Just Another RepRap

SO much progress since the last time I’ve written about it.

..and I’m actually at the cusp of printing.. So close I can literally smell it.. but I’m in a whole different category of problems now – problems getting things to print, not problems building the machine! So that’s actually one reason I want to post now – to cover all of the progress up to now so I don’t have to do it later. And there’s a lot to cover.. Huge media bonanza in this post – enjoy.

Ok, where to start.. Well, in my last real post I gave a quick preview of two youtube videos, so I’ll include them here. The first one shows a strange (new, replacement) problem I was having with the leadscrews:

Here’s a video from a month ago talking about the wiring that I’d started. Wow that seems so long ago:

Here are some pictures I took while wiring back then:

Then I had weeks worth of round-the-clock work on work stuff, so I couldn’t touch my RepRap for just about a month. But then this week I was on vacation with my family, and snuck in some nights and a few daytime hours of RepRap work. In a previous post nuttzy had suggested that I bring the RepRap out into the pool with me – I did almost that, by bringing it upstairs to work on.

More pictures..

One of the apps I have on my iPhone is a small app that makes it easy to take 3D pictures – you take a left and a right shot, line them up, and it’ll make either a parallel or crosseyed version for you (although sometimes it seems like it’s mixing those up, so I go with whatever works best). Here were two of those that I couldn’t resist taking.. It works great on an iPhone, not so good on a big screen TV. You want it to be small enough that you can cross your eyes and make the two images overlap then click, then the 3D effect takes place and you can easily make out depth. The downside to the first one is the fact that the front-most part is out of focus, which kills the whole 3D effect if you try to focus on it (which is a downside in any 3D pictures/movies).

I can’t get WordPress to not scale these images to be bigger, and scaling ruins the effect, so I’ll provide links to them.

3D-image-1

3D-image-2

The next thing I did was scratch an itch that had been driving me crazy ever since I got the Z-axis installed – I made feet for my Mendel. I also cut out the removable top MDF sheet to place on top of the bed, so that after a print I can just remove that and take the pieces off without reaching around inside the Mendel.

Another thing that had been bothering me was my power connection. Ultimately I’ll get around to obtaining a 12V power brick and connect that to an XLR connector for power, but I’ve been busy for the past month or so, so for now I have an ATX power supply. I’d been connecting two wires from the ATX power supply into the chocolate block on the bottom, but got sick of that, and figured what the hell – I’ll buy a 2nd XLR connector for when I eventually get an old laptop 12V brick, and free the current one up for me to wire it for use with the ATX power supply. However I didn’t want to wed myself to any particular power supply, since they often die, so I needed a molex connector.

I couldn’t bear to cut one off of any of the countless computers/accessories in my house, and I actually started looking at mouser.com for raw parts. Crazy. Instead, Y-adapters for 4-pin molex power connectors are like a buck. I had one in the other room, but again thought of the many times I use it. I went online, bought 3 more at around a buck a piece (they’ll arrive later in the week), and that freed me up to butcher that cable. Here’s the very nice result.

By the way, I LOVE the feel of an XLR connector – the locking feature is nice, and it’s just a very satisfying click. They’re more prevalent in Europe, from what I hear (at least for power) – I’ve never seen one used for power here in the US.

Next, I had the X-axis built up again to the point where it was moving (after having mostly fixed my Z-axis problems, and temporarily disabling my X-axis in the process). As for the details on this next video, I don’t think I was actually using the home feature at this point. I probably moved X, Y, and Z to some non-zero coordinate, then told them all to go to 0. I don’t think that at that point my Y opto flag was working, and Z didn’t even finish during the video, but I was still happy about it:

Next, on to the extruder. I decided to use a purchased 5mm nozzle, but make my own Mendel-sized barrel. In fact, I made it 4mm shorter so that the total length of the barrel and nozzle combined would be the same as the standard Mendel barrel. There was less space for the nichrome heater, but that’s good – I want to keep that heating zone short anyway.

Here I show why I’m building an entirely new heater barrel even though it was one of the first things I’d completed, and showing the old and new one side by side:

So I used that new barrel, purchased nozzle, and purchased PTFE insulator to build the heater assembly. I wrapped nichrome wire around it, wrapped that in Kapton tape, attached my thermistor, wrapped that in Kapton tape, and taped up the wires. Then I took it upstairs to test the thermistor:

I’ve noticed that since I’m so close to being done, the idea of printing SOON has raised the stakes of everything I do. So quite often I’m filming video before doing something, realizing that that might be the video filmed before a moment of tragedy or triumph. Here was one such pre-event video, about the simple task of cutting notches in my purchased PTFE insulator (I’ve made at least 9 on my own, but during my last makerbot.com purchase I threw in one so I’d have a perfectly centered, professionally machined one in a pinch).

