Tuesday, 23 August 2011

High power hot-end for fast printing - dual resistor & easier thermistor

This is the first of a few new hot-end designs I have been testing, I mounted it to my new ‘Bling’ machine for some speed and temperature testing, it’s performing well, here is how I made it - 

I decided to try a new shape for the heater block, normally you want a low profile block with a short hot-zone, but I specifically wanted to try a triangular shaped block for a number of reasons / goals - 
  • I wanted to mount 2 x Resistor heating elements in the block so a wide base made this easy.
  • I wanted to try a longer heating zone to see how fast I could extrude.
  • Triangle gives smaller mass at the top, hopefully to reduce oozing due to the longer heating-zone.
  • I wanted to be able to easily wrap the block in PTFE sheet, a triangle gives a easy to wrap shape and is insulated at the top with two thickness of PTFE.
  • For the future I want to be able to extrude fast @ 300 degrees C with less than a 4 degrees C temperature drop while printing solid layers.

Next I REALLY wanted to switch away from using the tiny dipped thermistors the size of pixi-dust with ultra thin connecting wires, they are such a pain to use and don’t really have many plus points apart from the low thermal mass/quick response. I now only use a ‘normal’ Axial style. I can fully recommend doing this, it’s 1000 x easier to build and much less fragile, I also think I get much better temperature reading stability, response time is fine.

Thermistor Type:NTC; Resistance:100kohm; Thermistor Tolerance:± 10%; Beta Value (K):3974K; Operating Temperature Range:-60°C to +300°C;

You can buy them from Farnell -  only £0.25p

I have used 2 x 10ohm Vitreous enamel resistors in parallel; this gives approximately a 5ohm load where you would normally use a single 6.8ohm resistor.
The resistors are either side of the nozzle, giving nice constant heating.
The axial thermistor is slightly above the resistors and is just touching the side of the nozzle barrel, this placement gives me a quite constant 3 degrees C lower reading than the inside of the nozzle throughout the temperature range, which is fine.
Wrap the thermistor body and legs in Kapton to insulate.

Use Aluminum foil or tape around the Resistor body, it should be a good tight fit in the block. 

All in place and secured with Fire cement.

The part make-up looks like this. (Ignore the PTFE washers; I didn’t use these in the end).

I also wrapped the heating block in a strip of 0.5mm PTFE sheet.

Still using PEEK in this design, I expect to remove this part in the next version.

All wired up using high temperature heat-shrink and silicon coated wires.
And assembled.
I added a small heatsink, but I doubt this is really required.

The PEEK block is only luke-warm to the touch even after 4+ hours of printing – Note I am using a small 40mm directional fan aimed at the PTFE.
I will do some more tests to see if I can switch off this fan and still keep a cool cold-end.

My Fan-duct for Greg's Linear bearing X Carriage is on Thingiverse here.

It’s printed just over 2Kg of material so far at ~160mm/Sec and it’s performing well; I’m really happy with it. All of the multicoloured Faberdashery printing in my previous blog update was done with this setup along with some ABS parts and build sheets.

I have not yet found an upper speed limit as I need to do more testing, but it manages to print PLA at 180mm/sec with only a 2 degrees C drop in temperature when doing whole build-plates of full infill layers.

I will be testing ABS more extensively with it over the next week and seeing how fast I can get it working with different materials.

My main observation now is that you do need to run this style of hot-end at a minimum of 80mm/sec, and I need a 1.5mm extruder reverse to make sure oozing is not a problem. At speeds lower than 60mm/sec it’s better to use a smaller heating zone/block.

Next time I will still use a triangular block but make it slightly smaller overall and a little shorter.

Thanks for looking, I would be really interested in any comments or opinions on high-speed printing and hot-ends designed for speed and also high (290+ temperature) - you'll see why in a future post....



  1. Hi Rich,
    Another very informative post, nice one. You mention oozing around the discussion of minimum print speeds. Please would you tell me the symptoms you experience when printing below 80mm/sec?

  2. Hi Woodsmoke,

    When the filament is in the nozzle at temperature (~215 deg C)it naturally wants to push out of the nozzle end, most nozzles do this anyway but with this design the natural flow is slightly faster at due to the length of the heating zone and slightly higher temperature (because I'm going to run it faster). What I can see with this design is that with a modest reversal move of 1.5mm as long as you run at 80mm/sec or more the physical speed and stop-start movements are fast enough that the natural flow/expansion does not result in any extra filament being deposited during the start-stop or between print moves.
    Otherwise you will see blobs or extra material 'oozing' out during the start/stop or during in-air movements.

