-
Event chaining.
02/14/2019 at 07:09 • 0 commentsIt was getting cold here. I fired up the space heater.
That blew the breaker, as I had let the printer cool down too much, from wanting to write the last log entry. That is my fault.
The excessive draw was caused by my desire to have a fast heating bed, which is hand wound, runs on 110V, and is of significant wattage. That is my fault.
The printer lost calibration as I had not shut down since the last calibration. The calibration settings are only currently saved on shutdown, and I wrote that. That is my fault.
That caused my last 0.4mm nozzle to crash into the bed. That flattened the end. Not having a spare or a drill to fix it, is my fault.
That caused my filament to strip, AND push the PTFE out of my mount. My PTFE doesn't use fittings to save the weight and so is just threaded in. That is my fault.
Everything above is sorted and I'm ready to go again.
Now, I forgot some stuff at the store, and have to go get it. That is my fault. :)
Looks like I'm the center of the universe here.. Back in an hour...
-
New petals.
02/14/2019 at 06:22 • 0 commentsEvolution.
-
Push it real good (or not, cause that sucks)
02/14/2019 at 05:47 • 0 commentsPulling with servo's is easy. Pushing requires planning, thicker stock, etc...
The new design requires upward force on the center to deploy the petals. That means I'm pushing at that point. That sucks.
When I was pulling, I was actually planning on using the wires feeding my Dotstar modules to also animate the rose.
I'm trying to think of how to invert this logic, perhaps making the center static and pulling on the outer rod. Or perhaps making the petal mount itself the pulling bit, and a static central shaft. I'm undecided. Time to decide..
-
Petals finalized.
02/14/2019 at 04:33 • 0 commentsThe curved petals meshed very nicely, but my opening method sucked.
I was going to push them open with a wedge from the inside. That resulted in all the petals stacking up and a 'linear rose'.
So I've changed the petal design to have an extended bit, inverted the mount so it can be thinner, and made the shape a little closer to an actual rose petal (as close as I can get and still have them fold/unfold).
12 to print.
-
Fooled ya..
02/14/2019 at 01:33 • 0 commentsWhile testing the folding, I had a visitor..
-
Perfect petals.
02/13/2019 at 22:10 • 0 commentsDollar Tree chandelier bulb, about 10 seconds warmup, 10 seconds on, shut off and smooth, wait 30 seconds, remove. Perfect petals.
Now to pick the right radius so they mesh and fold.
-
And... not TPU
02/13/2019 at 13:42 • 0 commentsMy TPU is not cooperating today. More precisely, I've blown away all my Cura settings and I don't want to take the time to dial it in again.
I think I have a better idea anyway.
Difficult to explain, but it gives me better mechanical advantage in that I'm now opening the rose with the servo instead of closing it. This also allows the stem leaves or a braiding band to be the spring which closes it. Ultimately simpler. Simple is good
Experimented with bending the petals.
Turned the heated bed up to 85C and let them sit on it, then pressed them against a light bulb.
Mostly worked, but not good enough. I need them to curve in two dimensions and I'm not getting them hot enough to smooth out. Perhaps the toaster oven... or turn the light bulb on. :)
-
Yay!
02/13/2019 at 09:40 • 0 commentsWas able to recover the original image, and undo my mess.
Printer is back up.
-
Bah.
02/13/2019 at 07:20 • 0 commentsPrinter won't boot. I didn't exactly shut it down correctly last time...
Removed SD card and fsck'd it. Put it back, but still won't boot.
Now I get to pull it apart and hook a serial monitor to it just to see what is going on. The printer cape prevents access to the serial lines.
If I even enabled serial..
Working on it.
<EDIT>
Yes, I enabled serial.
Yes, I corrupted my boot drive from not shutting down properly.
No, it is not recoverable. fsck's fine, but the SD card itself gives block read errors on boot.
Ugh.
I suppose I can't complain, as I've been shutting it down improperly for *years* at this point.
Rebuilding my printer OS from a freshly updated image, on a fresh SD card. Given I coded some stuff specifically for the old image, that could take a bit.
</EDIT>
<EDIT2>
Nope, I probably screwed it up more here.
eMMC somehow got the bootable flag set. So it was attempting to boot from the onboard flash instead, and given there isn't a boot disk any longer on the onboard flash, this was failing.
In the process of figuring that out, I seemingly hosed my existing image. Ah well. I still got the configs I need. Just 2.5 years of distro updates to deal with now. :)
Reset the bootable flag on the eMMC partition to no.
Moving on...
</EDIT2>
-
TPU bends...
02/13/2019 at 06:41 • 0 commentsI have TPU.. what in gods name do I need a hinge for?
Modeling the hinges as strips of TPU. Eliminating the return spring and modeling the slots in the base at an angle, so it naturally wants to open.
Had a brainstorm.. What if I used fishing line instead of TPU strands, and pulled on the outer edges? The petals may curl in when closing!
Going for it.
I can't actually model the holes for the fishing line as this would interrupt my single layer print causing an uneven pattern. I'll have to wing that part.
So... the petal.
The raised part in the center will be clear PLA, acting as a light pipe to the central clear PLA of the first layer.
Firing up the printer, and getting it loaded with TPU and PLA.