Published: Tuesday, 10 March 2020 11:25
Written by Kyle G
I had seen on the HeatSync Slack that we were having some issues with the 3D printers. The next time I stopped at at Heat Sync, I saw the Ultimaker 2 just had a note:
So the next week I made a date with the UM2 to see what was going on. First thing was first: test print. As a first test for just about any FFF printer I like to slice the following gcode
*A.K.A. Corkscrew mode A.K.A. single perimeter A.K.A. spiraled contour mode
- 20mm Cube
- Vase Mode*
- 1 base layer
- 0 top layers
- ~.2 layer height
I like this kind of print since it's fast, and you can get a lot of information from it. In this case I wanted to see how bad the underextrusion was, and with the first print I got this:
First thing's first, check the basics. PLA was loaded, settings on the printer matched. Cooling fans were working correctly and running unload/reload showed that the extruder itself was (probably) working fine. But once the filament hit the hotend the extruder started having a hard time pushing, which meant that there was be a problem with the extruder or that there was a jam in the hotend. As the latter was far more likely, I started with some basic operations to try to unclog the hotend.
Just incase the printer had been loaded with ABS or another material that needs a higher temp, I heated up the extruder to 265 and extruded several hundred millimeters or plastic. When that didn't help much, I tried Ultimaker's Atomic Method (which is more or less a manual cold-pull). While this did pull out a bunch of dirt, The printer was still having extrusion problems.
At this point if I had a new nozzle I would have tried that - but if there are any spares, I don't know where they are. So I took apart the hotend
- Remove the 4 phillips screws connecting the fan shroud to the toolhead
- Remove the 4 hand screws to disconnect the heatsink/hotend from the toolhead
- Remove the 2 M2.5 bolts from the top plate that connects to the heatsink
- Remove the phillips screw what goes through the hotend into the bottom heatsink plate
- Unscrew the heater block from the coupler on the other side of the heatsink plate
Note: The heater cartridge and RTD are held in with friction and hope, so they may pop out on you.
Once I had the hotend free I used the hot air rework gun to heat the inside of the nozzle. A mix of cold pulls, tiny picks, a tiny drill and some percussive maintenance left me with a clean(er) nozzle.
I reassembled the hotend and re-ran my cube. Finally, there was a vast improvement over the original. There was still one spot that had some small underextrusion, but I saw a similar spot on the second print at the same spot, which makes me think there could be an overflow in the firmare or the extrusion just needs some fine tuning (rather than continued nozzle jamming). Though we should still get some new hotends to be safe.
The other problem with the print was that the first layer was awful. I'm guessing that when I reassembled the toolhead, I inadvertently changed the height.
Running the UM2's bed adjustment (on the LCD menu, Maintenance -> Build Plate) was very easy. At first I decided not to mess with the set screws, but after my first print showed up lopsided I went through the whole thing and was pretty happy with the results - for my final print I did the same cube, except wider (100mm in X/Y) and about twice height (40mm in Z).
So I'll leave it to run overday with a "real" test print. Hopefully it stays working for me to see what's going on with my nemesis The Choocher.
Published: Monday, 30 September 2019 23:05
Written by Kyle G
A couple of weeks ago I was at Heat Sync labs during open hours and a couple of guys were having problems with the CooCheer printer. I hadn’t printed with it before but I decided to take a look.
The main issues at the time were
- The coocheer has no way to load/unload filament from the LCD
- Connecting to the printer from my laptop would crash my laptop
- The wire connecting the control board to the bed’s heater pad was disconnected / melted.
Obviously #3 is a big problem, and had to be solved first. I’ve seen this kind of melting before on pluggable terminal blocks. Usually the problem is that the stranded wire is only contacting on a couple of strands, which can lead to the strands heating up and slowly melting the connector. Oxidation on the strands can also cause a similar problem. Sometimes however, the connector is just not rated correctly so too much power goes through and melts the connector. .
Just adding a ferrule to the wires and putting on a new connector would likely fix this problem, but I had am external MOSFET board meant for powering a bed heater, so I decided to get that hooked up.
