Josef Prusa's Next Gen Hot-End (Print Nylon, Peek & PC)

Traumflug Wrote:
-------------------------------------------------------
> The solution would be to use resistors with 3.3
> ohms, which would produce 44 watts of heat.

> There are also drawbacks in using oversized (as in
> high wattage) heaters: temperatures change
> quicker, local overheats are higher, so you get
> more material aging. With or without a cartridge.

Actually there is also drawback in efficiency because, when powered from voltage source with non zero internal resistance, lowering the load resistance will lower power transfer efficiency into that load.

eff = (Rload) / (Rload + Rint)

The closer you get with Rload to the Rint the faster the efficiency will decay. Using load with resistance less than internal resistance of power source will additionally lower power delivered into the load, so doesn't make sense at all. From efficiency point of view the heater should be just powerful enough to heat hotend in reasonable amount of time and keep it hot. Difference may be, depending on ones Rint and how overpowered heater he plans to use, still insignificant, but is there.

Quote
Idolcrasher
eMakers hot-end... requires the use of a small bit of PTFE in the stainless steel collar ... and no one out side of RepRapPro uses it.

It's "low popularity" is maybe due to the fact that there is only one manufacturer/seller and they were often out of stock of spare parts.

But they are great ! I only used V2, V5, mendel-parts V6 and V9, Parcan2, (I've also a J-head and an Arcol v4, but never tried them yet) and the eMaker is by far my preferred.
Also, if you don't count the fan, the emaker is the shortest (the one made by prusa is nice but too long for my taste).

Hence I'm using it in all my FoldaRaps, and I'will certainly use it in every of my others/upcoming machines/designs
(but I tweaked it a little to make it even more compact, or to add eventually a second nozzle : [reprap.org] )
(I'll also make a carriage to use them on the Orca's at the school.)

Quote
Pazu
Funny, I have the original emaker huxley and the hot end does not have a ptfe inserted. I dunno why they decided to change that cos IMHO, that they had an excellent design that is extremely robust.
I once had a failed print that stopped for unknown reasons and the hot end was frozen at 230C for 8hours leaving the ABS completely charred. All I had to do was to restart up the machine and manually turn the extruder and it cleared.

As said Pazu, I'm curious about why they dropped the first version which didn't used the ptfe, I don't see a difference between the new ones and the old (I'll keep it preciously to eventually extrude some fancy polymer).

On my eMaker Huxley I had only two hotend problem in a year and I use it a lot (not 24/7 but it's still my main production machine and it printed most of the 600 parts of the 30 Foldarap I'm making). The first time the tip of pulling back the filament after letting it cool down a little was enough, the second time I had to disassemble it, but the 3 pieces are easy to handle, and make it modular (even if I'm quite fan on one piece hotend ideas, as I never switched a nozzle just to change the resolution).

---

about // liberapay // flickr // wiki // thingiverse - github

Emmanuel Wrote:
-------------------------------------------------------
> On my eMaker Huxley I had only two hotend problem
> in a year and I use it a lot (not 24/7 but it's
> still my main production machine and it printed
> most of the 600 parts of the 30 Foldarap I'm
> making). The first time the tip of pulling back
> the filament after letting it cool down a little
> was enough, the second time I had to disassemble
> it, but the 3 pieces are easy to handle, and make
> it modular (even if I'm quite fan on one piece
> hotend ideas, as I never switched a nozzle just to
> change the resolution).

Yeah I only had clogged nozzle once in the whole time I've used it. I blame it on the cheap chinese filament I use. But I still use plenty of cheap chinese filament and issues are far and few between. I love the shortness of the whole thing too and its compactness. I've never use some of the other popular nozzles like J-Head but I feel they are over-engineered. The only thing I don't like was the bowden which I got rid of. LOL.

Here's how mine looks now with optional 0.3mm nozzle installed.

Pazu in your photo it looks as if the (pink/purple) filament is unconstrained between the extruder and the hot end - is that the case or have you just separated them for the photo?

Oh its actually red filament with a short PTFE tube that was part of the old bowden setup. Won't work if it's uncontrained as you had probably guessed.

