|
Re: An alternative to buying a bad kit for $300-500 October 24, 2016 10:34AM |
Quote
the_digital_dentist
I ran a multipart print last night to see see how well the glue works- it does. Soaking the plate in cold water for a few minutes released the parts. The print came out OK except for some overextrusion that I expected (I haven't calibrated it yet) and some slight shifting in the Y axis when the extruder moved from part to part (50 mm/sec vs 40 mm/sec print speed). Acceleration is down to 500 mm/s/s, so I'm going to play with the Y axis motor current a little and see if I can fix it that way before reducing acceleration/junction deviation further. I'll calibrate the extruder today, too.
Way to go
Looking fw to see the print result, keep us posted.
|
Re: An alternative to buying a bad kit for $300-500 October 24, 2016 12:03PM |
Quote
TheJones
Quote
Dancook
I don't plan to use the original extruders, but I honestly can't see a reason to keep them... I was trying to find a way to use them as bowens.
Why not just clamp a bowden fitting where the hotend used to go? Seems pretty straightforward to me.
Quote
the_digital_dentist
As you can tell, I like reusing the existing cables as much as possible.
Are you planing on using the stock hotend wires? They seem a little thin to me, and I'm contemplating replacing them. Thought, to be fair, they could be fine and I might just be paranoid from my recent experience with my heated bed wiring.
I'd like to try and use the original extruders as Bowens. I bought a couple of smaller/more user friendly Bowen type extruders for the time being, but if TDD figures out how to use the optical sensors, they may be valuable in the future for a way to pause the print in the event of a filament flow issue.
|
Re: An alternative to buying a bad kit for $300-500 October 24, 2016 01:13PM |
@Dancook - Check this out I think it is similar but uses different sensors.
[www.thingiverse.com]
Out of the box thinking is easier when you never fitted in the box to begin with.
[www.thingiverse.com]
Out of the box thinking is easier when you never fitted in the box to begin with.
|
Re: An alternative to buying a bad kit for $300-500 October 25, 2016 10:13AM |
I've been running a bunch of test prints to tune up one of the extruders and I can report that the Z=0 switch is working perfectly. Once I established the z=0 that got the prints to stick, it hasn't been necessary touch it again after at least 10 prints so far.
I'm printing Coex3D PLA at 215C (after a test run with a temperature sweep found that below 200C the layers separate relatively easily), 0.5 mm layers, 0.7 mm nozzle, extruder calibrated and delivering line width of 0.7 mm, here is what the surface seam looks like before tuning:
Here's what it looks like tuned up:
The seam has all but disappeared to the naked eye. Now to test at different layer thicknesses...
Regarding temperature... I am using Steinhart-Hart coefficients in smoothieware to describe the thermistor. I ran the calculations for the coefficients based on the temperature tables in the EPCOS data sheet for the 230k thermistor. I checked the temperature by removing the heater block from one of the extruders, heating it to 210C, and shoving a thermocouple into the block in place of filament. It read 200C. The thermistor has +/3% accuracy, and the thermocouple +/- 2.2C, and who knows how good the controller board's conversion of voltage to temperature works, so I'd say the temperature measurement is about as good as can be expected. That's why you have to run temperature tests with filament to see where it prints best.
Coex3D recommends 210C for colored PLA which matches up well with my measurement and temperature swept print results.
Edited 1 time(s). Last edit at 10/25/2016 10:21AM by the_digital_dentist.
Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
I'm printing Coex3D PLA at 215C (after a test run with a temperature sweep found that below 200C the layers separate relatively easily), 0.5 mm layers, 0.7 mm nozzle, extruder calibrated and delivering line width of 0.7 mm, here is what the surface seam looks like before tuning:
Here's what it looks like tuned up:
The seam has all but disappeared to the naked eye. Now to test at different layer thicknesses...
Regarding temperature... I am using Steinhart-Hart coefficients in smoothieware to describe the thermistor. I ran the calculations for the coefficients based on the temperature tables in the EPCOS data sheet for the 230k thermistor. I checked the temperature by removing the heater block from one of the extruders, heating it to 210C, and shoving a thermocouple into the block in place of filament. It read 200C. The thermistor has +/3% accuracy, and the thermocouple +/- 2.2C, and who knows how good the controller board's conversion of voltage to temperature works, so I'd say the temperature measurement is about as good as can be expected. That's why you have to run temperature tests with filament to see where it prints best.
Coex3D recommends 210C for colored PLA which matches up well with my measurement and temperature swept print results.
Edited 1 time(s). Last edit at 10/25/2016 10:21AM by the_digital_dentist.
Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
|
Re: An alternative to buying a bad kit for $300-500 October 25, 2016 03:00PM |
I'm tuning for 0.25mm layer height now and found some odd behavior, so I started at the beginning again. Everytime I think I have it figured out, I realize I don't.
Here's the test part, with retraction enabled. The start of each layer is where the bump is inside the red circle, which is the start of extrusion after a retraction. It prints the inner perimeter first, then the outer perimeter, then the infill, then a retraction and lift occur at the gold bump, and new layer starts at the blue bump. At the end outer perimeter the movement is straight over to the infill as you can see by following the thin red line. That transition tends to leave a bump where the transition occurs.
If I turn on "wipe" at the end of a line (the outer perimeter in this case) the path runs along that outer perimeter before crossing over to the infill. The problem is, all I did was turn it on with the setting at 0 mm. With the wipe at 0 mm, the path should be the same as if wipe is off. Hmmm.
