Thermal Camera: The Dream

Source? *It came to me in a dream.

I don’t know if I first saw this on TV or in a dream, but I have had this idea to make glasses which can see CO2. In trying to find out if it was in fact a dream, I did a quick internet search to try and establish whether or not something like this had ever been conjured. I came across a group of people who did in fact not make a pair of glasses but rather a camera which has the ability to visualize CO2.

After further investigation and whilst simultaneously assessing my abilities and the scope of what this project would require I started to understand the 3 main ingredients necessary for a CO2 capable camera.

These types of cameras are called Optical Gas Imaging (OGI) devices, and they use a highly sensitive sensor made from a rare compound of InSb or Indium Antimonide. This material is well suited for detecting infrared radiation but are emitted by objects as heat. One thing which is necessary to understand is that these cameras are thermal cameras first and have the ability to also be detect gasses.

A filter is added so that it is able to capture infrared radiation between 4200 to 4400 nanometres (this is the spectral range at which CO2 becomes visible due to it absorbing the radiation.

Furthermore, the cameras are equipped with coolers (called sterling coolers) which cool the filters so that they can function and detect properly. This process takes around 5 min for the detection to even begin as the devices are cooled to -198 degrees Celsius or -324 degrees Fahrenheit. These devices visualise the gas emissions in a similar way to which thermal heat imaging is portrayed and the gaseous emissions of, chimneys, cars and airplanes are visible.

Reality

Getting to know and understanding the capabilities of this camera better I realized that if I tried to make or develop the design of a similar camera, not to mention if I try and make it a optical wearable, it would no doubt take longer than 1 semester at MIT and cost me more than the tens of dollars which we have allowed to us for the HTMAA course.

Taking into account the scope of the project, my limited understanding of electronics and the idea of spiral development I decided to simplify the design to something more achievable and aimed to understand/design a thermal camera and maybe turn it into a wearable if things go well.

Thermal Camera: Electronics Design

In my pursuit of designing a thermal camera. I have come across the schematics of someone who designed the camera for very cheap. Under $50 they say. Some might call me mad when I say that I am trying to understand this device through drawings, but as an architect this is how I do things.

This week we started off by trying to understand the tools we would be using to debug our electronics, which will almost inevitably become a thing which we would have to do. The reason for debugging is to try and find what is at fault with your work – weather it be a mechanical issue like incorrect soldering, PCB manufacturing errors or other things which I have yet to fail at. Issues which are electronic in nature can relate to code or closely relates to mechanics.

In the group assignment in week 5 we watched Quentin use the Multimeter (an item which can read voltage, ampere or resistance), the Oscilloscope (a device) which can read voltage difference over time, thus giving the debugger a deeper understanding of the fluctuations within the voltage of a system) and another implement which allows us to view digital fluctuations (the oscilloscope being analogue).

Reflection

I learnt that voltage can be negative or positive depending on the directionality of the black and red measurement stick. Also, that a change in voltage can also relate to the commands or codes which we tell an item to do, and in understanding this, we can see if an item is functioning correctly.

For example, if I code a servo motor to move at 4000ms intervals, this will show up on the digital or oscilloscope as spikes every 4000ms. The oscilloscope also measures human input and can fluctuate depending on a human’s interaction with a device (for example a spike when a human presses a button and let’s go of the button.

Electronic design Workflow

If I was going to try and understand how the PCB board was made, I had to take a step back. I had already drawn the components on paper and connected them on photoshop, but this told me nothing about the actual sizing of the components.

The next step was going to be to see how compact I can make the wiring in KiCAD. Although this week’s exercise might not result in a PCB for the eventual design, it might lead me further down the path and closer to the goal. I am essentially working towards the middle with the final product as an example and my own understanding as a starting point.

To see how small I can make the PCB board I decided to move around the components to see how they relate, and which configuration might be the most optimal.

Much like in architecture, by redrawing things you start to understand individual components and the greater scheme much better. I understand that I do not yet have all the answers to accurately understand how this system functions, at this stage I need to be able to ask accurate and the correct questions.