But really it could have just been any form of filament holder.

Quick question regarding heater blocks (regardless of resistor or cartridge method):

If increased heater block size is viewed as a detriment to melt zone temperature stability due to convective and radiative heat loss during movement (and even more so if a part-cooling fan is blowing across it), would it not make more sense to cover the heater block walls in a material with a low thermal conductivity and high thermal stabilitiy (perhaps some sort of fiberglass impregnated epoxy or even a heat shield of sorts)? In this way, more of the heat generated is forced to conduct into the nozzle allowing for a more stable and predictable temperature with less reliance on the PID routines (regardless of efficiency, effectiveness, efficacy, et cetera or in the unlikely cases that the laws of physics are temporarily suspended -- look for intricately detailed theories and reports of these incidents in your favorite journal of psychology...

Sure, insulating the heater block makes sense. Most simple way: a few layers of kapton tape.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

At least a few hot end designs insulate the heater block with ceramic tape to my knowledge.

Nophead - can you think of an easy failsafe for the cartridge heaters? I started looking at thermal links, but didn't find anything in the 260-300c sorta range I wanted.

I like the premise, but I've run one of the cartridges up to 400c before I chickened out - and ceased to have an accurate way to measure the temperature. I suspect based on the math I could get it up to close to the melting point of AL at 600 with the 38amp 12 volt supply I'm using..

Really applies to resistor based heating as well - I'd like something that could prevent an overheat of the hotend regardless of the state of the driver board - a true failsafe.

Right now I've settled for a smoke detector right above the printer!

... you can use a short piece of solder, tin (232 degC) or another 'lowtemp-melting' metals or alloys with higher melting temps as 'oneway-fuse'.

For regenerating fuses you can disassemble electric water jugs or steam generators ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

bobbitay - thanks, I admit I haven't been around the reprap/3dprinter world very long -- I have lots of research to do

SystemsGuy - Not sure which calculations you're using to get the watts-to-temperature relation (not saying it can't be done -- would be interested in seeing the math), but I would think that a blade-style (automotive type) circuit breaker might work for the application. These are commonly available in the same amp ratings as automotive fuses (6, 8, 10, 15, 20, 25, 30A) for around $10-$15. Different ratings would of course correlate (in some way) to different maximum cut-out temperatures and save you the cost/hassle of replacing fusible links.

I also haven't found any high temperature thermal fuses. I simply used 6R8 resistor which maxed out a little below 300C. It seems people are more impatient nowadays and need more power.

I have some HMP solder with a melting point of 300C so perhaps that could be used as Viktor suggests, although using a home rolled fuse with a cartridge heater doesn't make a lot of sense. Also I think solder tends to oxidise away when it spends a lot of time close to its melting point. It probably needs to be in an inert gas.

Edited 1 time(s). Last edit at 01/14/2013 12:35PM by nophead.

---

[www.hydraraptor.blogspot.com]

... I've opened some comercial 'one-way' temperature fuses with specified temps of 102 degC and 115 degC - it's essential a spring sligtly stretched and soldered with a droplet of solder at one end, so it will separate/release instantly, when the droplet get fluid.

They had a common TO-220 housing, but could be, that it was airtight closed with inert gas filling ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

I
'Ve had some really small resistor sized ones before ill dig them out and post the part numbers and sources

thejollygrimreaper Wrote:
-------------------------------------------------------
> I found if you want to make a practically non
> existant transition zonea long constriction works
> very effectively and so far i've managed to get a
> meltzone down to 7mm long with abs and pla flowing
> like it was butter on my last prototype i flared
> the inside just after the constriction and
> interestingly even though there was more molten
> plastic in the meltzone it reduced the extrusion
> force a little bit, the problem then being is that
> retraction wasn't as effective,
>
> i think the ideal hotend should have less than a
> 5mm meltzone with a heater block that acts more
> like a heat capacitor than just a mounting for a
> heater cartridge or a resistor, and this is where
> heater blocks could be made of brass or somthing
> that "holds" the heat rather than dissipating it
> anyway and everywhere,

Ok, now hold on - heat capacitors?
I suppose they are 100% efficient at producing heat too?
Nophead?

