nozzle filament output temperature

It occurred to me before, an aluminium block with a thermistor or thermocouple pushed into a hole a few mm from the thread. Often with an airgap not close to the nozzle tip.
Given potential lag trying to maintain a set temperature. especially with air attempting to cool the extrusion and nozzle. which can lead to premature hardening of pla
causing weaker layer bonding.

There is a video somewhere on youtube, where they had kapton tape attaching a thermistor direct to the nozzle. anyone ever try it? what are the pros & cons.
Digital Laser Infrared Thermometers are not overly expensive. How difficult would it be to hack one to use instead of conventional temp sensors.
Pointed at the tip the extrusion temperature would be more accurate.

It always falls or is knocked off.... causing the hot end to keep heating and eventually fire.
Less of an issue with modern firmware that notices its putting more and more power in the hot end that isn't heating up and will stop.

This is why people started attaching hot ends more securely

The difference between filament temp in nozzle
and the temp at the thermistor is miniscule.

The A to D (analog to digital) conversion is probably less precise!

If you really needed to know you could compute
temp effect between :
filament and brass nozzle
Brass nozzle to aluminum block
Al block to material around thermistor
material around thermistor to thermistor

Non contact IR themometers have too large a
measurement area and are dependent on reflectivity of material

So you using a filament material that you
must control temp +- 0.0001 deg C?

Could use:
Infrared sensor of the thermoMETER CTLaser series are used for precise, industrial measurements. Using two laser beams, the measurement spot is marked which ensures a precise temperature measurement. thermoMETER CTLaser temperature sensors are used in numerous applications. From extremely low temperatures (-50°C) to the highest temperatures (2200°C), these IR sensors measure precisely and reliably.
but minimum spot size 0.9mm +- 1% deg C and about $10000

Good Luck


BTY a small alligator clip keeps a thermo-couple on nozzle very nicely

I don't think you need to get too fancy. You can switch to a plated copper block and nozzle with currently available parts, and that will behave similarly to moving the heater and sensor closer to the nozzle. Its possible that combining the nozzle and heat-block into a single component would make the nozzle respond even faster, but that comes with major disadvantages.

If you tune your PID with fans turned on or even while extruding it will hold temperature in use. You could also program software feed-forward since the printer knows how much filament is being melted at all times (and should therefore be able to calculate the theoretical required energy input, only relying on the sensor to deal with slower drift).

So you are saying to copper plate the hotend?
a thin film of copper doing all this?

What pHD in thermal dynamics has stated that this would help?

Perhaps a FLIR thermal camera to show where the polymer
is cooling too fast or is off by 0.0001 deg C

Insanity
-- set temperature +- 10 c should be OK for any polymer

Quote
cozmicray
So you are saying to copper plate the hotend?
a thin film of copper doing all this?

What pHD in thermal dynamics has stated that this would help?

Perhaps a FLIR thermal camera to show where the polymer
is cooling too fast or is off by 0.0001 deg C

Insanity
-- set temperature +- 10 c should be OK for any polymer

Well said!!!!!!!!!!

No, not well said.

Firstly, 691175002 said "plated copper block and nozzle" from which I assume a block and nozzle made of copper and plated perhaps with nickle to stop it becoming oxidized, secondly, even if the nozzle was steel and plated with copper the copper with a conductivity of 386 w/m K versus steel with a conductivity of 36 W/m K would contribute a substantial improvement.

Learn a tiny bit of physics to back up your certainty of knowledge.

Mike

Quote
leadinglights
No, not well said.

Firstly, 691175002 said "plated copper block and nozzle" from which I assume a block and nozzle made of copper and plated perhaps with nickle to stop it becoming oxidized, secondly, even if the nozzle was steel and plated with copper the copper with a conductivity of 386 w/m K versus steel with a conductivity of 36 W/m K would contribute a substantial improvement.

Learn a tiny bit of physics to back up your certainty of knowledge.
Mike

If you Assume¹ 691175002 "plated copper block and nozzle" then you do know.

¹Assume isn't only used to mean "accept as truth without checking"; it also means "take on the form of." It might be safer if one not assume.

I was thinking of these products, other companies produce analogues as well.

[e3d-online.com]
[e3d-online.com]

For what its worth, you generally put the metal in front if you are describing the coating, and afterwords if you are describing the base metal.

Hmm, there is no emoji quite good enough for this one.

691175002 wrote "plated copper block and nozzle", not "copper plated block and nozzle"

my assumption is MUCH more likely than yours

Mike

You could be correct but I doubt that statement, because you are still assuming.

Would not but a ruby tipped nozzle and now will never buy this nozzle so in trying to make sale you have only produced one to despise¹ the Product.

¹ to regard as unworthy of one's notice or consideration.

I should remember from the Desiderata, "avoid trolls for they are vexatious to the spirit"
Mike

So now you strife to name calling to try to prove your pointless antics.

Clif

Should have stated nickle plated copper block hotend
Copper block and copper nozzle make all material the same
but nozzle is nickle plated -- heat break?
is there copper jacket on heater cartridge?

Whats the point? Keeping polymer within +- 0.001deg C
when +- 10 deg C is just fine??????

