General question about hot-end temps: when I command a hot-end temperature (say, 200C) and my printer i/f (in my case: Pronterface) signals “achieved”, is it reasonably certain that some part of the hot-end assembly is actually at 200C? Can I expect to verify that temperature using an external temperature probe?
I’m trying to understand if “my” 200C —on my printer— is approximately equivalent to “your” 200C — on your different printer. Within, say, +/- 2C.
Short answer: If it is working correctly, yes.
Long answer: All you can really say is that the thermistor is “probably” reading 200C. It may be buried inside a hot end which has an R(theta) (heat conductivity) such that the outside of the hot end is cooler than that.
On my Replicator I put a Fluke meter with the thermocouple attachment on the nozzle and verified that when the system said 240C the nozzle was at 240C.
That said, the typical ‘hot end’ has a temperature profile that is as cool as possible where the filament feeds in, then has a sharp up slope in the ‘melt zone’, maintains sufficient temperature for extrusion at the nozzle diameter without enough back pressure to push melted filament back up the feed path, and then is slightly cooler post nozzle to insure uniform diameter of the extrusion.
Generally there isn’t sufficient feedback on the extruders to actively manage this profile and so all of the printers I’ve seen so far essentially model the profile based on a fan speed to cool the feed side and a temperature reading from a thermistor in the thermal sink of the hot side. Of course those models are susceptible to errors from variations in humidity (changes the heat capacity of the air flow), nozzle composition (brass vs copper vs aluminum vs steel etc) and thermal conductivity between the ‘cool’ side and the ‘hot’ side. Thus a lot of learning about a particular printer, sitting in a particular place, are what are the parameters that work for a particular filament.
This is a more complex question than meets the eye.
What we are trying to do here is measure the temperature of the molten plastic in the hotend. Now we can’t just stick a thermistor right into the molten plastic, so we’re stuck measuring the temperatures of some part of the hotend.
First thing is the thermistor itself. Most are 5% thermistors, i.e they may vary in reading by +/-5% from reality. Additionally the balancing resistor on your board may also have a pretty shoddy resistance tolerance which could make things a bit worse than 5%.
Next is placement, where the thermistor lives in the hotend affects what it reads, generally the closer to the plastic the better.
This now all gets a bit complex when we consider that when you’re extruding faster the temperature of your extrudate is actually going to be lower than if you extrude faster. This is because faster extrusion means less time inside the hotend to equalise with the temperature of the walls.
TL:DR Don’t listen to people telling you that you need to print such and such a material at exactly 207C for the best results, as each setup is going to vary wildly and you’re just going to have to learn how to tune your machine for each material you’re printing.
Two really great answers above! They are both correct.
A great idea to verify temp is through the use of a fluke or similar meter. I myself check this variable in the same way.
I wanted to add as well, an issue I’ve seen time and time again, is making sure your firmware is correctly set to the thermistor you are using.
I.e. Marlin allows for use of a multitude of thermistor resistance. If your thermistor resistance setting varies from the thermistor you are using, you could be in a different “ball park” all together.
A lot depends on the design of the hot end and the nozzle, and even on the feed rate. Adding fan cooling, for example, can cool the tip of the nozzle where the filament comes out, which in turn cools the filament, but only when the filament is feeding slowly. Every machine and every machine placement (unless you’re inside a heated chamber) will be a bit different, but usually not enough to make a complete mess of things.
I found my fluke thermocouple to be very sensitive to ambient temp fluctuations. It may be that only the most expensive ones do proper ambient compensation but mine (116) is useless. If the AC/heat cycles in the room it can swing 5-10 degrees reading. Funny because this is the “HVAC” model.
I insulate my heater block and stuff the thermocouple probe inside the insulating jacket.
Thanks VERY much for the great answers!
Visual effect: noob doing lots of head-scratching
I asked because I was using mfg-recommended hot-end temps and getting poor printing results. I attempted to measure hot-end temps and correlate the results with temp settings — a complete fail. Maybe due to poor instrumentation — a low-end non-contact infrared device, and a sharp-pointed thermocouple-probe device— or poor technique. Both handheld and very shaky.
Then I ran some experiments, noting results every 10C. I started getting better prints.
At this point, I’m considering the temp setting an index that’s hopefully very linearly correlated with actual temp, all other things being equal. (Yeah, that’s the trick, making all other things be equal.)
It seems to me that the next step for me is to take the time to make a series of test prints at temp steps from can-barely-extrude to the maximum achievable. (If I start too low, try to print too cold, am I risking damage to the printer? What should I worry about at the high end – is it possible to achieve a damagingly-high temp?)
Print what? What critical features should be included?
Are there standard designs for this purpose? Seems like there ought to be…
If its too low you risk a clog. If you 're ok with taking apart the hot end and clearing it you should be fine.