-12 gauge silicon wire
Note - 14 gauge would probable have been a better pick as it met the power requirements and wold work better with the terminals I had.
-8-10 stud size Insulated Spade Terminals
-BIQU btmos v2
Note - there are newer models out there, this is just the one I had one hand.
-Misc ferrules and an extra pluggable terminal block found at Heat Sync
Here’s a simplified block diagram of the wiring. The 12v supply had empty terminals so I decided to wire directly from there instead of trying to piggyback of the 12v input on the main boards to help balance the power lines.
Ideally I would gotten the 5v control signal directly from the board, or reprogrammed an empty pin and controlled the BTMOS from that. Getting the control signal before the board’s built in MOSFET would involve an ugly solder job, and extra pins on the ANET board are in short supply. So while this might change in the future, I’ve decided to control the BTMOS directly from the FET output on the control board. There will be hardly any power draw, so the port melting again is not a concern.
Luckily the BTMOS accepts 12v in, so wiring directly to the FET output was not a problem. The BTMOS does not have markings for polarity on the input because it doesn’t matter, as there’s board compensates if the polarity is ‘wrong’.
The Coocher’s HBP has six wires coming off of it - 2 are form thermistor and the other 4 are for power. I haven’t yet checked if this is because it has two separate circuits or if the extra wires are just for balancing the power of one circuit on two sets of wire. Either way, I connected both positives together onto the positive output of the BTMOS and both negatives to the negative output on the BTMOS.
Following Jasper’s advice (and my own experience) I decided to put connectors on all the stranded-wire connections on the printer.
2 x 24v lines (board side)
2 x Hotend lines(controller side)
- 3 x AC input lines
- 2 x 24v lines (power supply side)
- 2 x Build Platform Heater (controller side)
Fixing What I Broke
During my various fixes I managed to crack one of the acrylic pieces holding the Z-Motor. Luckily there was some scrap acrylic in the lab and I cut a replacement part on the laser CNC. I used a part from this project and converted the STL to a DXF so I could cut it on the laser. Hopefully no one minds the new clear look.
After setting everything else I set the starting height and the bed level by eye. It’s not the best way to do it, but it’s good enough for a first print. I’d like to take another pass at leveling the bed and setting the starting height, possibly using some calibration prints. I’m a little spoiled with my M3-ID’s touch probe and built in calibration prints, but I might be able to adapt something for the Coocheer.
I had to leave before the first print was done, but the first layer was…ok. It was at least good enough that the first several layers looked good. I’ll post an update when I retrieve the first print.
I need to do some research on what version of firmware the board might have (an “M115” just responds “ANET anet3d.com”). But I’d like to load some new firmware onto the board, if only to add more options to the LCD (like load/unload filament)
I originally intended to replace the electronics on the printer. I have a RAMBo I can donate to the project, or a board with some nice TMC drivers would be great. However all of these options mean some significant changes to the machine. None are directly compatible with the A8’s LCD (at least without re-wiring ) and some boards have some drastically different configuration methods. So for now I’m going to leave the stock board on there.
I love Octoprint. Organizing, starting and monitoring prints is so much easier, and the plugin system makes it really easy to add notifications/ triggers and other cool features. I’m going to get a Pi 2 or 3 and then get Octoprint running in the near future.
This machine needs some documentation for Heat Sync users, so after I hunt down any existing docs, I’ll get to work on that.
I really prefer to set starting height with a firmware offset as opposed to a physical screw offset like in the Coocheer. The screw is vulnerable to vibration / changing the starting height. This will need several changes though (physical, software, firmware) or just a new Z-home indicator like a probe.
I really like the Nextion LCD so that octoprint can be controlled without needing to login from a computer.
Overall the machine could probably use a tune-up in terms of rods and belts, but that’s for another day.
Misc - Computer Shutdown
Earlier I mentioned that my computer would reboot after plugging into the A8’s board over USB. It turns out I had an old driver on my Mac for the CH340. I updated the driver and now I can communicate with the printer.