    I'm sure I could also go slower than 80mm/sec but with a longer reversal move, but I don't really want to increase the reversal too much. It does not want to be too long or the time to do it also has an impact and you also don't want to be sucking air back into the nozzle. So for this extruder I'll run it at 80mm/sec or faster.

    That's my understanding at the moment, if anyone has any better knowledge about hot filament flow with regards to length of heating zone I would be very interested in any feedback.


  3. I'm skeptical that the triangle block is going to distribute temperatures significantly different than a square block. In my experiments I have been quite amazed at how well aluminum equalizes its temperature. I've found that you need an actual thermal break to fight the equalization. After the break you even want more thermal mass not less so that it can more easily dissipate the heat.


    btw. I am going to give those thermistors a try. Not even talking about fragility I think they are a win when using through hole resistor based heating because of how nicely you secured them to the resistor leads.

  4. Hi Billy,

    That's an interesting hot-end design in your video. it looks like you are using quite a lot of force to extrude, what is the nozzle size? also you may be better off using brass for the nozzle, good luck with it, I will keep an eye on your progress.

    The main reason why my heating block ended up as a triangle is that I needed room at the base for two resistors for heating and I didn't want too much mass so I just cut it in a triangle, other than giving a longer heating zone I do agree with you that it's not much different from a square block, but maybe a little less overall mass.

    I'm sure you will like using the axial thermistors, they are great.


  5. I've not seen too many videos of people pushing filament by hand so I don't have a lot to compare against. Do you have any up anywhere? The only real comparison I have is with my MK5 hotend and my little test metric is to see how much I can push for the count of 10. For my MK5 that's about 10cm. This one will do 12 or so. I think my record for all of my random hotend designs is about 20.

    I put up a new video where I'm pushing using as much pressure as I'm comfortable with while it is mounted in my printer.

  6. Hi Billy,

    I don't have any videos up with hand pushing, but you should be able to push about 5-6mm of (input) feedstock per second and it will be extruded a little faster than in your last video, but it all depends on what speed you looking for, your hot-end should work upto about 60mm/sec by the looks of the video. Let me know how you get on with it.

  7. nothing to say really...except this is awesome! keep pushin the limits & tryin' new things!

  8. Thanks, I have some really cool things planned for the near future, keep tuned.



  9. Out of curiosity, why are you planning on losing the PEEK block?

  10. Hi Jeff - Yes that is the plan, I would ideally like to not use PEEK or PTFE in any extruder design.

    My limited experiments with a all (apart from the nozzle and heater block) Stainless steel hot-end is that it works but has limited extrusion speed. I may need a better filament drive mechanism or better cutting of the stainless parts.
    My other experiment used a SS insulator instead of the PEEK, but still needed a PTFE tube, this works well but I imagine the PTFE will fail at some point, it is very light-weight mind.

    I have some other ideas, but have limited tools to manufacture hot-ends well.

  11. I don't have clear why it cannot print slow - is it not enough lowering down the temperature?
    Also, I would like to try your thermistor solution, is there a ready-made table to use in the firmware?
    Thanks for sharing!

  12. Hi Syncra - it can print slower, but it's not the best design for very slow printing as it has a long heating zone, this is great to get lots of power into fast moving plastic material, so it prints really well fast at a high temperature - the plastic is moving so fast that it only just gets to the correct extrusion temperature even if its set to 225 Degrees C.

    But if you want print slower then due to the longer heating zone it makes the plastic ooze out of the end of the nozzle, even with a low (180 degrees C) temperature. That's why this is good for higher speed, which is exactly what I was trying to do with the design.

