. During a period of -22 degrees Celsius, I did get some interesting results in the furnace room where I can readily see the furnace turning on and off. While the IAQ remains good, it does get close to level 3 'medium' when the furnace runs. Still not bad, but it may be time to get the furnace checked!
3) Home – Daughters Bedroom (aka: The Reptile Horde)
My young adult daughter's room is in the basement, where she keeps her pets and bugs. She presently has 15 creatures (geckos and snakes) in tanks. Along with this collection are literally hundreds of bugs that she breeds for sale and as food for the reptiles. One of the geckos has recently become a new mom and has two new arrivals!
My daughter's room did not smell great, so I had asked her to purchase an air purifier which helped a lot. However, I was concerned that the IAQ may be an issue with so many creatures in a small space. The results of extensive testing in this space showed that the air quality is not as bad as I feared. There are periods where the IAQ readings reach close to 3-medium but that was when the room was occupied. Testing for minimum 24 hours with and then without the air purifier showed that the purifier did not seem to have much impact on IAQ in this room, however the purifier did dispel odor rapidly.
The Reptile Horde IAQ
The demonstration software also includes a separate odor detection mode, which generates a control signal if a sudden change in IAQ is detected and the threshold exceeded. Running the software in this mode shows brief periods where the odor gets out of hand, the red vertical lines display when the control signal is active. The in-room air purifier fixes this quickly, however, there may be a future purpose for the HAT to control an external ventilation fan in this room.
4) Home - Garage
We have an attached garage which also has a furnace, though this furnace is much older than the one inside our home. The worst readings were found here, where the air quality hovered above IAQ 3-medium most of the time. The peaks show when the furnace was on and the two dips in the graph shows when a burst of outside air came in through an open door. I didn't expect garage air to be high quality however, it is clear that this furnace will need to be checked as well.
5) Home – Washrooms
The Avnet Renesas ZMOD4410 Indoor Air Quality HAT was placed in a few washrooms in our home.
I probably only have myself to blame for the highest peak which exceeded level 3-medium. Still, not as bad as I expected, which would have been in the red.
The HAT responds well to chemical changes. My wife had cleaned the powder room surfaces and the sensor indicated a rapid change. The readings stay close to 3-medium for around 20 minutes.
As a final test, I decided to breathe on the sensor after sampling a couple wee drams of Scotch whisky! The Dalmore 18 is one of my favorites with spicy chocolate notes and smooth finish. The Tomatin Cadenhead Small Batch is also very nice but at 110 proof it is closer to the look and taste of a fine gasoline as reflected in the results. The Tomatin presented the highest TVOC reading I saw in any test. Should an IAQ profile accompany whisky tastings? If so, I have a lot of sampling to do – bourbon is next!
Scotch tasting - a more objective way
Scotch under test
In summary, Avnet's Renesas Indoor Air Quality HAT is well constructed, responsive and provides valuable air quality data. The MOX sensors provide data on a wide variety of VOCs. The humidity and temperature module is a nice bonus, however, it does not provide accurate temperature readings when installed directly on the Pi.
The demonstration software looks polished, provides informative graphs, and works sufficiently to demonstrate that the HAT functions. The software falls short as a useful tool, crashes often, and has significant performance issues.
The Avnet IAQ HAT is a great value considering the capabilities it delivers. At under $50 USD, it is very affordable.
Avnet's Renesas Indoor Air Quality HAT - gecko approved!