Yes they are. All the electrical energy they consume is turned into heat. They don't consume much though. In fact, a perfect capacitor doesn't produce any heat at all because the current is 90 degrees out of phase with the voltage so the product is zero. A practical capacitor has some resistance and also dielectric loss, both of which produce heat.

---

[www.hydraraptor.blogspot.com]

I didn't realize that. The only time I've seen heat capacitors is right before they burn up.

So, I don't really want to go here again, but if you built a resistor with the current 90 degrees out of phase with the voltage...

Edited 1 time(s). Last edit at 01/21/2013 01:26PM by Tekwizard.

To me it looks like that by term "heat capacitor" thejollygrimreaper meant object or device with considerable ability to accumulate heat, not electrical charge.

Tekwizard: You can of course build resistor with current (almost) 90 degrees out of phase with voltage, it just will be called capacitor or inductor, with all of its properties. Its analogous to growing apple that looks and tastes like orange - it will not be apple any longer, but orange.

essentially thats what i meant, the ability to hold onto energy in the form of heat , it's bit of balancing act

the more thermal mass .(.? is that the correct term) the more lag you generally have when the pid loop reacts to a temperature drop because you have to heat the whole "heat capacitor" where as going with smaller and smaller heater blocks you can "run out of heat energy" before the pid loop has enough of a chance to catch up and can cause a jam up or can have an effect on the print

this is why it's very very important to make sure you have a accurate thermistor table and the well tuned pid loop, i think a very large proportion of hotend troubles have been to do with temperature control rather than the design of the hotend,

the all metal hotends i'm producing now really don't need any active cooling on the cold side even with pla, providing that the pid loop is tuned properly, which in most cases it isn't and isn't possible to tune properly because of other factors, eg wirelengths, noise, fan placements, powersupply, temperature table doesn't properly match the adc output stuff like that

perhaps i missed it a little in previous posts but do people realize that up to 1ohm resistance is in the wiring itself? I see posts about using lower resistance resistors which will cause the power load on the wire to increase as well. so don't just worry about the hot end. perhaps a fuseable linkage in series to the wire to extruder, and using a lower gauge wire?
[en.wikipedia.org] also talks about fusing current of copper wire.
a little bit of Google search here is a web page that talks about recommended current for wire gauge [www.powerstream.com]

some people don't, i've seen plenty of prusas with a huge rats nest for wiring and the owner wondering why their temperature is reading funny or one of their steppers aren't turning, but cannot track down where the problem is because the wiring is a mess and too long

Wow!
About that light bulb analogy. Yes a 100 watt lightbulb produces less heat than a 100 watt heating element, because some of it left the bulb as light. If you had an led with a 100 watt input you'd get a lot more light and less heat. However put each of them in same size boxes of the same material and the boxes will obtain the exact same temperature, because all the light striking surfaces will be absorbed and turned into heat, the lowest form of energy.

I work with pumps, and frequently have to figure out whether a system will overheat or not. With a pump all of the energy eventually winds up as heat.

Now on to a more serious issue. Why stainless steel? I'd like to suggest that it was used because it has great corrosion resistance which is probably going to come into play at higher temperatures. However more important is that relatively speaking it is a poor conductor of heat. So that means the heat introduced at the tip will not be able to make it's way all the way to the cold end very well, which is the goal here. Also the fact that it is a poor heat conductor only means that a bigger temperature differential between the heating element and the molten plastic will be required. So for the same plastic temperature you'll probably have to crank up the heater power a bit.

Don't get me wrong, I love the RepRapPro style hot-end. I was just pointing out its low rate of adoption. I think folks would love it if they tried it.

On the thought of insulating the heater block... I have been thinking about the possible effects of a 3mm thick layer of kapton (read: many layers) on the heater block... That is the thickness often used to insulate parts and pieces of satellites in space...

For the first few pages of this thread I would have voted to ban tekwizard as a troll. Did he seriously equate conservation of energy to thinking that the world is flat?

