The Centrifuge Camera — Will it separate?
This channel provides a live view inside a lab centrifuge, showing you the process of separating different types of (viewer suggested) samples. The separating is achieved by spinning the samples at 2500 times the gravity of earth, thereby the samples are separated into different layers each with their own densities. The goal and idea of this project is to find collaborations within the different realms of science to generate interesting content that can for example be used in educational purposes. Making the video’s by themselves, story starters, that will hopefully be appealing to a brought audience besides science.
How did I come up with this idea?
As a former lab technician, I was always fascinated by the centrifuge process, what would the process look like from within the centrifuge? Whilst working at the lab, I simply had only visual knowledge of the “pre-centrifuge” sample and the “post-centrifuge” sample. Information regarding centrifugal force and the effects of it on different types of samples can be found described by different scientific perspectives accompanied by illustrations and animations. However, a live view of the separating process I have never seen included.
Now, many years later, I wanted to see if I could actually visualize this process from within the centrifuge. So I started to focus on finding a centrifuge that would allow me as a visual artist to achieve this dream. There are many centrifuges available in the second hand market, my dream was to find a centrifuge that had relatively large buckets allowing 1KG+ per rotor position, cooled and allowed between 5.000–10.000RCF.
Unfortunately, that amount of budget was not available at this moment. Important now was to create a minimum viable product that was able to record interesting visuals from within the centrifuge. From a budget point of view, a smaller bench top centrifuge with swing out buckets.
Finding several possible candidates that were interesting from a budget point of view, it was time to recreate a 3D replica. This would allow me to pre-determine if the hardware needed would fit the centrifuge. Unfortunately, the standard buckets were not big enough to accommodate the hardware needed to record the sample. So A custom bucket design was made, unfortunately this also resulted in a few candidates that needed to be dropped. Size and weight wise, it was not possible to make the designs work. Lucky enough, two of the candidates showed promising results! After a few weeks of design iterations, I trusted myself enough to take the leap and buy one of the candidates, a Hettich Rotofix 32.
A new chapter in the project! It was time to turn the 3D designs required for this project; buckets, inserts and other small parts into reality. This was all done on my Ultimaker 3D printer with PLA filament, since this would speed up the process of making quick iterations. Once I’m happy with all the designs I will depending on the tensile and compression stresses choose a material suited for a long term solution.
But before that I needed to test fit my first designed and 3D printed parts into the centrifuge with all hardware placed inside of the inserts. Allow speed run at 500 RPM — 1000 RPM was done to see if the parts would react to low g-forces.
This process continued for many months, during that time lots of design changes have been made to solve tensile stress & compression that were affecting the prints by the high g-forces. But also many ergonomic design changes had to be made so loading of the swing buckets, inserts, camera module, sample etc. would be easy and consistent. Besides that, technical design changes had to be made to accommodate the camera and other electronics.
Now a few months later I’m proud to present the first working version of The Centrifuge Camera that will spin and record samples at 2500 g-force (RCF).
From the render above, most black parts are 3D printed out of black ABS. The rotor and round bucket are original. The “silver” bracket around the 3D printed bucket is laser cut out of 5 mm Aluminum and bench-pressed in to shape. The other aluminum parts needed for cooling are also laser cut to size with all it’s mounting holes, accommodating the electronics and camera.
In the next “Chapters” I will go more into detail regarding each of the parts.
Power bucket:
Delivering power to the camera is done by utilizing the original centrifuge bucket. This bucket was large enough to enclose two 18650 batteries, including its holders. The only thing that needed to be 3D printed was a flex-able sleeve to secure the batteries in place and the cap to close the bucket.
The two 18650 Li-ion batteries have a build in BMS and provide, 3500mAh and are connected in series to provide a minimum required voltage of 5V to the camera. The power consumption of the camera itself is 5W, so the battery should last up to 4–5h.
Camera & Sample bucket:
The camera and sample bucket is made out of two parts. The 5 mm thick aluminum bucket holder (sheet metal; laser cut and bench pressed) and the 3D printed bucket.
The bucket holder is made out of 5 mm thick aluminum to be able to withstand the centrifugal force. The 3D printed bucket that fits in to the aluminum bucket holder will make sure the 3D printed layers are compressed. If we leave out this bucket holder and 3D print the whole design, it will be exposed to tensile stresses and the layers will rip apart. Creating the complete bucket out of aluminum would not be a reasonable option.
By using this combination I was able to reduce not only weight but also production costs. Sheet metal is relatively cheap compared to completely CNC the bucket out of aluminum. The bucket holder also gave lots of future creative freedom, since this would be using the maximum amount of the “real-estate” available for future internal iterations. Iterations that can easily be 3D printed at a low cost to accommodate future needs or fixes.
The Camera and sample bucket currently holds the following important parts:
- 4k Camera + control board (stream + DVR with micro SD)
- Heat sink to dissipate the heat of the control board that gets very hot, 70C +
- Counterweights pockets
- Tube holder that is interchangeable to support different tube sizes
- Centrifuge tube with a sample.
Light setup:
To be able to see the samples in the centrifuge on camera, a led-light ring was implemented in the top lid of the centrifuge. Previously the light was placed inside the “camera” bucket, due to ergonomic reasons; not having to recharge batteries and connecting wires the top ring light was introduced.
This also gave me the opportunity to choose “smart” LED light options, this NeoPixel led ring with 16 RGB LEDs is fully addressable and can be accessed through Wi-Fi since it is connected through a Node MCU running Tasmota.
First results:
Now that I’m able to create content with The Centrifuge Camera, I’m trying to figure out a good format for a YouTube Channel. Below are already a few videos that are shot 4K from within the centrifuge. In addition, for those interested, there is a live stream on Twitch as soon as new samples are being spun in the centrifuge the stream will be live.
What would you like to see centrifuged?
Future plans:
For the future, I want to improve the following technical parts of this centrifuge:
- Smooth dark background by printing the current designs on a resin printer.
- Ergonomic improvements to even quicker access all parts within the buckets.
- Higher quality test tubes, unfortunately these are quite scratched and often break at sustained high g-force load. Besides that, they can’t handle high density samples. In the next few weeks, a shipment will arrive with tubes that should be able to handle these current issues.
Besides technical improvements, I want to improve the quality of the video’s:
- Find a good format for the video’s, include pre- and post- process of the sample, besides only inside the centrifuge?
- Music replace the current centrifuge sound.
- Voice over explanation (depending on available information on what we see during the process).
- Add change current overlay to give a better idea on how fast time is elapsing in this sped up footage.
- Create a nice leader and other branding.
I hope this will improve the overall quality of the videos and create a higher attention span.
Next to that, I’m looking for help from the science community.
- Test samples, suggestions on what to centrifuge and/or ways to do this accordingly.
- A more in depth explanation of what happens in the individual samples.
- Possible ideas / collaborations.
- Any recommendations tips and or tricks on how to get better visual results.
- Leads on a larger and higher RCF centrifuge to expand the capabilities of the project.
If you have any questions, feedback, tips, tricks and/or (collaboration) ideas please feel free to contact me.