And the follow-up result video:

(No, I still don’t have a lathe).

So now I was all ready for another task I’ve been dreading – gluing my PTFE insulator into my Mendel extruder piece. I’ve been dreading that because with the Makerbot I went through at least 5 or 6 PTFE insulators, and this one will be permanently welded to my pristine Mendel extruder piece (which I can’t print any more of, not having a working printer).

As with many of these worry-about-it tasks, there wasn’t much to worry about (at least until the PTFE insulator dies someday, but hopefully that will be after I print out a better extruder piece).

My JB-Weld experience:

JB Weld works by mixing two agents. The first is this black goop:

then you mix it with this (which I put right on top of it in this pic):

Here’s what it’s like when you mix it:

Here was the result before drying:

The next day:

So then, getting really close, I hit a huge roadblock. While others get to use new cool geared extruders like Wade’s Geared Nema17 Extruder or Adrian’s, I’m stuck with the RP parts I have, and that means I’m using the standard Mendel Extruder piece for now (at least until I can print a new one). That means the extruder shaft needs to directly push the plastic against a bearing, and a bare smooth extruder shaft won’t do anything but slip. I’d previously purchased this S10T08M012A0505 small gear wheel to put on the shaft, as suggested under “Adding a small gear wheel” here.

As soon as I went to put it on, though, I realized it wouldn’t fit inside the Mendel Extruder part:

Ignoring that problem for a while, I continued on with tasks I still had to tackle. I built a small stripboard with two 4-way connectors, as was suggested in the RepRap Wiring video – one set of four is for the extruder stepper motor, and the other is for the heater and thermistor connections. I built an 8-way ribbon cable with two 4-way connectors to mate with that. Here’s it all put together:

Then I got the final Z opto flag remade in tin and working great (somewhere before this I’d remade the X and Y opto flags in tin as well). The original Coke-can ones I’d made were just too flimsy – a slight breeze would knock them out of alignment. These new ones were made from a tin can of mandarin oranges (thanks Emily!). I also had to replace the X opto endstop board with a spare I’d made, because it stopped working for some reason (I’m using polarized connectors, so it wasn’t the polarity).

Finally I wanted to see my extruder actually extrude (which I originally envisioned I’d do waaaaay before assembling any of the Mendel itself!). I rested the extruder on the top of the frame, hooked it up, and did the test (pushing plastic in by hand). Despite what the first video says, I went with 170 degrees C (who knows what temperature it really was – I haven’t done any thermistor calibration or anything). I was using UltiMachine 4042D PLA.
Two part video:

Ok, so that left one remaining problem.. Getting the extruder motor to be able to grip the plastic to push it in by itself. Not being able to use the gear wheel, I decided I’d try cutting splines in the shaft, as described under the “Getting your stepper to put splines on its own shaft” section of the Pinch Wheel Variations page.

I built this rig to do the trick (from the above instructions):

Then here I describe and use the rig:

Did I mention I was close? 🙂

Laurie and I were up pretty late, as I told her I was so close that I needed help bringing stuff upstairs.. I brought the RepRap, and she brought the lazy susan with PLA on it. 🙂 As I untied the twist ties around the PLA, I remembered when I’d tied the PLA up previously – it was when I had to return the Makerbot; that was a sad moment, this was happy. We set our expectations low (rightly so), hoping just to have it extrude. We then had the X-axis start to fail (aaaggghh!!!), but Laurie figured it out (she’s a genius!). 🙂

And there ya go.. That’s all I have. I started typing this up Saturday night, ended up typing it through Sunday night. Actual prints probably not coming until next weekend sometime.

(and congratulations to anyone reading this far.. Did you actually watch all of those videos?)

Thanks go to Tom Royer for the continued use of his corded Dremel (vs my cordless one), and to Reed P. for his Dremel extension (which I’m returning tomorrow as I now have my own). While I’m at it, thanks to P. Newman at work for helping me with various mechanical issues that I couldn’t figure out (like how to remove a stubborn set screw which was driving me crazy, among other things). Thanks to Chris for giving me someone to talk to in person that’s reading as many of the online reprap blogs as me (your parts are coming soon I hope!). Thanks to Laurie, Emily, Alicia, Cara, and David for giving me a day or two of our vacation to do this latest push, and for being the best support in the world.