On a somewhat unrelated note:

Just so everyone is aware, the stratasys patents are publicly availaible and have dozens if not hundreds of detailed drawings and explanations of how to build a variety of different hotends. It even clearly describes the construction and disadvantages of their initial designs.

Not only that, but they include an overview of all viable hot end materials (aluminium, steel, graphite, ptfe, peek...), coatings (to lower friction), and even suggest bore diameters and geometry to minimize jamming and improve performance...

Of particular interest is their work on the location of the melt zone which can significantly improve print quality.

Consider the following diagram showing identical hotends with different melt zones:

Figure 2c shows an extruder at rest, where the heat has traveled up the filament and raised the melt zone. Figure 2b shows a working extruder where the introduction of cold filament has cooled the top of the hot end and lowered the melt zone.

Note that the volume of plastic in 2c is greater than the volume of plastic in 2b. In other words, even if you are pushing plastic through the hot end at a constant rate, plastic will not leave the extruder at a constant rate unless the location of the melt zone is held constant. The patents describe both the conditions that cause the melt zone (or meniscus) to move, as well as methods of fixing the melt zone to a particular location and controlling its length.

[www.faqs.org]
[www.faqs.org]

I've yet to see a hotend design that makes any use of the work stratasys has done.

Emmanuel Wrote:
-------------------------------------------------------
> It's "low popularity" is maybe due to the fact
> that there is only one manufacturer/seller and
> they were often out of stock of spare parts.

They have sold plenty of hot ends... with their kits. However, they are available (on and off) as spares. Like someone else said, the RRP hot ends don't have the wider support of different extruder designs to mount them on, so perhaps have a lower general awareness. Also, they are designed for 1.75mm filament with no 3mm filament version, so perhaps another reason they aren't as popular.

> Funny, I have the original emaker huxley and the
> hot end does not have a ptfe inserted. I dunno why
> they decided to change that cos IMHO, that they
> had an excellent design that is extremely robust.

From what I understand, the first batch of hot ends didn't have a ptfe liner, but when the ambient temperature was high (ie Australia at 40C) there were a few problems with blocking, so the short ptfe liner was put back in.

This is the general problem with hot end designs; they work for some people but not for others. Sometimes there's no obvious reason, and it takes a number of design iterations to work out the bugs. That's why I think the J-Head is so popular and successful. It built on previous work with some sensible improvements, but it took time. When you take a different path, like Josef has, you're going to have to work through the issues. Personally, I can't see what stops the heat from rising up the barrel - there's no active cooling. Yes, using SS with it's low thermal conductivity slows the heat creep, but on the flip side, it doesn't shed it's heat to air well, so once it's hot all the way up the barrel, it stays hot! Add in high ambient temperatures and it's quite possible that it will be easy to block. I've used a mendel-parts v9, which is basically a long brass (and so very thermally conductive) barrel with a tiny (and slightly ineffective) PEEK break in the middle. To get this to work effectively and not block, it can't be left without extruding for even a short time, and needs to be switched off immediately at the end of a print.

Also, the direct drive extruder on the i3. Everyone is jumping on the stepper motors with gearboxes. Yet they seem to need a heatsink and/or fan just to cool the motor! Surely that's not a good step forward? It reduces the complexity of the extra printed parts, but increases the complexity in another area.

691175002 Wrote:
-------------------------------------------------------
> For the first few pages of this thread I would
> have voted to ban tekwizard as a troll. Did he
> seriously equate conservation of energy to
> thinking that the world is flat?
>
> On a somewhat unrelated note:
>
> Just so everyone is aware, the stratasys patents
> are publicly availaible and have dozens if not
> hundreds of detailed drawings and explanations of
> how to build a variety of different hotends. It
> even clearly describes the construction and
> disadvantages of their initial designs.
>
> Not only that, but they include an overview of all
> viable hot end materials (aluminium, steel,
> graphite, ptfe, peek...), coatings (to lower
> friction), and even suggest bore diameters and
> geometry to minimize jamming and improve
> performance...
>
> Of particular interest is their work on the
> location of the melt zone which can significantly
> improve print quality.
>
> Consider the following diagram showing identical
> hotends with different melt zones:
> [img32.imageshack.us]
>
> Figure 2c shows an extruder at rest, where the
> heat has traveled up the filament and raised the
> melt zone. Figure 2b shows a working extruder
> where the introduction of cold filament has cooled
> the top of the hot end and lowered the melt zone.
>
> Note that the volume of plastic in 2c is greater
> than the volume of plastic in 2b. In other words,
> even if you are pushing plastic through the hot
> end at a constant rate, plastic will not leave the
> extruder at a constant rate unless the location of
> the melt zone is held constant. The patents
> describe both the conditions that cause the melt
> zone (or meniscus) to move, as well as methods of
> fixing the melt zone to a particular location and
> controlling its length.
>
> [www.faqs.org]
> [www.faqs.org]
> /
>
> I've yet to see a hotend design that makes any use
> of the work stratasys has done.