    I use this Thermistor table for the Axial part -
    #define NUMTEMPS_1 61
    const short temptable_1[NUMTEMPS_1][2] = {
    { 23 , 300 },
    { 25 , 295 },
    { 27 , 290 },
    { 28 , 285 },
    { 31 , 280 },
    { 33 , 275 },
    { 35 , 270 },
    { 38 , 265 },
    { 41 , 260 },
    { 44 , 255 },
    { 48 , 250 },
    { 52 , 245 },
    { 56 , 240 },
    { 61 , 235 },
    { 66 , 230 },
    { 71 , 225 },
    { 78 , 220 },
    { 84 , 215 },
    { 92 , 210 },
    { 100 , 205 },
    { 109 , 200 },
    { 120 , 195 },
    { 131 , 190 },
    { 143 , 185 },
    { 156 , 180 },
    { 171 , 175 },
    { 187 , 170 },
    { 205 , 165 },
    { 224 , 160 },
    { 245 , 155 },
    { 268 , 150 },
    { 293 , 145 },
    { 320 , 140 },
    { 348 , 135 },
    { 379 , 130 },
    { 411 , 125 },
    { 445 , 120 },
    { 480 , 115 },
    { 516 , 110 },
    { 553 , 105 },
    { 591 , 100 },
    { 628 , 95 },
    { 665 , 90 },
    { 702 , 85 },
    { 737 , 80 },
    { 770 , 75 },
    { 801 , 70 },
    { 830 , 65 },
    { 857 , 60 },
    { 881 , 55 },
    { 903 , 50 },
    { 922 , 45 },
    { 939 , 40 },
    { 954 , 35 },
    { 966 , 30 },
    { 977 , 25 },
    { 985 , 20 },
    { 993 , 15 },
    { 999 , 10 },
    { 1004 , 5 },
    { 1008 , 0 } //safety

  13. I was just asked a little more about the resistors and Aluminium bar used in this hot-end, so here is a bit more info for anyone else -

    My triangular hot-end was made with a standard length of Aluminium Bar stock, it's not the size I would use again in the future, but really size is not that critical, just make sure it's big enough to hold the two resistors and not oversized for the sake of it. I think it was 20mm x 20mm bar to start with that I cut two angles on. - Next time I make one I will not make it so high, so maybe a 15mm x 15mm Alu bar.

    It's quite important to get the correct type of resistors, you can get wire-wound silicon (usually grey) that look like they would work ok, but usually fail after a few months.
    the Type to get are usually Green with a shiny glass coating (enamel) - you will still need to wrap it in a layer of aluminium foil (I use alu tape) before pushing it into a tight hole in the heating block.

    I got my resistors from Farnell - http://uk.farnell.com/welwyn/w21-10r-ji/resistor-10r-5-3w/dp/9503625

    They are made by Welwyn - Datasheet - http://www.welwyn-tt.com/pdf/datasheet/W20.PDF

    Mendel-parts have a picture of the type of resistor you are looking for - http://www.mendel-parts.com/index.php/catalog/extruder/extruder-heat-resistor.html

    As a single heating resistor you can use 4R7 or 5R6 or 6R8 (2.5W or 3W)

    If you want to make a dual resistor hot-end I would recommend using 2 x 10R in parallel, but you could use 2 x 12R if you can't find any 10R (2.5w or 3W are both fine) don't use the 6W ones as they are much bigger.

    Good luck.

  14. online printing

    Wonderful to end up being visiting your website again, it is often months to do. We need this article to perform some of our my assignment succeeds. Many thanks for Giving.

  15. Nice Work.

    I have an idea in mind to get rid of the PEEK. At the university I had to design an experiment to make Lithium vapor. We ended up mounting the ss oven on ceramic balls. The oven ( a little cylinder 3cm diameter and about 5 cm long) had in its base 3 holes, d~3mm (spaced at 120 deg) and a threaded hole in the center. Then U place in ceramic balls of about 5 to 6 mm diameter in the 3 holes, then a heat shield out of 1mm metal sheet, with exactly the same holes, just the center hole is bigger to pass the threaded bolt through without touching, again 3 ceramic balls, add a second head shield, now again 3 ceramic balls and then the base plate, everything gets compressed by a SS screw. The screw head sits on top of the base plate, its end gets crewed into the center of the oven.

    Now for an extruder holder you want to have the fillament in the center, so use 3 SS screws on a 120 degree spacing 60 degrees relativ to the ceramic balls.

    This setup has the advantage that it will work up to very high temperatures (>600 C). If U stack all parts and drill the holes for the ceramic balls and the screws in this stack U get an easy precion allignment. The mount can take a lot of force and abuse, even at high temperatures.

    U have to take care to choose the size of holes and ceramic balls in such a way that they are not touching in the 1mm heat shield.

    The ceramic balls are a standard item U can buy of the shelf. But for lower temperatures U may even get away with using SS balls from an old bearing.

    Hope it helps.


  16. I don't know what to say. This is definitely one of the better blogs Ive read. You're so insightful, have so much real stuff to bring to the table. I hope that more people read this and get what I got from it: chills. Great job and great blog. I cant wait to read more.
    Screen printing

  17. NTC temperature sensor

    Thanks for the article, the performance of the machine looks very good.
    I think it definitely makes sense to use standard thermistors.


Note: only a member of this blog may post a comment.