Learn how to tame the polycarbonate beast and what treasures await once you’ve mastered it!
Have you had any success with it so far?
http://www.youtube.com/watch?v=a2E6xTOFVTI
Here’s a littleRP-branded form1-labs twisted rook: http://www.thingiverse.com/thing:399545
You can’t use a thermistor above its maximum rated temperature for a couple of reasons. First, it won’t just give you a less accurate temperature reading, the lookup table it uses to convert ADC readings to temperatures just ends, and there is no way to calculate any temperature for these unspecified values (which in many cases are pretty close to the end of the ADC’s scale anyway). Secondly, and more importantly, you cant just add values to the end of the table because the NTC (negative temperature coefficient thermistors we use are designed to operate in a relatively narrow range of temperatures in which the normal relation between temperature and resistance (resistance rising with temperature, for a positive temperature coefficient) reverses. Outside that range the relationship between temperature and the ADC reading will reverse, so assuming the thermistor is not destroyed (which, as a safety feature, they probably should be engineered to do), it will begin reporting a lower and lower temperature as the temperature increases above its maximum temperature.
The maximum ratings of a temperature sensor is not a suggestion, and your firmware should be configured to stop heating well before this point.
BTW, you should try Protopasta’s PCABS. It’s an awesome material. Some of the best bridging and overhangs I’ve seen, and it works better for the diagonal printing method I’ve been using than any other material I’ve tried.
He’s not kidding about the PC-ABS. It’s amazing, the closest to stratasys parts I have ever cranked out. Bridges, overhangs, strength, stiffness and toughness. It’s got it all. Support material removal is incredibly easy too.
@Whosa_whatsis
Having designed ultra high accuracy temperature measurement instrumentation for over 30 years using thermistors (and other types of temperature sensors), I feel compelled to correct the misinformation posted above. First, you CAN use a thermistor above its rated temperature, and it will not result in a “less accurate” reading. The only issue with exceeding the manufacturer’s rated upper band is that the thermistor can shift somewhat to a higher nominal resistance, however in the real world, this happens only occasionally. Second, using a lookup table is a very lazy way to convert the A/D reading to engineering units, you need to use an equation, such as a high order polynomial. If you are too close to the end of your A/D input, then you haven’t designed your analog circuitry properly. Thermistor are NOT designed to operate over a narrow range of temperatures, and in fact are easy to use over ranges of many hundreds of degrees C, particularly the glass encapsulated models. Thermistors are NOT “engineered” to be destroyed (for safety or any other reasons). Thermistors do NOT report a lower and lower temperature as temperature increases above their maximum temperature, they do not “report” anything. They simply exhibit a lower resistance.
I’ll bow to your 30 years of using thermistors outside their manufacturer’s specifications, but you obviously haven’t been doing so on 3d printers. They use lookup tables because they need to frequently calculate temperature on an already overtaxed 8-bit microcontroller that runs at 16 Mhz and doesn’t have a lot of hardware acceleration to run those polynomial functions. The equation is used to create the lookup table so the calculations that would be taxing for the processor only have to be done once (on a regular computer, before compiling firmware) rather that requiring a slow microcontroller to try to run them several times a second.
I didn’t say that thermistors were engineered to be destroyed, I said that they should be engineered to do so, with a failure mode that will stop any temperature measurement from occurring before they exceed the point where their readings become ambiguous.
Thermistors obviously don’t report anything. The circuit that includes the thermistor combined with the software that interprets the data does report a temperature. How do you ever get anything done if you can’t interpret such obvious shorthand?
There are a lot of things in the design of these low-cost 3d printers that a “real” engineer would be appalled by, and assuming that everything works the way a “real” engineer would have designed it is a good way to break things. The fact is, though, that these machines work, some of them better than the machines designed by those “real” engineers, for a fraction of what those machines cost to own and operate.
@Whosa_whatsis some of the thermistor tables actually go up to 350C or even further. Agreed, there be dragons, but like you’ve said, we’re using plenty of things outside of their specs already and most of those work, too.
Exceeding the temperature limit by 10 or 20K probably isn’t going to make a huge difference. Sure, Steinhart-Hart doesn’t really produce accurate values anymore, but printing temperatures are trial and error anyways.
The only real downside is that it might age the thermistor prematurely, but again, that’s probably not much of an issue because of tiny amount we’re exceeding the specs here.
If i get a chance, i might do a test run on a couple of thermistors. I have the MF58, Honeywell 135-104LAG-J01 and ATC Semitec 104GT-2, which i can benchmark against a thermocouple.
@Sanjay_Mortimer1 will do! Eventually…
@ThantiK i’ve seen that model before, but while it looks similar to the real deal, there are quite a lot of differences to it that might have your print turn out completely differently.
@Thomas_Sanladerer you’ll need to compile new thermistor tables into marlin to do that. I know that when I was first trying PC, the table for the Honeywell thermistors we were using only went up to about 270-280, and the dragons came out and started wreaking havoc when we tried to set the temperature higher. I believe the one in marlin now is the one generated, which ends at the manufacturer’s maximum rating of 300c.