Comments on: Calibrating the CupCake Part 1: Nozzle Temperature

By: Keith Neufeld

Keith Neufeld — Thu, 24 Sep 2009 18:06:30 +0000

Hoop, because I don’t have a thermometer I trust, I was looking for a physical phenomenon with a known temperature. I’m less certain of the boiling point of mineral oil than I am of that of water, so that’s not a good fit for me.

By: hoop

hoop — Thu, 24 Sep 2009 17:25:13 +0000

In the Past we’ve used mineral oil for heat dissipation , especially in conductive situations, – a Paint can full of oil- as dummy antenna load for 2kw linear RF amps.

a similar mineral oil filled can may be just what the doctor ordered, fill your spam can with unscented mineral oil and give it another go – the mineral oil holds heat much better than water, is electrically non-conductive , and DOESNT bother anything you may have soldered, (or kapton tape)

By: Luis de Rivas

Luis de Rivas — Wed, 16 Sep 2009 03:32:38 +0000

With regards to IC/solder temp/probe(thermistor) temp: IC based solutions are a bad idea; reliability and longevity are poor when stressed to the extremes needed for thermo-plastics. NTC Glass encapsulated thermistors are reliable, quick to respond, stable and have better than 3% tolerance. Platinum RTDs are choice for precision & long term stability.

By: steve

steve — Tue, 15 Sep 2009 18:50:44 +0000

I’ll throw an idea out here for you. If you used an oil, like peanut or safflower, instead of water you can measure to temperature closer to where you want to operate. They have smoke points around 450 F (232C). If you used a good kitchen thermometer, like a fry or a digital probe thermometer, you can plot your thermistor reading over a wider range of temperatures. I doubt that the thermistor has a linear curve so I would want to have multiple readings and have it calibrated accurately to the temperature I want to operate at.

By: Cathal

Cathal — Tue, 15 Sep 2009 12:03:25 +0000

I’ve had a good look and got a response out of National Semiconductor on the high-temp IC idea. Turns out they’re soon to release a new chip that can sense as high as 200˚C! Great, but not quite there yet. Perhaps for lower-temperature plastics? I gather PLA prints at lower temperatures.. though I’m still looking for a supplier of PLA in coil form.

By: Phillip Kliewer

Phillip Kliewer — Tue, 15 Sep 2009 01:34:56 +0000

+1 for Spam eating and can usage!

By: Keith Neufeld

Keith Neufeld — Mon, 14 Sep 2009 18:27:21 +0000

Nophead — yeah, before assembling the whole thing would have been a great time to calibrate the temperature.

Thanks for the note about using PWM but no PID for the heater. I think I need to have a lookie at the extruder code — I can see by watching the indicator LED that it’s doing bang-bang, and maybe we can do something better.

By: nophead

nophead — Mon, 14 Sep 2009 16:54:40 +0000

Hi Keith,
I calibrate my thermistor with a multimeter / thermocouple combination. I put the thermocouple inside the nozzle. It is easier to do when the nozzle is new and does not have plastic inside of course .

You don’t need PID to get good control. I get swings of +-1-3 C with simple on off. The thing that is different is that I convert the temperature to ADC value once on the PC and send the value to the micro which then compares that with the ADC and turns the heater on and off thousands of times a second. I don’t have any table interpolation either.

The black stuff looks like burnt filament. If you heat ABS to extrusion temperature and leave it in the presence of air then it will burn. Best not to leave stationary filament cooking for more than a few minutes as it can block the nozzle.

By: Dave

Dave — Mon, 14 Sep 2009 15:54:42 +0000

For the temperature/pressure of boiling water, you can always consult the CRC Handbook of Physics and Chemistry, which I used to compute about a 1.7K difference in temperature due to your altitude.

Dave

By: Keith Neufeld

Keith Neufeld — Sun, 13 Sep 2009 13:44:53 +0000

Cathal, the IC sensor is a tantalizingly simple and accurate idea.

The snag, I’m afraid, is that common leaded solder melts in the 180-190°C range, significantly lower than the 220°C range of ABS plastic. I suspect that manufacturers won’t have been strongly motivated to make IC sensors with measurement ranges that would have them floating loose from their solder connections.

I had a quick look at National’s parametric search and none of their sensors go about 150°C.

Of course one could crimp (rather than solder) wires onto a TO-92 package. If you find something, let us know!