Oh, and again, thanks to you for reading this far. Seriously, who made it this far? 🙂

The problem I described in my previous post has been resolved! It turns out that I didn’t fry *any* of the stepper driver boards, and now I have two to spare. 🙂

The boards I thought were bad were suffering from not getting 12V or GND connections on both sides of the board. Contacts would be made somewhere, but not everywhere. Enough was connected that the green power LED would light up, but no power would make it to the stepper motors.

I desoldered the screw terminals on both fully-modified boards and fixed them. To make sure I was getting contact on both sides of the board I figured I’d play it safe – I took one of the metal contacts I’d ripped out of an RJ45 jack and soldered it in along with the screw connector. First, I soldered it on the top:

Then I bent it beneath and cut it to length:

Video:

I soldered the screw connector on after that and everything worked fine!

I mounted the boards on the Mendel again, and stopped for dinner.

One point worth mentioning for anyone following down this path.. The way I was determining if a board was working was by connecting it via ribbon cable to the motherboard, connecting power, and then looking for lit LEDs next to the ABCD stepper motor connectors (such as the incredibly bright red LED in the first video above). However, I later discovered that this method only works when connected to the X or Y stepper driver port on the motherboard. The Z stepper driver board won’t light up those LEDs, even though the board is fine. (I went through a quick discouraging moment where I thought I had a bad cable or Z port on the motherboard). Actually using the RepRap host to move the motor, however, works.

Once I realized this, I remembered reading that the Z motor is only enabled when it’s needed (since it’s used far less during a print). I think that even might be configurable in the firmware, if memory serves. That’s also probably the enable line that everyone was talking about before (either in the comments in the last post, or in various threads on the Mendel forum).

Now my biggest obstacle is the summer. 🙂 Trying to get our pool usable is eating up daylight, and once it’s open, that will eat up more daylight. Maybe I’ll be building by night. 🙂

p.s. Just because this post is so short, here’s a quick video showing the acceleration that RepRapHost (or actually probably the FiveG firmware) does during simple move commands:

Agh!!

Ok well the good news is, my 16 day hiatus is over, and I can try addressing my stepper driver board problem.

The bad news is, one of the two replacement boards that I purchased to address the two bad stepper boards is now bad too! I’m afraid to touch the last one.

I was grounding myself the whole time, so I don’t believe it was a static problem. I used a 15-watt soldering iron. I had the board grasped by the corner of the board, in a vice, and made sure it didn’t come into contact with any of the components or circuit lines. The page describing the necessary modifications is here.

First (before any modifications, and before apparently breaking one) I tested both replacement boards (unmodified) by connecting them (one at a time) via ribbon cable to the Y-axis port on my motherboard, which was connected to the computer. I connected one of the makerbot.com Kysan stepper motors (with the clip still on it, so the motor wiring was right), to the stepper driver controller board. I connected the molex 4-pin power connector from a PC power supply to the stepper driver board.

I turned it on (this worked for each of the two boards), saw not only the green power light but bright red and green lights next to the stepper motor connections. Those bright red lights will be on even if all you have connected is power and the motherboard connection, the stepper motor cables don’t need to be connected.

Here’s a picture I took later of what it looks like when it works, on an unmodified stepper driver board.. The ribbon cable is going to the motherboard, and on the left I’m holding the red and blue wire (red=12V, blue=gnd) coming from an ATX power supply’s yellow and black cables (multitester says it’s the right polarity and it’s 12V):

When the stepper motor is connected on a board that lights up like this, it can be controlled via the RepRap host software on the XYZ panel. I got motion from both boards before modifying the first of the two replacement boards (which is now not working).

I only made ONE modification, then immediately tested it, and it was already dead (the green power light goes on, but none of the lights near the stepper motor connector go on, and if you connect a stepper motor you can’t control it via RepRap host – it’s completely unresponsive). The only change I made was removing the 4-pin molex connector and soldering on a two-way screw connector in its place. Here is a picture of the first of the two replacement boards, apparently broken (note the green power light in the bottom right and its reflection on the RJ45 jack, which is the only lit LED).

I’m extremely irritated/concerned/disheartened.. When I modified the original three boards, one of those worked and two didn’t. Now one of my replacement boards doesn’t work already.

Has anyone else had this happen to them?

I’m open to suggestions. I’m afraid to modify the other replacement board, lest I destroy that one too. I’m seriously confused because I was extremely careful – maybe even 15watts is too hot? (I made sure not to use my 35watt soldering iron). I’ve checked the board and there are no obvious shorts etc.

Aarrgghh!!!