|
Cory Parsnipson posted:I have no idea how to source parts that aren’t electronic components. https://www.mcmaster.com/ - 16,591 results for "springs," time to go hog wild 
|
#
¿
Jan 20, 2021 00:00
|
|
|
|
| # ¿ May 15, 2024 03:18 |
|
|
Cory Parsnipson posted:Hmmm I see. I guess the key might be to add some artisanal twist to it like hand painted figures or some assembly. I hate to pile on but this is exactly what babyeatingpsychopath meant when they said all the niches were filled already. There's no way to make back cost on 3D printing with a single machine and a single person that hasn't already been fulfilled by ten thousand people on MyMiniFactory or Etsy or wherever. ~ artisinal objects ~ take up far more time and effort than can be made profitable even if you value your own time at zero. Just enjoy your printer and use it to make things that satisfy you! Cory Parsnipson posted:form a chunk of wires that route to the GPIO "spires" When you get to the PCB design part of this adventure it may help to know that the chunk of wires is called a "bus," and a well routed bus is a thing of beauty. Yours is looking great on paper! If I were you I'd think about starting to explore PCB design at this stage. Like you said cutting copper tape for this will be a bitch and a half and with all the overlapping traces on that bus (in either design candidate) it might not be practical to do in two dimensions. Even if you use very small magnet wires to do the jumps you're going to start taking up far more vertical real estate then you'll actually need in a final design and it'll be hella fragile to boot.
|
|
#
¿
Feb 5, 2021 22:59
|
|
|
Agreed, a port expander is exactly the way to go. I'm not sure how good the Pi's i2c bus is but you can surely scan it faster than button-mashing reaction time, and if you can receive GPIO interrupts (can the Pi do this?  ) then chips like this even remove the scanning requirement by having an interrupt on expanded pin state change.e: seriously though the MCP23017 is an excellent solution. It comes in DIP for breadboarding and QFN for when you get to designing a board thin enough to fit in the case e2: lol look what showed up on Hackaday today csammis fucked around with this message at 17:49 on Feb 9, 2021 |
|
#
¿
Feb 9, 2021 15:50
|
|
|
Just to be clear I linked the Hackaday article specifically for the use of the MPC23017, not because I think you need a full-on coprocessor or anything...that would be overkill for your application at this stage of development. A coprocessor hosting input functions is a valid solution, some embedded devices I'm familiar with farm out input streams to a peripheral microcontroller but that's for some fairly specialized situations:
|
|
#
¿
Feb 9, 2021 20:21
|
|
|
Cory Parsnipson posted:vvv Oh I see. Yeah this chip is quite powerful, that's nice. At 3 bux these capabilities seem like a steal. $1.32 at Digikey and Mouser Hell I should probably pick up a couple with my next order just to have them on hand.Also in case there was any confusion after my post about skipping the cardboard phase: I loving love this thread so much and please don't change. Every time I see the thread title it makes me smile and your posts are outstanding stuff.
|
|
#
¿
Feb 9, 2021 23:16
|
|
|
Cory Parsnipson posted:If I call it here, the next thing to do is to figure out how to use KiCAD and make some real PCB's I guess? I have to figure out how to break this down into manageable chunks... Yessssss join ussssssssss There are a ton of resources on leaning KiCad and quite a few of us in the Learning Electronics megathread use it. There will be a lot coming at you - schematics, make your own symbols, make your own footprints, so on and so on - but it is totally manageable to take things one step at a time from the vast stock library and build up knowledge as you go. Plenty of people have posted their schematics and PCBs for feedback in that thread so definitely don’t be shy about it! Cory Parsnipson posted:Wiggling the wire around causes the connection to be degraded. Wiggling which wire, the stranded wire tacked onto the end of the copper tape? If so I wonder if the solder joint is not good. Could be that the end of the tape was dirty and the solder didn’t wet onto it properly, or perhaps the iron’s heat melted the tape a bit so the connection isn’t the full width of the tape?