have you seen an arcol hotend, or one of my hotends,?

both hotends have a constant meltzone length, on mine my experience is that pla and abs filament is solid right until the end of the constriction and practically have no "transition zone" making them virtually "unjammable" even with the crappiest filament,

the 1.75mm V2 and the 3mm V1 of my hotends that are about to go up for sale (possibly this weekend) both have thermal barriers that are so effective that the cold side heatsink is allmost not even required and now serves as nothing more than somthing to mount the hotend with, even the fins now are really there more for decoration and are somthing i'm looking to get rid of in future versions , both depending on the filament that you use are almost retractionless but not quite

I personally don't think much of the work stratasys has done simply because it's based on using filament that has a stupidly low tollerance much like the infamous up! printer, the filament we tend to use has very different properties and as such has different design requirements,

to put it in perspective we have filament drive mechanisms and hotends which can print with a very high degree of quality with filament that ranges in size from 1.6mm to 1.9mm (i'm talking the worst of the worst chinese filament) with a reasonable selection of materials eg PLA ABS NYLON PC and in most cases at much faster printing speeds,

a classic example is with up!, the up! extruder has a very fine tollerance on the filament it can use simply by design, the pinch wheel and gear setup requires filament that is not less than 1.7mm otherwise the gear will strip the filament, the hotend has a tube size of 1.80mm (pp3dp have verified this i can show you the email) so if the filament gets too big it'll jam up, infact when i tested a brandnew one there was an incredible amount of resistance just passing filament that measured a perfect 1.75mm through it, which is why the oem filament is actually made to a 1.73mm target....... yet they publish a filament tolerance of 1.75+- 0.1mm ( doesn't add up does it, i actually had the argument with them over that as well, they seem to live in a world where filament that can get to 1.85mm in diameter will still fit down a tube 1.8mm in diamater)

if anything (and i hate to blow ones own horn and sound arrogant, but i don't do it very often) they should be learning from us (the reprap community) as they have had decades to build and develop 3d printer components and are still stuck on designs that require crazy tolerances and proprietary feedstock

I was inspired to have a go at making one of these with a few changes.

The design has a 0.4mm wall thickness all the way up to the M6 thread and additional heat sink to prevent binding and heat travelling up the barrell. The heatsink is not connected to the block and only connected by the tube. It is drilled with a 0.5mm hole and will start testing soon.

Reason for build it to reduce the stress on the Peek as it would soften during printing of abs and jam and push the head off.

that heat sink needs to be further up, and the stainless steel barrel needs to be shaved just above the heater to reduce heat transfer above the hot end. i would say 25mm up if barrel shaved, and 50-80mm up if barrel is not made thinner above hot end. even with stainless the cool region begins way to soon.

this hot end has not been tested yet has it? edit: and this is not the prusa designed hot end.

Edited 1 time(s). Last edit at 01/29/2013 05:20PM by jamesdanielv.

Hello

pruse love this design =) - i belive in keep it simple and nice


i have a new design for extruder that im testing right now

my extruder very easy to assamble ' doesn't need cooling at all and can work with 400 celsius no problem .

i can print 200 mm per sec with 190 celsius with direct drive .

ill post pictures soon

elidor

Elidor, 200mm/s? what plastic and filament diameter?