Last modified: 2023-11-13 21:06:20
< 2023-11-12 2023-11-14 >Tidied the office.
Fixed the constant paper jamming problem in the printer by following this video: https://www.youtube.com/watch?v=rA54lzHAdss
Fixed the laptop (the issue was the "0" key would keep repeating once pressed) by tipping some isopropyl alcohol into the switch and hammering the key a few times. Great success. Time to install NixOS on this puppy.
Fitted the pipe in the alley with just one screw. I think this will be adequate, but can always fill the join with silicone if it proves not to be.
I am having trouble getting the laptop to boot the NixOS
installer, annoying. I don't know what is wrong, no error
messages, just doesn't boot. I'm going to try putting
the PopOS installer on the same USB stick and see if that boots.
Hmm, I can't even see the USB stick in dmesg.
I think it might be a bad USB stick, I'll try and find another. Great success, NixOS installer boot menu loaded up on first try.
I can't fix the chopping board, too difficult to get the edges square. I've chucked it in the wood pile, maybe can use it for some CNC parts.
So I'm finally trying out NixOS. First impression is that
"tap to click" doesn't work on the Touchpad. Hasn't this problem
been solved years ago in sensible distros? Apparently I need
to set some configuration.nix setting to add a kernel
parameter.
The other issue is that, OK, "the system is declarative"
because you specify everything in configuration.nix... but
what about the things you don't specify in configuration.nix?
Like I changed Gnome to dark mode using the "Settings" window,
but this obviously doesn't do anything to configuration.nix.
Same for setting up Ublock Origin and DuckDuckGo in Firefox.
It's the classic problem with supposed "declarative" systems like Ansible, Puppet, Chef, Docker. They work fine as long as you stay on the happy path, but they fall down as soon as you deviate.
Also I'm used to being able to drag a window up to the top in Gnome and it automatically maximising, but this doesn't seem to work, maybe that's a PopOS-specific thing.
I'm 3d-printing the plastic escapement model from yesterday. The white filament has all cracked where it goes into the printer, annoying. Why is white filament so brittle?
Also the printer seems to be under-extruding slightly. I've increased the flow rate from 95% to 99%, but not sure why it has changed from before.
If it turns out that the balance wheel isn't heavy enough, I can probably redesign it to have some holes around the perimeter that nuts or bolts or ball bearings can get pushed into.
Shafts/pins needed:
Also need to bring in drill bits to open up 1mm and 2mm pivot holes.
Well I was making the shafts and it was all going well, and using the 4th axis to cut pivots, and then I was trying to jog the machine around to zero the X axis ready to cut another pivot, and the left/right arrow keys stopped working, and then all of a sudden the left arrow key acted like it was stuck down, and it drove the tool against the work piece and snapped the tool off. Very frustrating.
Is this a Bluetooth problem? Or just the key not working very well? Need to watch out for this happening again.
Was thinking I could really do with a CNC lathe for parts like this, instead of pissing about with the 4th axis.
OK, great success, I've made all the shafts and pins, just waiting on the 3d printing.
I measured the under-sized pivot that I made the other day by accident, and it was about 0.20mm, and felt reasonably robust, so maybe I could go quite a bit smaller than 0.5mm for the watch pivots if necessary.
Let's look at fitting some of these pins/shafts into the parts that I have already printed.
Well I built it up, using the old mainspring as a balance spring:
It runs, but needs quite a lot of weight and keeps stopping, I think I need to clean out the pivot holes and maybe oil them. Or maybe look at the gear teeth. Not sure where the problem is exactly.
The back plate is too flexible, and the pallet fork is too flimsy.
The spring is obviously not centered on the shaft, that could do with tweaking.
But it looks like it works pretty well so far. By far the highest amplitude of any escapement model I've yet built.
The capacitor has arrived, it is even bigger than I expected:
Time to try it out...
Whoa! Actually a great success!
It is running backwards though, but I could just swap the coils over or something to solve that.
It has so much torque that when I try to stop it with my finger it tips the whole clock over, and also it has broken the ratchet by forcing it around backwards.
Yes! Great success.
Where am I going to mount this???
For now I'm going to take the ratchet off, and swap the coils over to hopefully make it run the other way.
No, not swap the coils, just swap orange for pink.
Good, that worked. Now if I switch to the 6v coil does it still work?
Yes! Seems good, and still really hard to stall. So I don't need the 9v transformer after all. And in fact I could probably get away with even less than 6v. Need to see how hot it gets, but this is probably acceptable.
I could consider trying to source a physically-smaller 100uF capacitor, or maybe try to make a feature of this one.