Hardware for the affordable lab
Overview:
Until we found a lab we could work in in Oakland, we were limited to our highschool lab, which gave us a good starting point with a large amount of glassware, but very little specialized equipment. We quickly realized this limit when our team leads ran a class to provide everyone with lab experience to get ready for the actual experiment. To try and overcome this, our team began to design cost effective lab hardware and made it available for aspiring lab scientists to use. This hardware also makes designing a basic home laboratory a far less expensive endeavor.
The two main pieces of equipment we chose to design were a thermocycler (for PCR) and a microcentrifuge.
Design goals:
For both of these we began by looking at already existing models both commercially and ones in the 3d printing community. Here is what our research came up with, and why we decided to design our own.
The cheapest available thermocycler was 250 dollars, and could run five tubes at once. My main qualms with this design was the lack of any cooling element. The design only had resistive heating, and no way to cool tubes. We aimed to design one below that price that did have this feature.
Centrifuges go from very small to very big, so we focused our attention on 10,000 rpm centrifuges, strong enough to create cell pellets, extract DNA and perform other common functions. The cheapest we found was around $80, limited to 10,000 rpm with no variable speed, and a roughly 5 cm radius rotor. Once again, we aimed to design one with both a better feature set (variable speed, larger rotor), and a lower price tag.
Design process:
Thermo cycler:
For our thermocycler design our team decided to use Peltier modules, after one of the team leads had bought them for another project. We later found that the old version of the pocketPCR (what we deemed the cheapest available thermocycler) actually did use a Peltier module. For control, we used an esc 32, because they are cheap, have an in-built ADC and our team lead once again had them lying around.
Our first challenge was designing something to cool the hot side of the thermocycler. We used a cheap metal heat sink initially, but quickly realized we needed a fan to blow air over the heatsink if we wanted any chance of our Peltier module not overheating. During this process we also had to purchase new Peltiers treated for higher temperatures as denaturing dna had to be done at 98 degrees celsius and our old Peltiers were limited to 85 degrees celsius.
So we ended up designing a 3d printed solution to hold the fan below the heatsink. Another big challenge was somehow cutting a heat block to hold these tubes, which you will read more about in the build section.
You can see the end design here on github, with pcb designs available.
Microcentrifuge:
As one of our team leads already had spare 1000 kv drone motors and escs, we wanted to utilize those in our design. For convenience we controlled our motors using a raspberry pi 4 our team lead also had, but in the final design we only needed a 5 dollar microcontroller such as the raspberry pi pico.
One disadvantage of using this already available motor was the limited rpm we could reach. Our
motor running at 3S (11.1v) could spin at a maximum of 11,000 rpm. The reason this mattered is that the
amount of gs a sample is exposed to in a centrifuge is 
. ie. double the rpm meant 4 times the g force. To compensate for our limited rpm, we decided to
design a fairly large 7 cm radius rotor, to aim for a g force of ~8000. This, while less than what
some protocols call for, was enough for most of a small lab's needs.
You can find the design files here on thingiverse with additional instructions.
Build:
Thermocycler:
Step one was manufacturing the heat block. We ended up choosing aluminum as our material, and initially tried our school’s cnc to cut it. This turned out to be a bad idea, as without water cooling our school’s cnc was not able to handle inch thick aluminum, and we ended up shattering a titanium carbide bit (yay). Eventually we found a better solution was to use the drill press our school had, which turned out to work far better.
To measure temperature we had to calibrate a thermistor for meant 3d printing using 3 lab thermometers, which gave us a curve to convert the resistance (measured by a voltage divider) into a temperature value.
Microcentrifuge:
This was fairly straightforward, and thankfully, the model our team lead designed worked correctly the first time, due its similarity to a normal drone rotor. For our initial runs, we slowly built up speed, and eventually we had a functioning microcentrifuge.
Testing and Application:
Thermocycler:
Our first test for the thermocycler was for the synthetic biology class we ran for our team to gain lab experience. Using the mini pcr bio ptc taster reagents, we extracted cheek cells and used pcr to determine the genotype of our team’s ptc tasting gene. The thermocycler process was nerve racking as the first time we tried this our Peltier module fried itself, so we had to last minute order more, and make things run.
We found that to reduce the load on the Peltier module, insulating the sides of the heat block with cardboard worked wonders, and significantly increased our ramp speed.
Additionally, since our PCR did not have a heated lid, we had to use something to prevent the samples from evaporating, which turned out to be easier said than done. Normally, you would use molecular biology grade mineral oil, but since that was in short supply (i.e. we were running out of time and couldn’t buy it) we ended up using vegetable oil. In the future we hope to add a heated lid to this design.
Nonetheless when we ran our gel in class we did get pcr bars supporting that our thermocycler worked!
Microcentrifuge:
To test the speed of the microcentrifuge, we marked one of the tubes with a black dot, and took 240 fps slow motion video from a phone. With that, we calculated that the estimated rpm of the rotor was around 8000 rpm at top speed. While a little bit slower than we wanted, this centrifuge was still quite powerful and separated finely ground spice powders from water effortlessly. In the future, we hope to try out motors rated for even higher rpm, say 3000 or 4000 kv with smaller rotors to hit incredibly high g forces.
Safety:
Thermocycler:
The thermocycler thankfully runs relatively safely. The main concern are burns from touching the heat block while the machine is one, which we tried to minimize the danger of by adding a cardboard lid.
Microcentrifuge:
The microcentrifuge was significantly more risky than the thermocycler due to the fast moving rotor. Our first step to ensure our safety was to design a rotor that we were certain would not shatter holding eight tubes spinning at 10k rpm. We did this using the known tensile strength of pla as well as the estimated weight of the tubes and rotor at max possible g force. In our final calculations each of the 2 gram tubes weigh a maximum of 35 pounds at 8000 g force. With the tensile strength of pla being about 5000 psi and our smallest cross sectional area on the tube holder 1/2 sq in * 0.2 (20 percent infill). We get an estimated strength of 500 pounds of force, more than good enough for our expected 35 pounds from the tubes. Of Course this does not take into account the force felt on the spinning rotor from its own weight, but that is fairly well distributed across a large cross sectional area.
For additional safety, we plan to design a plexiglass shield to place over the centrifuge while it is running.
Comparison:
Thermocycler:
|
|
Our design |
|
|
Cost |
$45 (electronics) + $5 (3d printing) + $60 (cnc/sls) = $110 |
$250 |
|
Other features |
No heated lid, no gradient block |
No heated lid, no gradient block |
Microcentrifuge:
|
|
Our design |
|
|
Cost |
$25 (electronics) +$10 (3d printing) = $35 |
$80 |
|
RCF |
~ 5000g |
~ 5400g |
|
Other features |
Variable speed |
No variable speed |