Claghorn wrote:I just stumbled across the new (at least I'm pretty sure it didn't say it when I got my printer) recommendation in the solidoodle FAQ that says the appropriate temp for the extruder is 215C, unless I have the old motherboard, in which case it is 195C.
Why does the motherboard change the way the thermistor works?
Is the calibration table just different in the firmware? Is the pull-up resistor different leading to a different temp value? What is going on here?
There must me 10000 places that say you should never operate close to 215C or your PEEK will melt, now it says that should be normal? If it is the table in the firmware, will I get yet another different reading if I compile and build my own copy from (for example) lawsy's repo?
And, of course, in the exact same FAQ a few paragraphs above, it still says the nozzle size is .35mm.
How many other totally contradictory recommendations are people following? Is it any wonder so many people have so much trouble getting good prints?
I told the guys who maintain the faq to update the nozzle size. If you are doing work that is highly dependent on nozzle size exactitude, then I highly recommend you measure the relevant proportions with a caliper.
The current set up reads the temperature more accurately. IIRC The table is not different, nor is the resistor. I personally did not make the adaptations for the printrboard. From what I remember, we simply got hold of a printrboard friendly build of marlin, and moved all of our custom Solidoodle settings there. The thermistor tables might in fact be superior, or there could be other factors involved. I haven't had the time to rummage through the firmware to figure out the exact issue.
As for altering the tables, I suppose that would work in theory. The last engineer to play with the firmware insisted to me that the temperature inconsistency was not linear, and would therefore be trickier to correct. Just off the top of my head that sounds bogus, but maybe there is something to it. Certainly, I would imagine that if it were that easy, someone in the community would have made the modification already.
I think development wise, there is a danger in making this particular mod to the firmware. The tables should reflect a very accurate temperature under ideal circumstances in the hardware. It feels kind of like a kludge to change it, but if we aren't going to change the hardware for a while, it kind of makes sense.
At the end of the day, the heating inconsistency is due to the fact that the thermistor is taped to the outside of hot-end, rather than embedded within it. This is something we hope to correct in future generations of hot-ends.
This isn't anything you need to worry about too actively. Any Solidoodle received in the last 6 months or so should work fine with the 215c temperature. The inconsistencies in the documentation are a shame. As a community, we need to work on both the wikis. If you've noticed, there has been a major revamp of the documentation available on the website. I hope you guys see it as a step in the correct direction. Of course we understand the a formal manual would nice as well.
If you see anything on the main website that needs changing, feel free to contact me directly and I shall make amendments. It's sad, but Solidoodle can't afford an army of tech writers. Sometimes this stuff get's written by interns. There *will* be mistakes.
If you have an early Solidoodle 2 or 3 (pre-printrboard) you may want the lower temperature, but the firmware limit should save your butt even if put in a ridiculous temperature. Essentially,the printers should not be able to achieve a temperature over 215c. Maintaining "Legacy" Solidoodles is a problem we are just starting to recognize, and deal with. I'm totally open to suggestions on this front.
In the very early days people were attempting things like 250c. That didn't work out, so we put in the firmware limit.
For the last year, the main source of burn outs have been:
1) Defective hot-ends. We had a rash of those about a year ago. Not nearly so much an issue these days (insofar as they have an acceptable life-span - they still have a lifespan)
2) Thermistor errors leading to faulty readings leading to continuous heating. This is something that still happens, but only in an extremely small number of cases. We may see it once every 2 or 3 months.
In the here and now it is not a problem.
I should also make it clear that there is no single magic temperature that works best for any Solidoodle. Nearly every setting on the Solidoodle will work best if it is zeroed in. Different prints will even have different optimal temperatures. We just try to provide a sensible default for new users.
I can imagine for the future, applying some tidy algorithm to the question of temperature. The print would be analyzed, and an optimal temperature for each layer would be calculated, with the idiosyncrasies of the slicer settings as well as the hardware taken into account. Perhaps even the quality of filament would be taken into account. If anyone here is an expert on control theory, or thermodynamics I'd love to hear your thoughts on how might eventually accomplish this, because I really think it's the future.
For the time being, perhaps we should invest some time into creating a tidy set of community reviewed profiles? Solidoodle is about to release a few new profiles and I'd love to see the community's ideas about how we could make them better.
Former Solidoodle employee, no longer associated with the company.