Firstly, I was trying to understand my role in the design, to try and separate component from design and PCB. By studying the existing PCB boards, I learnt that GND points are not necessarily connected to one another, however they need the same voltage to act as a baseline for the rest of the system. I also learnt how KiCAD functions, however I have yet to master it. I associated the Schematic symbols to their footprints manually because the components were not necessarily available at MIT or otherwise locally.

My initial attempts at recollecting my very limited memory of electronic design. I did this to form a lexicon for the purpose of reading the existing electronic schematics. This was going to be my Rosetta stone.

What I discovered when I redrew the images the first time is that each wire number corresponds to the correct port in of a different component. This gave me a vague understanding of the relationship between some of the components.

With this image I was trying to understand the functionality of the system. I started off by arranging the components according to how the user might use the device, and to keep the wiring as simple and efficient as possible. In essence I was simultaneously trying to become more confident in my understanding of the relationship and importance of each piece of electronics.

This previous step helped me when I repeated the connections in KiCAD (so to eventually try and optimize the design). Understanding the interface of the program was relative to manage as this is my first time using this program. I am also trying to understand the scale of the electronics so as to know if adjacent wiring will interfere with each other or have enough space to properly function.

At this stage I gave up (for now) trying to produce a fully reproducible PCB board for the thermal camera. Instead, I tried focusing on what the lines mean (yes, I am at that stage). How the components relate to the wiring, and to what extent I can use KiCAD to try and simplify my workflow.

The actual sizes of the components placed up to the wiring diagram, this shows the space for optimization in sizing of the components and gives you a better idea of what the eventual scheme will be like when the PCB is produced.

I had consulted Anthony and some of the other TAs about my project throughout the semester. Anthony told me that I could essentially simplify the existing design by at least 50%. I removed the backlight control, the on and off switch, the battery voltage sensor, the battery charging, the 3.3v regulator, the USB-UART/ programming circuit, auto reset circuit, ecternal I2C and some other smaller redundant switches and components.

This significantly improved the chances of me getting a working prototype at the end of the class. I could tell that my limited knowledge of electronic design was increasing because my questions to Anthony were less likely to be “what do these lines mean?”, as the semester progressed.

I also set out my limitations and the scope of what I hoped the end project would be in the bigger scheme of things.

Schematic and PCB design

The final PCB and Schematic has several considerations.

Firstly the correct size and components needed to be understood so that the initial design could evolve in the correct direction from the example.

I was also considering whether I am going for a handheld or a wearable device. My initial ideation was relating more towards a pair of glasses, but I ended up deciding on just a normal hand held, and tried to keep the form quite small.

Then the PCB scheme had to be as simple as possible to be able to accommodate the most components, in the most effective way. I also tried to keep jumps from the one side of the PCB to the other to a minimum.

I would probably say that the adage ‘Measure twice, cut once’ was most applicable to me. Especially when you are setting up your project. Especially because of limited resources. For me personally it also helps to understand the component and maybe try to have a physical one in my position before I try and start up. Perhaps I am paranoid but I would like to eliminate the most amount of errors before I start.

I think keeping track of which way the reverse board is going is also quite difficult. I would equate it with trying to figure out the inverses of inverses during the molding and casting week.

Another thing which I noted during the PCB design, was that adding resistors to jump over cables disrupts the netlists in KiCad. I had to resave everything and move the PCB, Schematic and General file into a new folder to have the PCB and Schematic be ale to communicate and update accordingly.

Mistake 1

After I finally fixed the writing and matched everything up. I realized that because the back of the PCB would be flipped. I made a mistake in inversing the XIAO.

Me and Anthony were speaking past each other, but once we got on the same page about what is ‘Up’ and what is ‘down’ I started to understand things better. I decided that the minimal effort would be the latter and to mirror the Thermal camera and the TFT screen.

I also ordered the parts and made sure to take into consideration shipment time. I had to choose between the 110-degree camera and the 55-degree lens. I chose the 55 degree lens because I was more interested in focused and accurate imaging as opposed to landscape shots in thermal.