• aditya mehrotra.

filament updates: starting to build the system [updates]

So today, I spent some time putting together parts of the filament extruder from the parts we got in the mail. Here's the rough build-log and some electrical calculations.

So we started by cutting the auger's end off and drilling a 1,75mm hole in the end-cap. Cutting the auger's end means the spiral tip that goes into the wood won't interfere with us trying to move plastic and the hole is so that when plastic comes out of the tip it's 1,75mm in diameter (that's the kind of 3D printer we have) and it should be melted!

The next step was to attempt to mount the semi-finished heating/melting zone we made last time. We cut a hole in the wood, and mounted the L-bracket from the underside.

Then I started doing some research - and we want the heated region to reach temperature of 400-500deg F. Which means we're using ~28 gauge wire and it takes, according to the table, like 3 amps to reach that kind of temperature. So we added some winds in parallel to the current winds, and we ended up with this.

That's roughly 15ohms of resistance. And at this point we're only heating the body directly not the nozzle itself. So we thought - let's power this at 110V AC and then add like two feet of the 4ohm/foot wire and that'll get us up to 5A-6A of current at 110V AC.

So there's the result of that. We didn't test this directly but it has about 20ohms of resistance and at 110 V when we're powering it that's like a little over 1000deg F we're expecting this thing to reach up to. That's an insane temperature and thank god we have some insulation around it. We might need to use the relay a lot more than we initially thought - or add some resistance in series to this coil. We can also think about adding a heated hopper in series but I'd rather not.

So that does mean we can have a slight problem - in that maybe the fiberglass insulation we're about to surround this tube with its going to melt. And hopefully the insulation keeps things out because otherwise the wood that makes up the framing just might catch fire. But the heater coil won't be on all the time. What will happen is that we'll control the heater coil until the thermocouple on the nozzle reads 220 deg F.

Now we can try something else. The original recipe called for 20ft of the wire and 4 ohms a foot - 20*4 = 80, 110/80 = 1.375A! So if we have something that's 20ohms just power it at 110/20 = 25V. We could do that maybe?

Or we don't need to turn on the coil for very long you know. But the much hotter coil temperature allows us to heat the system faster and maintain higher temperatures.

So what we'll do is use the thermocouple on the surface to tell how hot the coil is so we have one method of control there - and the thermocouple on the tip to determine when extrusion should begin.

Then we started to wire some things up. This includes mounting the relay and wiring the resistive wire and the power socket to it. Then we wired the relay to the arduino. Then we hooked up the motor controller to the arduino and the only thing left is the thermocouples and the logic. There's some other mechanical parts to place but in-general, it might be ready to extrude. We need a hopper and things but all the circuity is in-place to have it turn the motor and then also heat the chamber. We're going to use some basic control for that - bang-bang control actually for the temperature and PID for the motor. Maybe we should fix the insulation though so it's metal that's facing both inwards and outwards. Yea let's do that.

As you can see - it looks exactly the same but I promise we flipped it. At least I don't think we'll have to worry about the fiberglass exploding or melting. And because of the insulation - we probably don't have to worry about the wood either. There's another piece of fiberglass insulation we're going to put between the hopper and the entrance to the melting zone just so the hopped doesn't get any funny ideas and like melts or something.

So to clarify here's how this system works. When we turn it on - the arduino is the one that's on and it reads the temperature from the thermistors first, then we put in the material we're trying to extrude and it says "okay I have a target temperature let me turn on the relay." It flicks the relay on and it waits until the first thermistor reads a temperature like 320 deg F, and then it turns the relay off, the moment that drops below 280 deg F, it turns the relay back on, then off, then on, then off, you know - bang-bang control. What that means is heat is transferred better to the pipe. At the same time as its controlling the pipe temperature it attempts to control the nozzle temperature. If the nozzle temperature exceeds 220 deg F the whole system shuts off, if it is below 220 deg F or whatever target temperature, then the system goes back to controlling the temperature of the tube in the same fashion as described above.

When we're at the target temperature, a PID controller takes over and starts moving the motor forwards at whatever speed we desire.

So where and how are we going to write the code? We'll install the standard firmata sketch on the arduino and use: https://pypi.org/project/pyFirmata/ and python to write the logic. The only thing is we'd need to know the port the arduino is on. And before we do any of this we kinda need to wait for the thermistors' resistors to show up. So what that means is in the meantime we get to do something super exciting!!!!

Clean up the basement ... yea... I know.

Also before anyone roasts me, this is supposed to be a quick prototype OK!

#updates #omgrobots #yay #filament #getting_there

0 views0 comments
© copyright 2019 | aditya mehrotra