Heated Chamber 2

Heater arrived yesterday. I got it mounted reasonably well. Not pretty but it’s secure.

Also the power cord I’m going to turn into a power cord(…) for the heater element arrived.
I don’t want to intefere with the 110V wires going to the Predator power supply, so I got a IEC320 pass-through (C14 to C15) cable and I’m going to cut it up and tap the line and neutral wires for power to my heater. The predator casing has a internal power cord going from outside the wooden case to the predator’s original power socket. I’m going to have that pass-through this cable.

Wiring done and I tried out the heated chamber. After 20 minutes it didn’t get the chamber thermistor to read over 36 degrees. Not quite what i’m looking for. I think it needs a more powerful fan. The plastic sheet right nearest the heat element got really warm. Like too warm to touch. This indicates that its radiating too much heat. I need that heat to spread around the chamber. so mor airflow. More convection spreading the heat. I’m also switching to a bare thermistor that should more quickly respond to changing temperature.

Heated Chamber

So when I built the enclosure for my Predator 3d printer, I envisioned at some point adding a chamber heater. I made allowances in the design and construction for adding this additional element. Well, i’m finally moving forward with this project. I have a $10 heat gun, a $15 Solid-State-Relay, and some spare time.

I spoke of my planned chamber heater implementation on a forum for 3d printer enthusiasts, and was quickly warned away from using a heat gun. They are not rated for use this way, and another user had experienced a catastrophic failure doing exactly what I intended to do.

So I dropped some cash on a better solution. I should have all the parts I need by this weekend.

Even without all the parts I’ve so far managed to set up the SSR(Solid State Relay) and the temperature probe for the heated chamber.

Now I just need the heater itself and a fan to move the hot air around. The trick is not to move the air too much. I want mostly calm air around my print, as moving air will cool it too fast. Even the hot air is cooler than the extruded plastic. In an effort to duplicate the successful chamber heater of someone on the duet forums, I’m using a 220V fan being driven by 110V so it runs at half-speed(or therabouts).
Configuring the firmware was not difficult. RepRapFirmware version 3 is very well put together. I only needed to add a few lines:

M308 S2 P"ctemp" Y"thermistor" F0 T10000 B3950 				; Chamber temp thermistor'
M950 H2 C"exp.pa21" T2 ; Create chamber heater pinout
M141 H2 ; declare chamber heater on Heater2
M143 H2 S60 ; Set chamber max temp 60
M307 H2 B1

Perhaps I should have the chamber heater be heater 3 rather than 2, but its only a single-tool machine. Heater 0 is the bed, and heater 1 is the extruder. So heater 2 is the next one by count.

This is an awful lot of work to go to in order to be able to print ABS without warp. But I do so like the look of acetone-smoothed ABS. And this should help me make better parts which I can then smooth.
PETG is a fine material, and I’m largely migrated to it from ABS, but I have a large stockpile of ABS filament that I want to use up, and the heated chamber will let me do it without warp.


I have signed up for the MS150, which seems a little inaccurate, since the longest part of the ride is 102 miles. that’s the leg my team is doing. Improving put together a team for the MS150 this year. Now I’ve got 6 months to get myself in shape enough to ride that far. I think I’m about halfway there. I bet I could manage 50 miles today. Taking breaks, of course, every 15 miles or so, or whatever the interval is. I’ve managed 30 mile rides and still felt pretty good about it. another 10 or 20 doesn’t see so far-fetched. My bike needs its shifters adjusted however. I should get on that. I’ve been kinda sorta maintaining the tabata habit. My goal is to keep it up MWF, which is fairly attainable. Acing every week is really hard, given how cold it is in the morning. So given that I’m maintaining the HIIT training I think I’ve got the aerobic threshold needed to do this ride. It’s just the endurance that I’m really lacking, but that is what the next 6 months are for.

In other news, I’ve stopped the daily blogging. I’m going to try and focus my blog posts on a topic instead. My role models for blogging do not blog daily, but instead blog by a topic, even if that topic does not lend itself to a lengthy post.