It seems a bit rich for cozmicray, the master of using only one word when 10 are needed, to say that 691175002 should have stated "nickel plated copper block hotend" instead of the already clear "plated copper block and nozzle". Having said that, let's go to the more important point: cozmicray claims that ±10° is O.K.
My tests show that ±3°C is pretty awful with bands of shiny alternating with matt. ±1°C is better and of course ±0.001°C (or earlier ±0.0001°) are neither achievable nor even asked for by anybody.
The reason I know is that I made some nozzles with a female thread fitted onto a stainless steel tube but the stainless steel acted like a heat break and caused wild temperature swings. I measured the temperature in a pocket drilled in the side of the nozzle with a 0.25mm diameter thermocouple.

As to why I would do such a thing when perfectly good nozzles are cheaply available? I did it to look to see if the nozzle hole aligned with the filament hole, if the sharpness of the edge shown in the diagram was important (it needs to be radiused) and if the inside needs to be polished (it does, up to about 3 times the nozzle hole diameter). Having a female thread made it possible to look at the inside with a microscope - something almost impossible in the deep well on a nozzle with a male thread.

Mike

My twopence worth, I've got some tungsten nozzles [www.kickstarter.com] and they definitely transfer heat more effectively to the filament than brass, when I installed it prints were stringing and looking shinier than before, at the same indicated nozzle temp.

See the deltaprintr mini hotend [deltaprintr.com] which has a hole in the nozzle for the thermistor and a clever heat proof sock that holds it on. Works in this design as the heater block is cylindrical.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

What filament material are you using?

Would like to know what material and manufacturer
requires this tight temperature tolerance?

Is the length of the melt zone a factor here?

Would a hotend insulator help?
[www.thingiverse.com]

In the perhaps forlorn hope that cozmicray is here in his participant aspect rather than his troll aspect, and in the hope that this thread does not become completely negative.

As to the material, I was using black PLA - cheap black PLA that shows up surface defects like nothing else. This was by intent as I was looking for problems.

For the material and manufacturer: I have no idea, nor does it matter what the manufacturer's temperature tolerance is, only how that filament behaves with that hotend, not to mention print speeds, nozzle diameter, cooling etc.. If the hotend temperature is mid range for the plastic it is possible that there may be a better tolerance although I still doubt ±10°C which is a span of 20°C. If however you are working low on the range to get fine detail then temperature control is much more important.

cozmicray entered this discussion with a claim that:-

Quote

The difference between filament temp in nozzle
and the temp at the thermistor is minuscule.

Which I have found not to be the case. He also cryptically said:-

Quote

So you using a filament material that you
must control temp +- 0.0001 deg C?

I can't remember ever seeing anybody claim that that degree of precision is needed.

He was however completely right about the IR thermometer, even 12mm spot at 300mm will be about £200

In his next posting, cozmicray completely misrepresents a posting as talking about a copper plated hotend, an error that he has, albeit grudgingly, admitted to. I will skirt over his faux-pas regarding PhDs, the conductivity of a copper film etc..

On his last two points:-

• The length of the melt zone is not likely to be a factor, this particular hotned had longitudinal ribs on the inside of the heated zone to facilitate heat transfer.
• The hotend is insulated by a shell made of cuttlefish bone which has a lower thermal conductivity than polystyrene foam.

Mike

3DS-QUAD asked the original question
and have not heard from him/her?

Is leadinglights a thermal engineer for
a polymer extrusion development company?

All PLA NOT the same,
all kinds of different material to make up colors.

Many manufactures don't even test or make up
spec sheets?

But your the X-spurt?

Found a good article that may show some temperature effects
on PLA, dramatic surface effects

He does state
"I also destroyed one of these vases at the time to see if the temperature made any obvious differences to bonding strength. I could not find any difference while splitting and ripping the different temperature printed areas."

the original question.

Temperature effects with PLA printing / RepRap News

[richrap.blogspot.com]

Quote
cozmicray
Found a good article that may show some temperature effects
on PLA, dramatic surface effects

He does state
"I also destroyed one of these vases at the time to see if the temperature made any obvious differences to bonding strength. I could not find any difference while splitting and ripping the different temperature printed areas."

the original question.

Temperature effects with PLA printing / RepRap News

[richrap.blogspot.com]

Precisely what I said. Edit: Do you agree then that ±10°C is too broad - at least for surface finish?

On your earlier point, I am no X-spurt or even expert, but I do try things out for myself.

Mike

Edited 1 time(s). Last edit at 10/06/2017 03:14AM by leadinglights.

I'll agree that +- 10deg C does have consequences
if
you will agree:

type of PLA
manufacturer
melt length
glass temp --- melt temp
Flow rate
cooling

Have just as much effect on extruded PLA

sounds like lots of thermal mass, heater power
and a tight PID required

Quote
cozmicray
I'll agree that +- 10deg C does have consequences

Who knows, perhaps the hippies will be proven right and peace will break out all over the world.

Quote

if
you will agree:

type of PLA
manufacturer
melt length
glass temp --- melt temp
Flow rate
cooling

Have just as much effect on extruded PLA

All true and many other things e.g., in the nozzle shown earlier it was actually made worse by the relatively good thermal conductivity of the brass nozzle, had it been a stainless steel nozzle the swings would have been much less.

Mike