The Semitec thermistor in my E3Dv6 started going open-circuit occasionally after about 5kg of filament printed at various, usually pretty high temperatures (so there’s a single data point right there). Never had a problem with them before, so it might just be a fluke with this one.
I do have a spare Semitec at hand, but i wanted to try an axial thermistor instead since i was really happy with the one my Budaschnozzle used to have. So far, the axial one works well after adding an extra hole to the heater block and some fiddling with the PID values, which makes me wonder - why are E3D not using axial thermistors as the default option? @Sanjay_Mortimer@Joshua_Rowley@David_Lamb
Wow, funny I’ve been having thermistor “noise” problems on my Budaschnozzle recently and finally replaced the axial thermistor last night after almost two years running it. Will test it tonight.
I bought 1% Honeywell models (Lulzbot lists 10% on their parts list) and these were $5 each from Mouser… I didn’t buy 200 I agree that axial feels more secure for this critical item.
@Sanjay_Mortimer said why they didn’t use it last time I was bugging them about it, but I don’t recall the answer, though I do recall not thinking it was a good enough reason.
I believe the reason is that it makes it impossible to offer pre-assembled thermistor harnesses, which E3D has kicked around the idea of doing. Personally, I don’t much see the benefit of axial over a properly fastened v6 with the SHCS+washer. If you wanted added security, I’d do a screw in thermistor over axial.
Also I’m pretty sure that MF58 is only good to 200C.
Да все просто - ток идет по поверхности провода а они мать их окисляются при высоких температурах и в следствии чего меняют свое сопротивление, значением которого мы собственно и определяем температуру… Короче просто спаяй контакты датчика. Скрутка тут не канает ёмае…
@Tim_Elmore that makes sense. Pre-assembled harnesses would be possible with an axial thermistor as well, but would require some design changes to the heater block.
Depending on what datasheet you read, the MF58 is either rated for 200°C or 300°C. Seeing that it is a glass-encapsuled thermistor, 200°C seems rather conservative, but 300°C would be quite surprising for such a cheap thermistor as well. They do seem to work at “normal” printing temperatures - what kind of issues would you expect from a thermistor being used beyond its rated temperature?
Reprappro use them on their Huxley/Mendel hot ends and I am very happy with it. The main advantage I see is that there is no more issue with the perpendicular stress point on the fixing head.
@Thomas_Sanladerer Resistors normally have a positive temperature coefficient of resistance (PTC). The NTC thermistors we use are designed to measure within a range of temperatures where the resistance curve reverses, but if you go above that range, the resistance will begin to rise again, assuming the thermistor doesn’t physically fail (e.g. crack the glass) first.
@Whosa_whatsis thanks for the insight. And that makes sense - even though i swapped out the thermistor table, the hotend still prints as if it actually runs at a higher temperature than what it’s reporting.
I broke my E3D thermistor today so my printer will not work next 2 weeks… time to get a new one, am far from all. I am thinking to buy a new hotend, too many issues whith this one. Thanks for sharing.
One of the major factors is that ATC don’t screw around, carry large amounts of stock, and deliver next day. We were having real issues with previous thermistors, we cleared out Farnell, RS, and Mouser in the space of 6 weeks and our entire business ground to a halt. When you’re shipping literally thousands of units a month supply chains matter, we lost a lot of money, and people had to wait a long time to get v5s because of silly little things like thermistors.
As @Tim_Elmore said, pre-assembly is a factor, and not having the thermistor tied to the block is a very useful factor. It also makes swapping/changing them out really rapid. Which means you can remove all the electronics from the hot areas in under a minute with a single allen key. No cutting of wires, no resoldering or crimping to reattach. @Thomas_Sanladerer How would you propose redesigning the block to use an axial thermistor yet not be tied to the block?
The reason we went with these particular semitec thermistors is because they have a really really snappy response time and you can get that little glass bead right down into the block with excellent contact and placement with low thermal mass.
Thanks for the detailed answer, @Sanjay_Mortimer1 .
First of all, i personally wouldn’t mind a DO34 or DO35 thermistor being tied to the heater block, since they are pretty robust little things even when you secure them just by bending the legs over. However, it could be user-replaceable if the heater block was a two-part design, with the second part being a clamp/trap for both the heater cartridge and the thermistor. Alternatively, using removable, spring-tensioned connectors (similar mechanism to WAGO Klemmen 273-100 3x1 5 Qmm 3polig 100 Stk online kaufen | eBay , i’m sure they’re available as miniature versions somewhere) would allow the thermistor to be used without any permanent connector attached.
I’m not sure how to what extent the different thermal mass and heat transfer between axial and radial thermistors has real-world implications, but my gut tells me it shouldn’t affect accuracy too much. I’m sure you’ve run some benchmarks on this topic at some point, so that’s where your hotend-making expertise comes in.
That being said, i still think the Semitec is a good solution the way you use them - i’m just exploring alternatives here and trying to see what sticks to the wall.
I agree that axial feels more secure for this critical item.