Cypress Talk

Cypress is an automation framework built in javascript. I use this framework to build end-to-end automated tests. It’s the primary function of my job at the moment. I’m delivering a talk on the subject on friday the 11th. It’s not the perfect solution given our web-based application is build in knockout.js. I don’t know that there is a perfect solution.
The talk went well enough. I managed to get half an hour of content which isn’t too bad given i have about 2 weeks to put together the talk. I bet if i did a demo of the kitchen sink or real-world app i’d fill some more time. Oh well. At least it’s over with and i can focus on work again. No more looming presentation angst.

Dice Towers

I have been bitten by the design inspiration bug again. This time my goal is to make the best Dice Tower I possibly can. My Yin Yang tower was a decent first effort, but this one will be the epitome of randomization. I’ll design a tower that prints at the limits of my machine, to give me the maximum potential kinetic energy to convert into tumbling.

It starts with a cylinder. From that cylinder I subtract a spiral shape big enough for dice to fit in. I found 33mm to be about the right size so dice don’t get stuck. Then, to increase tumbling, I add steps. This is the most challenging part of the design. I need to repeat the pattern of the step up and around the spiral to the top of the cylinder. Also the step can’t be too big or else it might stop the dice as it goes down the spiral. Given that the tube is going to be 33mm in diameter, the spiral must be at least that or else it will intersect itself. I chose 36 for the spiral diameter because that seemed to work well. And the same value for the pitch of the spiral. Given these parameters, the angle of the slant of the spiral worked out to about 27 degrees. I didn’t do any math to determine that, I just imposed the spiral on a flat plane and measured it from horizontal. I tested this spiral with a smaller dice tower of only 2 rotations, and dice didn’t get stuck. The full tower will have 10 rotations around at its full height. which logically follows if the pitch is 39 and i’m making it 400mm tall.
Looks like the latest incarnation is going to come out to a 33 hour print.

Designing for 3d printing

I gave a talk last night about design considerations for 3d printing. It went well. I’ll try and capture the gist of it here in a blog post.

Succinctly, one must think like the printer, anticipating how it will build the model layer by layer. Design the part from the build surface(ground) up.

To start off with you need a well-calibrated printer. Design intent won’t get you around calibration issues. It can compensate for them somewhat, for example if you have issues with interior spaces coming out smaller than modelled, you can just design them to be bigger, but the best course is to correct the issue causing your interior space to be smaller and design the part how you actually want it to come out.

3dBenchy is a good way to check your printer’s calibration after you first set it up. It will reveal a number of problems if your printer has them.
The autodesk torture test is another good one to help get your printer’s calibration and find its limits. There are more out there, but these are the two that i’ve found to be the most useful.

So when going from concept to contraption, the process is like this.
You have an Idea, a concept. Then you have to take that concept and turn it into a 3d model somehow. I am blessed with the gift of being able to see shapes in my head and I have learned how to express those shapes as 3d geometry in CAD software.
So once you have your 3d geometry in some kind of software, you have to export it as an STL. That is a file type that represents the geometry as a series of triangles, and its what pretty much every slicer is designed to work with.
The slicer is going to take your model and produce Gcode, which are the machine-level instructions that the printer will follow in order to produce the part. The printer of course takes gcode and produces the part.

