Friday, February 18, 2011

First Extrudate!

If you look close you can see the shiny gray steel filings that where pushed out with the first bit of plastic. I had to re-ream the hole with one of my tiny drill bits while it was hot to get this plastic out.

Making the hot end

I've built a SAE version of the all steel hot end  by lampmaker. I decided to use 5/16-18 hardware instead of the more normal 1/4-20. For you metric guys, that means it's thick. 5/16" is basically 8mm, rather than the more normal 6mm for heater barrels. I did that so I could ream a taper into the bottom, and have the transition zone tapered. I used the taper pin reamer that I mentioned in a previous post, and it worked well enough.


I don't have a lathe, so I made the whole thing on a drill press. I put the barrel in the chuck and the drill bits in a small vise that I left free-floating on the table. The most nit-picky part was starting the hole - touching the spinning barrel with a drill bit such that the drill bit doesn't wobble, which indicates it's in the precise center.

Started it with a 3/32" bit, just because I had a lot of those and thus didn't care if I broke a few. I did break one.

Finished the through hole with an extra long 1/8" bit that I bought just for this. Thus the use of a different bit to start the hole - I really, really hate spending money. I was surpised that the hole on the far end came out exactly in the middle. Didn't expect that.

Reamed it as far in as the 00-sizes reamer would go, so the tip was just above where I expected to turn it down.

Turning it down with a file on a drill press proved to be - well - trying. I did manage to get some metal off of it, but it ain't pretty!

Heater Block

The heater block is 3/8" x 3/4" aluminum bar stock. I drilled a small hole for the resistor - I don't know what size, as it was a worn bit from a box of random bits I had laying around. Then drilled and taped for the heater barrel.

I'm really, really bad with taps and dies. Nothing I've tried manages to get them straight, so it sits a little crooked on the barrel. After I chopped the almost complete block off the end of the piece of bar stock, I drilled the small hole for the thermistor, the 3/32" bits turned out to be perfect.


The nut is a brass-plated steel 5/16 dome nut. I wanted brass but couldn't find any locally. I also used steel half-nuts, as I couldn't find brass. If I ever go into production with these I'll go ahead and order the brass ones in.

The dome is relatively thin formed steel, and I had a heck of a time drilling that tiny hole. The drill press would hold the nut, but so far up that I couldn't reach it holding that little tiny drill bit in my fingers. So I ended up doing it with a hand drill, and the nut screwed onto a spare heater barrel in the drill. Since the nut was screwed onto a barrel, I couldn't do the normal thing of drilling it from the inside, and had to rough up the tip to get the drill bit to stop slipping off. I got it done though!

Testing the heater

After getting the thermistor to work, I was testing my heater, trying to get stable tempurature regulation out of it. My heating resistor is held in with permatex "high-temp RTV silicone gasket maker" , and this is apparently the weak link. If I put too much power into it it starts to smoke before the thermistor registers the temperature increase. Also it gets soft and sticky at higher temps. That might just be because the tube of it that I have is really old though.

I'm powering it with 24V, and it's a 7.2Ohm (nominally 6.2) resistor, a PWM of about 20 is enough to get it hot, (takes a few minutes) and somewhere around 17 is stable at 220°C if the fan isn't blowing too directly on it. I'm currently using a pretty hefty 24V fan, which is way too much. I can't even get it hot when the fan is directly on it, and had to improvise a duct/windshield out of a piece of paper.

In this last picture you can see the back side of my board mountings, where I've held the arduino, PWM controller V1.1, and a piece of breadboard on with twist ties and heavy copper wire (it was within reach, ok?) The twist ties are on the arduino, and you can see the blue USB cable that goes into it. The two white cables are the temp sensor, and the Y endstop. It's sheilded 18/2 stranded wire, which is the only sheilded wire I have on hand.

To walk through the math,
P=E²/R  ->  P=24²/7.2  ->  P=80Watts at full power.
17/255 = 0.0666, or 6.6% of full power, or 5.33 Watts to maintain 220C.
20/255 = 0.078, or 7.8% of full power, or 6.27 Watts.

I picked a PWM setting of 20 just because I could see it was going to work. It could be higher. I tried full power (that smoked) and then I tried 64 (1/4 power) and that overshot, I'd set it to 50C and it'd hit 70C or so before settling down. Tried 10 next, and that wouldn't get warm enough.

This is all without insulation of any sort on the heater block. Well, I guess the piece of paper is insulation after a fashion.

I'm using the Fived_on_arduino firmware of course, as it's what I'm working on and I only have an atmega168. I've modified it to report the PWM settings (accumulated and number of times it was set) which is how I can tell how much power it's pushing.


I need a 4K7 (4.7K, whatever) resistor to use the second thermistor. I've got all sorts of other resistors, but not that one. I wonder if I can bum one off of my electronics instructor so I don't have to place another mouser order? I want to get a second thermistor working so that I can get some definite figures on all this stuff. How hot does the top of the heater barrel get? I don't know. Cold to the touch? How high up the heater barrel is it still hot enough to soften ABS? I don't know. The bottom washer is ever so slightly warm. But I will know precisely! Eventually.

I wonder if turning down the outside of the barrel is really needed at all. With the fan ducted to blow on the barrel just above the heater block that might be enough to keep the melt zone in the tapered part of the barrel. I'll need a second thermistor to test that though. And I'll have to make a second barrel that's not turned down.

It's been a *really* long journey to get here. I've been trying since at least 1/8/2009, which is when I registered on the reprap forums. That's over three years! I've build a McWire-inspired cartesian robot that works, and now I have a working hot end and a sortof working cold end. I'm so close!

Unfortunately, I got a job, start Tuesday and I have much work to do to get ready. I'll be living in my van during the week in craig, colorado. It's only a three week call, so hopefully I'll be able to get back on this again then. Might get some work in on weekends. I had planned to load the reprap into the back of my van and work on it there. Maybe put it on the front seat. But my inverter isn't strong enough to handle the laptop and the reprap, and there's not really room for all the tools and such that I'd need. It doesn't help that this thing runs on 24V instead of 12V! If it was a longer call I'd probably figure out a way, but I'm not sure it's worth it for three weeks.