The bump on the surface of the print occurs where the inner perimeter ends and transitions to the outer perimeter. As the outer perimeter line is finishing and gets back to that point, a bump is created on the print's surface. One would think that "coasting" would be the answer, but that creates all sorts of problems (exaggerated here by coasting for 5 mm to make the effect visible):
Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Here's the test part, with retraction enabled. The start of each layer is where the bump is inside the red circle, which is the start of extrusion after a retraction. It prints the inner perimeter first, then the outer perimeter, then the infill, then a retraction and lift occur at the gold bump, and new layer starts at the blue bump. At the end outer perimeter the movement is straight over to the infill as you can see by following the thin red line. That transition tends to leave a bump where the transition occurs.
If I turn on "wipe" at the end of a line (the outer perimeter in this case) the path runs along that outer perimeter before crossing over to the infill. The problem is, all I did was turn it on with the setting at 0 mm. With the wipe at 0 mm, the path should be the same as if wipe is off. Hmmm.
The bump on the surface of the print occurs where the inner perimeter ends and transitions to the outer perimeter. As the outer perimeter line is finishing and gets back to that point, a bump is created on the print's surface. One would think that "coasting" would be the answer, but that creates all sorts of problems (exaggerated here by coasting for 5 mm to make the effect visible):
Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
|
Re: An alternative to buying a bad kit for $300-500 October 27, 2016 09:26PM |
Alright, so as promised, I modified my extruder to have a fully supported filament path. This should help with printing flexibles. Check out the Tom's review of the Flexion extruder, it's the same idea.
I'll let you know how it works out once I get everything up and running. But at the very least, it don't see how it can hurt performance, and it gives me a way to make sure that both the hotends are at the same height, as long as I cut both tubes to the same length, and make sure they bottom out in the heatsink.
Anyways, here are some pictures.
First, I stripped out the aluminum blocks, because I didn't want to get metal chips all over my extruder's bearings and motor. Then I used the bolts as guide pins to align the pieces, and a clamp to hold them in place. With that set up, I had to drill out the holes of the filament path. I drilled them to just under 5mm for the top (to mount the bowden fittings), and just shy of 4mm in the bottom (for the Teflon support tube).
Then I tapped the holes. The top one is taped for the M5 fitting, and the bottom one is tapped with an M4 to give it "teeth" to grip the Teflon support tube.
Here is a picture on the part that will hold the tube in place when it is all bolted together.
Next, I trimmed the tube to a point, so it fits all the way up to the wheels. This will support the filament all the way down into the melt zone, kind of like a short bowden tube, which should keep the flexy stuff from kinking. I also trimmed the tube to a length that will let it bottom out in the heatsink once it is all sandwiched together. As I said before, if I trim both sides to the exact same lenght, and make sure they are both fully seated, then they should line up at the same nozzle height. At the very least, they will be close enough to shim with coke cans.
I used a piece of filament to line everything up, and pinched the lower guide tube in place. Then, with it all bolted back together, I put in the top bowden fitting for the upper filament guide tube.
And here it is, almost ready to go. I still need to assemble/wire the hotend, but I plan on doing that when I'm putting in all together.
Edited 1 time(s). Last edit at 10/27/2016 10:43PM by TheJones.
I'll let you know how it works out once I get everything up and running. But at the very least, it don't see how it can hurt performance, and it gives me a way to make sure that both the hotends are at the same height, as long as I cut both tubes to the same length, and make sure they bottom out in the heatsink.
Anyways, here are some pictures.
First, I stripped out the aluminum blocks, because I didn't want to get metal chips all over my extruder's bearings and motor. Then I used the bolts as guide pins to align the pieces, and a clamp to hold them in place. With that set up, I had to drill out the holes of the filament path. I drilled them to just under 5mm for the top (to mount the bowden fittings), and just shy of 4mm in the bottom (for the Teflon support tube).
Then I tapped the holes. The top one is taped for the M5 fitting, and the bottom one is tapped with an M4 to give it "teeth" to grip the Teflon support tube.
Here is a picture on the part that will hold the tube in place when it is all bolted together.
Next, I trimmed the tube to a point, so it fits all the way up to the wheels. This will support the filament all the way down into the melt zone, kind of like a short bowden tube, which should keep the flexy stuff from kinking. I also trimmed the tube to a length that will let it bottom out in the heatsink once it is all sandwiched together. As I said before, if I trim both sides to the exact same lenght, and make sure they are both fully seated, then they should line up at the same nozzle height. At the very least, they will be close enough to shim with coke cans.
I used a piece of filament to line everything up, and pinched the lower guide tube in place. Then, with it all bolted back together, I put in the top bowden fitting for the upper filament guide tube.
And here it is, almost ready to go. I still need to assemble/wire the hotend, but I plan on doing that when I'm putting in all together.
Edited 1 time(s). Last edit at 10/27/2016 10:43PM by TheJones.
|
Re: An alternative to buying a bad kit for $300-500 October 28, 2016 04:58PM |
Quote
the_digital_dentist
Looks great, but what is the purpose of the Bowden tube connector at the input to the extruder?
To connect a filament guide/strain relief tube from the spool to the extruder. That way, the filament doesn't resist the movement of the carriage, as it zips around.
Kind of like the the tubes that originally went to the filament cartridges, but more like this.
|
Re: An alternative to buying a bad kit for $300-500 October 28, 2016 08:01PM |
More progress this evening. I'm working on the wiring right now, which will likely be temporary until I get my upper enclosure built.
Here is my solution for the z limit switch. I re-purposed a piece of acrylic that I had stripped off earlier, and used the waste bin height adjuster as a set screw.
It seems pretty solid. It will probably stay like this until I replace that back plate, them I'll make something similar from some scrap aluminum.
Here is my solution for the z limit switch. I re-purposed a piece of acrylic that I had stripped off earlier, and used the waste bin height adjuster as a set screw.
It seems pretty solid. It will probably stay like this until I replace that back plate, them I'll make something similar from some scrap aluminum.
Sorry, only registered users may post in this forum.