Hot End Fail

I’ve had a bit of a break from my RepRap project recently, mainly caused by my disappointment when my first hot end lasted all of about 1 hour continuous printing before failing. It looks like the force of the filament being pushed down into the barrel was sufficient to cause the threads to slip on the PTFE interface. Probably partly due to heat making the PTFE more compliant. It does look like some plastic has made its way into the thread, but I’m not definite that it’s not just remnants of the PTFE tape. I’ve ordered the new design hot end components so looking forward to assembling that.

Meanwhile I’ve made a couple of adjustments. I’ve moved the USB connection from the top board to the lower board next to the power connection, I’ve added a switch (salvaged from a broken kettle) to the 12V power, and also added an LED strip to the underside of the upper board which nicely lights up the build area. Although I am now starting to think that it would be better to have all the electronics encased in a project box and have the rest of the machine stand alone.

I’ve also been working on a side project; using my android phone as a remote control for my Samsung TV using an Arduino connected via Bluetooth. It worked a lot easier than I thought and showed me how simple (and cheap) it could be to add wireless capability to any Arduino project. Documentation is a little sparse at the moment but hoping to add more soon, but if you’re interested the info can be found here:


http://sites.google.com/site/jsrsprojects/androidir

Extruder Stepper Bypass using Pololu A4983

I have decided to bypass the onboard extruder stepper motor control to use a separate Pololu stepper driver to run the extruder stepper motor. The reason for this stems from the original RepRap which was designed to use a DC motor for the extruder and not a stepper motor, therefore used some electronic trickery to control the speed. Using some clever programming they hacked the driver to control stepper motors.

However this is now what is causing my stepper motor to emit that high pitched noise. It is possible to bypass the information sent from the motherboard to a separate stepper controller, and leave the extruder board to just control the hot end. This method is documented on the RepRap wiki and suggested in a couple of forum posts I have seen.

I have made a carrier for the Pololu board from strip board using Tony Buser’s layout as a guide (http://www.thingiverse.com/thing:3761). I ended up changing it a bit as I wanted screw terminals for the power and motor connection and not headers. As the screw terminal spacings are wider than the header’s I had to use more short jumpers to get everything wired correctly. I have tested it and it seems to work very well. It is now much quieter and the motor seem to run a bit cooler.

From Mendel

From Mendel

First Almost Print

Just thought I do an update on my initial testing of the machine. I was quite pleased that I have managed to get everything wired up correctly and the extruder heated up. I have spent a while trying to level the bed, a job that was more tricky than I imagined. I am not really happy with the way the bed bolts to the y axis plate. On the Prusa design there is a nut above and below the y plate, I find this a bit annoying when adjusting the bed level – I have to keep adjusting both nuts, I can’t see why there can’t be just one nut below the plate, I might test it out at some point.

At the moment I am using RepRap host but I have also been testing with ReplicatorG to see if that will work. I can connect with ReplicatorG fine using the “experimental” Mendel machine profile. The control panel jog’s the axis (although sometimes the distances seem a bit random and sometimes it moves 2 axes instead of just one). The control panel also controls the temperature of the hot end fine however it does not turn the extruder stepper. When I try and print an object with ReplicatorG it starts ok, but then extruder starts to cool down, it doesn’t seem to control the extruder temperature. Not sure how to access the temp settings, I assume it might be part of the Skeinforge profile, but not sure how to get to that at the moment.

With the RepRap host things are working pretty well. Below you can see my first attempt at printing the 40mm test cube object. As you can see I aborted the print part way through, the object is less of a 40mm cube and more of a 40mm filter. The extrusion seems to be patchy, and thin even when it is extruding. I think this could be caused by a couple of things. First my “hobbed” bolt on the Wade’s extruder could be not gripping in certain areas – or the springs set too loose. Second I might not have set the extrusion length correctly (the steps per mm value) so not enough filament is being supplied for the required extrusion. Third the temperature could be too low causing a blockage as filament is being melted. See the picture and video below and let me know if you think you know which of these it is. At least the x and y dimensions seem to be correct!

Another thing is the extruder stepper is getting HOT (on a long print I think it could get damaged), I have tried trimming down the pot but to get the current up enough to turn the stepper. I have decided to use a Pololu for the extruder stepper in the hope that the stepper will run cooler (and quieter).

Settings:

Host: RepRap Host
Electronics: Techzone Gen 3
Temperature: 200°C
Material: Black PLA
Extrusion Steps/mm: 2.508

From Mendel

Start Up Procedure

I was having a little trouble when I hooked my extruder up. Although everything seemed to be wired correctly I found that I was getting a zero temperature reading. I checked the resistance across the thermister and it seemed to be right. Looking on the internet I found out the 0 reading is given when the extruder board is not connected. I recollected having this problem when I first tested the electronics, see post here, but couldn’t remember how I solved it. All I wrote was that it was something to do with the order I started everything up. A little searching on the internet and I found this post that was very helpful. Following the comments shown there I worked out a starting procedure that works every time.

Start with everything off and disconnected.