So in order to really design for 3d printing we need to go a little deeper into what the printer is doing. It’s extruding plastic. If you’ve ever used a hot glue gun, you’re familiar with the mechanism. In fact its possible to build a model out of hot glue if you were so inclined and sufficiently patient. It would be very tedious, but can be done. Plastic melts and comes out of the nozzle in a circual cross-section, based on the shape of the nozzle, which is a circle. It is then pressed into the layer beneath it, turning it into an oblong slot-shaped cross-section. A rectangle rounded on both ends. You can do the same thing with a hot glue gun by pressing the tip down as you squeeze the trigger. Now the printer extrudes this oblong in a shape in the XY plane based on the model it’s trying to print. THen it goes up a tiny bit, and does it again with a slightly different shape. Over and over it does this until the layers are built up into the shape of the model that was sliced.
The cross-section of these layers is going to resemble a stack of pancakes. As long as you’ve got enough overlap on the pancakes, the stack won’t fall over. But you can’t stack a pancake over empty space, it will just fall down. That’s what we call an unsupported overhang. By the same token, you can’t stack a pancake too far off the edge of the pancake beneath it, or the stack will fall over. Now when 3d printing, our panckes are pretty sticky, and rapidy solidify in the room temperature air. Even faster if you use a print-cooling fan. So you can have some amount of layer hanging over the edge of the one beneath it. My printer, for example, can handle about a 30 degree from horizontal overhang without too much trouble. So the perimeter layer is going to be 1/3 touching the layer beneath and it will still stick and be supported.
Knowing this, I can allow my designs to have up to a 30 degree overhang without worrying about the print coming out bad. So this is just one more thing I keep in mind when designing my models in CAD software.
And I would point out this only applies to overhangs, not bridges. Bridges are like overhangs but between two supported points. Printers have a much easier time printing bridges. By extruding over the open space and then stopping just before the second point, it uses surface tension to draw the extrusion taut before attaching it to the other side, and this way it can get a nearly flat layer over empty space. The torture test can tell you how far your printer is able to bridge over empty space before the weight of the hanging extrusion becomes too much and it stretches out instead of drawing taut before attaching to the other side. A cooling fan and lower print temperatures can extend this distance, because cooler extrusion is less stretchy and has more surface tension.
So what is good about 3d printing? Well the way layers are laid down means that in the XY plane 3d printed parts are very strong. I’ve made tons of very strong T-slot brackets by laying them out in the XY plane. Yes I could have bought brackets, but only at very common angles like 90 and 45. I’ve printed them at 15 and 75 degrees or 42.5. Angles you’re not going to find in stores, and paid pennies in filament compared to what it would cost to buy brackets. Are my plastic brackets as strong as the machined aluminium ones? no. But I have yet to find a use case where a printed bracket has failed due to being not strong enough.
Where is 3d printing weak? Z delamination is the biggest weakness of 3d printed parts. The layers sticking together only so well. If a part is going to fail, its probably going to fail by layers not sticking together.
Another thing 3d printing is pretty good at is threads. As long as you do them normal to the XY plane, you should be able to print really nice threads in your part. I’ve done this on numerous occasions, and saved myself $700 on a custom set of taps for really rare thread profile. 7/16 ACME threads. Mcmaster doesn’t carry taps in that size, and they’ve got like everything, but that was the quote to have a set of 7/16” ACME taps made. $700. ridiculous. Why would I need threads in this weird profile, well i’ll tell you. American science and surplus has the wonderful 12V motors with a gear reduction to 190 RPM and 25A stall current. They’re designed to adjust car seats, so they can shift a great deal of weight, since those seats are designed to move even when you’re sitting in them. Anyway, these motors have that weird thread profile on the output shaft. So I can 3d print an adapter bracket with matching threads. Or I can weld something to the output shaft but I had trouble making it centered on the axis. So the 3 times I did that i’ve got wibbly wobbly shafts. Fortunately there is enough slop in the acme nut mounts that it compensates for that slop.

internet woes

So today is the 18th of November. Every day since the 2nd(inclusive) my internet service has gone down. Usually only for a few hours in the morning, and it’s back in the afternoon. I’ve now had 6 technicians come to the house to try and deal with it. They replaced my modem(twice), re-run the fiber lines going from inside the house to outside(also twice), and now today the technician reran the line from my house to the pole. I hope this finally fixes the problem. Time will tell.

I think I’m going to take a new approach to blogging. I did it daily for 140 (week)days, now I think I’m going to do 1 post for the week. I think that will give me enough time to write something more substantial. And if i’m not satisfied with the results, I can always go back to daily. But it seemed like the days just ran together and I was out of things to say.

So far(thursday morning) my internet service has stayed operational. It was out when I woke up every other day this week, so it being on when I awoke this morning is a positive trend.

I was able to work a full day without interruption in service. Given the previous trend of losses of connectivity, I’m optimistic about the problem having been fixed.

