[sudarshan] is a solar hobbyist and needed a way to cut solar cells for his projects. He had previously created a rotary tool saw but manually feeding them through was sketchy at best. With just a slight wrong movement of his hand or flex in the work surface would cause the cell to break. These cells are extremely brittle and break easily. He needed a method of cutting these cells that was free from jitters and would cut in a straight line. He looked around his junk bin and found an odd solution… a scanner. Yes, the type you would scan photos in your computer with. The scanner had two critically important features, a flat surface and a carriage mechanism that moves perfectly parallel with that flat surface.
[sudarshan] made a solar cell cutting mini table saw with that scanner and made the cutting happen automatically. He mounted a motor with a diamond saw disk to the carriage, that is responsible for the cutting. The blade was positioned just high enough to poke through the plexiglass that replaced the original glass bed. A power switch turns on the cutting disk motor and an Arduino was used to move the carriage, including the cutting blade, back and forth. Two of the stock scanner buttons were reused and wired to the Arduino to keep the saw looking good.
The first few passes of the saw were done to cut a slot in the plexiglass. In order to cut a solar cell, the cell is taped to the bed with the desired cut location aligned with the slot in the plexiglass bed. Once everything is set, hit the ‘go’ button and the saw blade is slowly pushed through the cell, leaving a straight, clean cut.
If you’re hands are steady enough to cut solar cells without shattering them, you may be interested in this Dremel-powered solar cell saw.
“If you’re hands are steady enough”
Really, HaD? Really?
Blade should perhaps run in a water bath like a tile cutter, and a bit more power. It should be feasible to have a set feed rate rather than having to manually advance and back it off.
I love the way he removed the ‘unwanted’ circuitry from the scanner PCB. Literally hacked lol He probably used a ‘hack’ saw lol.
While it definitely works well, I can think of two improvements.
One would be to have the buttons control the cut rate rather than just the movement in a fixed cut speed. IE holding the button down makes the stepper go faster and Releasing the button keeps it at the same speed.
The other would be to use a round motor because you can rotate the brush plate to optimize the motor for the correct compromise between torque and speed. The motor he is using is going way too fast and has little torque at that speed.
I was thinking use opto or hall effect to sense rpm and have the arduino moderate the feed rate to maintain a set rpm.
This would probably be my second choice.
As it stands now, it’s a motor that was designed to carry a 30 gram (slot car) at a reduction of about 3:1 and it’s driving a much higher load with no reduction (direct drive).
The phase is far too advanced towards speed so it has next to no torque at lower speeds, meaning that when the speed drops it looses torque and drops further. You can see this in the video as he backs off the cut to prevent a stall.
Feedback can regulate by dropping the speed but it can’t add torque.
Actually it can to a point. The controller compensates for lower RPM by increasing voltage or pulse width. That in turn increases torque. It’s only limited by supply voltage and the strength of the motor and its windings.
Put your car into first gear with the clutch in. Rev the engine to 6,000 RPM. Now very quickly turn the key off and drop the clutch. The tires squeal for a moment but the energy input to the motor is zero (ignition is off) so the torque input is zero. The energy is coming from momentum rather than torque. You can this in the video. The motor slows down (not enough torque) he backs it off and it re-builds momentum – repeat.
“Actually it can to a point” – yes agreed, he has reached this point. Any more voltage and the blue smoke will get out.
I tried to find a good link, RC hobbyists know this well with their electric model cars etc.
Take a round hobby motor loosen the locking tabs on the back so you can rotate the end cap with the brushes in it.
Turning it one way with power applied will cause the motor to progressively slow until it stops. At this point the torque is at maximum and the speed is at minimum (zero). As you turn it back the other way the speed will increase and as it does the torque will decrease until the speed is at maximum (back EMF is equal to the Electro-motive Force) at this point the torque is zero.
Torque and speed are a trade off against each other. This motor is set to a very high speed and low torque and hence the need to build momentum. This in effect is storing energy to compensate for the lack of torque.
The ‘strength’ of the motor as you put it is the angular relationship between the brush plate (end cap) and the commutator. This can be adjusted on some motors.
Most small motor are set for high speed and a gearing ratio is used to set the torque to load ratio. This isn’t applicable in direct drive so he needs to reduce the speed and increase the torque (as a ratio) and this can only be done by altering the angle of the brush place to the commutator.
I think he should laser-etch the newly created cells edges to prevent electrons ‘jumping’ to the other side and reducing the efficiency. But i like the creative table-saw!
So much noise for such a little saw!
You can dampen vibrations by adding mass to whatever is causing the vibration. In this case if you were to bolt a big heavy piece of scrap metal to the bracket holding the motor, then the whole thing would be much more bearable. Alternatively, you can make up big heavy blocks of epoxy granite/engineered stone/polymer concrete by mixing sand with enough epoxy to stick it together.
I don’t think that being around glass dust is very healthy.
Glass? It’s silicon. The problem is the dopants used in PV-cells, it contains Cadmium-Telluride if i’m not mistaken – VERY unhealthy :( There is a reason that PV-cells are excempt from the ROHS-directive. http://www.engineering.dartmouth.edu/~d30345d/courses/engs171/SauratRittoff-Photovoltaics_and_RoHS-2010.pdf
Silicon PV doesn’t contain cadmium or tellurium, although there are solar cells made of cadmium telluride. Those are deposited through an electrochemical process directly on glass. Silicon PV is made through diffusion of dopants into slices of crystalline silicon. The dopants used are boron and phosphorous. Silicon dust is a hazard, but only as a particulate, not for toxic chemical reasons.
The amount of dopant in the silicon is minuscule. You will eat more cadmium in your lunch today. And, like Jerrett said, Cadmium and Tellurium, are not the dopants used for this type of cell. The dopants would be boron (for p dopant) and phosphorus (for n dopant) in this case.
A more general point regarding cutting of solar cells – they should NOT be mechanically cut right through the thickness of the wafer. According to the solar cell people I worked with in my PhD, this creates defects and recombination centers (in simple terms, it buggers up the P-N junction at the front of the cell) leading to a drop in efficiency. Now, I never actually checked whether this was the case, since my cells were rather precious, but I’m inclined to believe it.
Much safer is to score the back of the cell and then snap it. Use the saw to cut part-way through, not reaching the junction, and snap. The company I used (NaREC, http://www.narec.co.uk/) actually uses a Nd:YAG laser to do the cutting, and they can carefully control the cut depth.
(As an aside, they also use the laser for patterning the front contact, instead of silkscreening on silver paste, so the shading factor is less, leading to a higher efficiency. Aherm….)
Sad to see the old type scanners gutted as they are extremely useful for scanning populated PCB and they are no longer made. New ones with contact image sensors just can’t focus that far, so gut them by all means.
You don’t bury your blade in the work. leave your cutting blade protruding high so your cutting with least amount of teeth at once. same with wet saw. least amount of drag is optimal for rpm’s. you can hear him bogging down the blade repeatedly. i know there is not much diameter to work with on that dremmel diamond blade. A better solution, might be an actual tile scoring cutter tool. low tech and not as cool as this hack but work much cleaner.
Find a motor that has more torque.
Mount the motor closer to the scanner rail/bearing to reduce flexing.
Reduce the feed rate.
Better shaft balancing to reduce noise and vibration.