1. Start PC up wait for everything to boot up (or if it is already on, leave it on).

2. Power up 12V on Mendel (I really need to connect a switch up to make this easier).

3. Connect USB to Mendel (Could I connect a switch up for this as well, maybe on the +5V line?)

4. Press Reset on motherboard. Wait for motherboard to resume blink.

5. Start RepRap host, some errors are reported (0 temp measurement).

6. Close RepRap host.

7. Open RepRap host again, no errors reported this time, now everything works!

All electronics and hardware up and working. Success sounds like a high pitched whine (where did I put that Pololu board?!)

X Axis Belt Tension

I noticed I little lashing on my x-axis, it looked like it was being caused by the slack in the belt. The tension is set when first using the clamps, keeping the belt tight whilst tightening the clamps is a little fiddly. So I decided to try and improvise a tensioner. I looked around the house for torsion springs that might be suitable, I found a promising one in a cheap clamp (Tesco DIY little hand clamps). The un-tensioned legs of the spring were set to about 110° apart, and had long legs that I could bend around to hook onto the belt. The spring works really well in keeping tension in the belt, I might need to do some more reshaping or use tape to ensure that it stays in position in the long term. Here are some pics:

From Mendel

From Mendel

From Mendel

From Mendel

Wade’s Extruder Build

I have decided to abandon my previous Adrian’s extruder built from cast plastic parts, the oval gears were a nightmare to shape down to circles and the never seemed to mesh quite right. The printed plastic parts bought from nophead came with the parts for a Wade’s extruder (much better quality), so I thought I may as well build one.

For the Wade’s extruder build I needed to make a hobbed M8 bolt. Originally I was going to use the M3 tap set up I had seen before, but I came across this method that uses a Dremmel to cut slots in a threaded bolt to create spikes that will grip the filament. I thought this looked a little easier to do so I thought I would try this method first (plus less likely to break my one and only M3 tap). I found it really easy to cut the slots, though it did eat through a couple of cutting disks to do it. My cuts aren’t as tidy as the ones in the wiki but hopefully should do the job, I’ll let you know what the results are when I get my new stepper motor and I can test it out. I think it would have been better to have used masking tape so I could have marked evenly spaced cuts.

The Wade’s extruder build instructions recommend the use of springs to keep tension on the idler to help grip the filament. I’m not sure if there are some standard springs that will fit, or if people are making their own. I didn’t fancy much making the springs myself so I looked on ebay for something suitable. I found these springs that were cheap (25 for 99p), they measure 9mm in dia by 9mm long with a wire dia of 1mm. They fit fine around the M4 bolts and held in place by M5 large washers and M4 wing nuts. They seem to give plenty of force when tightened up, if one spring on each bolt is not sufficient I do have room to double up to increase the load. The M5 washers do clash a little but by rotating them I can get them to move past each other.

Getting the idler bearing on the short piece of M8 bar was not fun, if doing this yourself I recommend using a short bit of M8 threaded rod, the bearing will fit much easier.

See my previous post here on how I made the hot end.

Here are some pictures of the build.

From Mendel

From Mendel

From Mendel

Prusa Video

I've uploaded a video of my newly constructed Prusa to youtube. You should be able to see it below. I still have some calibration to do and finish off the Wade's extruder before I can start printing for real.

Prusa and Wade Firmware Changes for Gen 3

I’ve now got the three axes of my Prusa up and working with some temporary wiring, woo! Time for some calibration, now when I first set up my electronics I was doing it with a standard Sells Mendel in mind. To get the Prusa to work correctly I’ve got to make some changes to the configuration file of the firmware.

The first one is the z axis steps per mm, now I think that as the z axis is now directly driven rather than via a pulley and belt, this needs to be altered. The pitch of M8 rod is 1.25mm, so for every full turn of the stepper motor it should produce 1.25mm of travel. My steppers have a step angle of 1.8° (for a full step), now I think Gen 3 works in “half” steps, so for one full turn of the stepper I have (360/1.8)x2 = 400 half steps. Therefore the z-axis value should be 400/1.25 = 320 steps/mm.

The second one is the x and y axis, as I am using newer design pulleys (9 teeth rather than 10) I need to change the steps/mm. The belt pitch is 5mm, so 9x5 = 45 mm circumference. Therefore the steps/mm should be 400/45 = 8.889. (Interestingly the value given in the firmware is 8.9307 I wonder if it allows a little extra).

Thirdly is the extruder steps per mm. I am now going to be using a Wade’s extruder rather than Adrian’s. This has a gear ratio of 39/11 rather than Adrian’s which has a ratio of 59/11. I’ve had a go at calculating the steps/mm using the equation given on the wiki. For my set up, assuming a pinch wheel diameter of approx 5mm, nozzle of 0.5mm, and 3mm filament size, I make it: ((400x(39/11))/(pi x 5))x((0.5/3)² = 2.508 steps/mm.

Prusa progress

I have found the build much easier with nophead's proper printed parts. It only took me a couple of hours to get this far. I think I should have a carteasian bot up and running in the next week or two.

Switch to Prusa

I've decided to switch my build to a Prusa mendel after a long time getting frustrated drilling the cast parts. The cost of printed parts is now much less for a Prusa kit. Last week I purchased a printed set from nophead, can't be more pleased the quality is fantastic and the geometry is spot on. It has taken just a couple of hours to assemble most of the frame. Progressing well now, I'll upload some pictures soon.