In all of this, one might think that I have lost faith in my service provider, and would be therefore looking for a new one. However, I am not. I really have no complaints about their handling of this issue, except maybe how long it took to resolve. But I can understand why it took so long. They did the cheapest fix each time, which makes business sense. Replace the modem, relatively inexpensive, and of course by that point it was working again, either on its own or because the modem replacement gave it enough go to work again. The next time, they replaced the modem with a slightly newer model, and by that time it was working again. The next calls came after the internet had come back on its own, so they just tested the wires and re-ran the run from inside the house to a small box on the outside of the house. The next time they checked the runs to the box also. It wasn’t until I had a guy out in the morning, when the outage actually occurred that they replaced the whole run from the pole. This could be due to it being actually properly down when the tech arrived, or because they’d already replaced every other less-expensive component of my service. Process of elimination, starting with the easiest, cheapest components first. I would have done the same process if I were calling the shots. Also the technicians had a tendency to arrive early in the appointment window, so that’s pretty good. Actually several of them came before their window, they were that early.
Also, ATT sent out a retention specialist at one point, which happened to be when a repair technician was here, the third or fourth one who had come out. A nice enough fellow, geeked out over my chair, and also played D&D, in fact, DMing the same campaign i’m running with shannon’s family. He applied a promo credit to my bill, as an apology for the trouble I was having with my service, and then of course had a pitch about getting TV and phone with ATT. I was spared this when I said I had google Fi for cell phone service. “I can’t compete with that.” he told me, and that was that.

In other news, I aced my workout this week. 5/5. And no internet downage this morning.

I need to put together a presentation about designing parts for 3d printing. I have until monday next week. It’s about how to design things for 3d printing. I’ll go over how to optimise parts for being manufactured with FFF.


Today is friday. Day 35 of examining my themes. But there are only 34 of them. So today I will summarize what I learned by this experiment; going through all 34 of my Themes. I learned that I’ve got a big brain. I’m really good at thinking about things, strategically. Not so much on the relationship building, since those don’t appear until the back half. Influencing is also something I’m not super great at. Executing i’ve got all over the place. The problem with executing is that my thinking gets in the way. As I said yesterday, I keep thinking up awesome things to do, even in the middle of executing a project, and that adds scope, or might derail a project altogether. But it has its upsides. I think up some project with a skillset just narrowly out of my own, then I learn the new skill and get to it. That’s actually kind of a problem I have with learning new things. I can’t really focus on learning a new skill unless I have a practical application for it. ‘Form follows function’ has pretty much always been my guiding principle. The stuff I make needs to work, then I worry about it being pretty. I have a hard time 3d-printing tchotchkes or other sculpture-like items, because they serve no real function. I’ve done it on a few occasions, but its very rare.


Today’s theme is Achiever. I 34 out of 34. I wish this was higher on the list because I like getting things done. But it turns out i’m not that great at it. Among my weaknesses, this is probably the one i want to overcome the most. The feeling of accomplishment is great, and I want it more often. One of the biggest problems I have with that is enthusiasm. I keep asking myself ‘wouldnt it be cool if ?’ and then expanding the scope of whatever project I’m working on until it gets too big or takes too long and then never finish. I need to say no to myself. I need to fight to maintain the scope of my work within reasonable limits.


Today’s theme is Woo. People(not me) with this them love meeting new people and winning them over. They derive satisfaction from breaking the ice and making a connection with someone. Yea, this one is definitely not a strength of mine. I couldn’t woo my way out of a wet paper bag. It just doesn’t occur to me to try and sway others around. Once I figure out where somebody stands, I see them as a static object, a pillar of whatever idea they have. You can’t move a pillar, at least not easily, you just have to navigate around them. That is what I do.

Internet went out again today. This time a senior technician came out and replaced the wiring going from inside the house to the box outside on the back of the house, and said an engineer would come to check the wiring going from there to the hub. Engineer came by about an hour after he finished. Really fast technical support from AT&T.
Also, a ‘customer retention’ specialist came by. Friendly enough fellow, he geeked out over my chair and we chatted briefly about D&D, since he is also DMing the Princes of Apocalypse.
I must say the handling of this situation by AT&T leaves nothing to complain about, other than that it keeps happening. And nobody can seem to figure out why.