|
|
#
¿
Feb 24, 2021 07:01
|
|
|
It’s not difficult from the machine’s perspective, in fact a Prusa makes it almost criminally easy compared to other printers I’ve owned. Prototyping in a different plastic is fine but as has been mentioned ABS presents quite a few difficulties compared to PLA or PETG which might involve having to redesign a specific part around those difficulties. ABS is mainly used for high temp applications like being stuck inside a car in Arizona all summer or hanging out around the hot end of a 3D printer. I would recommend sticking with PLA and/or PETG, which I loving love for enclosures. And PLA is derived from corn. You may have heard that it’s biodegradable but that’s only partially true...it will break down under industrial composting conditions, high heat and such. You can’t just huck it into your backyard compost bin and wait for it to disappear. e: looks like you had an easier time with your X axis lead rods than I did. Those bastards did not want to go in. It took me half an hour with a mallet and a lot of muscle with rubberized gloves for traction. I earned those goddamn Haribo 
csammis fucked around with this message at 19:40 on Feb 25, 2021 |
|
#
¿
Feb 25, 2021 19:37
|
|
|
If it helps (it might not) I have had zero luck with the powder coated textured steel sheet. I can’t get anything to stick properly even if the first layer calibration comes out okay. As a contrast I’ve had no issues with the smooth sheet once the first layer cal is done. If anything sometimes PETG will stick too well and take tiny nicks out of the sheet coating, but I just tell myself those are consumables and keep a spare on hand if it ever gets so bad I have to switch out. I’m on something like 350 days of printing on one sheet though. If you don’t get it figured out in short order come to the 3D printer megathread and they’ll get you straightened out. e: just to clarify, you said your first layer calibration was newly out of whack when you went to print the test object; that’s likely not exactly what’s happening. The first layer cal was probably not sufficiently close to the plate so the test print didn’t adhere but you got lucky with the first layer cal, /or/ you accidentally didn’t save the cal properly - when printing starts hit the button and go to Live Adjust, make sure it’s at the offset you previously calibrated, /or/ the steel sheet is being a piece of crap. Do make sure it’s clean, skin oil is print adhesion poison. Warm water and soap without touching it will do nicely. csammis fucked around with this message at 00:36 on Mar 1, 2021 |
|
#
¿
Mar 1, 2021 00:27
|
|
|
Nice! Makes me want to take another crack at my powder coated sheet, that surface texture is really quite nice.
|
|
#
¿
Mar 2, 2021 01:58
|
|
|
I don't think the Z offset has anything to do with bridging. In my experience bridging issues are likely to be related to travel speed, part cooling, and/or printing temperature. The seam on the smokestack looks like the layer seam where the printer Z adjusts to the next layer. Some - most? all? - slicers default to putting the Z adjust at the same X-Y position on every layer which can form a seam-looking feature as the extruder stops pooping out plastic while the vertical travel happens. There are slicer options to randomizing the Z adjust location to eliminate this problem but I'm not sure what the Prusa presliced Benchy's Z-adjust is set to. That rear hull looks...not good. I've never made a Benchy on my Prusa (i'm printing one now out of sheer curiousity lol) but I'm surprised that the Prusa-supplied gcode isn't revealing those details. The "holes" in the surface on the port side might be layer seam or they might be underextrusion. I can't recommend enough going to the 3D printing megathread and posting your Benchy to get everything dialed in correctly.
|
|
#
¿
Mar 4, 2021 16:42
|
|
|
Cory Parsnipson posted:Just popped in the printing thread. Those guys are... intense. That is a true story but also they're smart and experienced as all hell. Sagebrush's advice to just print what you want and deal with issues later if they pop up is solid. I think once you've got the plate adhesion sorted then I'd agree with him.
|
|
#
¿
Mar 4, 2021 19:58
|
|
|
Look it's not important how I know this but I want to help a fella out...so go to your account settings and scroll down for a while and  In other smooth contour news, look up "Variable layer height" in Prusaslicer. It really helps with the top layers on curved surfaces like your Bulbasaur.
|
|
#
¿
Mar 11, 2021 16:30
|
|
|
That definitely could have been done for ease-of-repair reasons - I would bet that the ABXY buttons get broken due to mashing more frequently than D-pad buttons, and of course replacing a smaller board with fewer expensive components is much easier and cheaper. I still don't understand this though: quote:if I separate the design into 3 boards it won't be as dense so the total area will go up You're designing the boards, each board can be as dense as you want it to be  How would the total area go up versus a single board the length and width of your console but with a giant hole out of the middle? The fab will charge you by the squared outer dimensions of your edge cuts, not the absolute volume of FR4 left after all the milling is done.
|
|
#
¿
Mar 31, 2021 18:29
|
|
|
AGND isn't literally driven by the decoder, it's serving as a clean (i.e., decoupled from digital ground) reference for the analog signal. I think Adafruit calls it an analog "output" because it's going to the speaker which is the audio "output." The UDA1334 builtin symbol in KiCad has the VSSA pin with the Power Input type which is indicating that it's not being driven. You'll still want a global ground net for the analog ground just for convenience and clarity's sake. It's a nice way of indicating that the net is part of a power supply system distinct from a digitally driven domain. The builtin "GNDA" is handy for this but any of the ground symbols would do the job.
|
|
#
¿
Apr 3, 2021 22:21
|
|
|
|
| # ¿ May 15, 2024 03:18 |
|
|
Cory Parsnipson posted:PWR_FLAG ... loving stupid Yeah the PWR_FLAG stuff is annoying at first. It's the double edged sword of nets being generic and versatile in KiCad - it doesn't know "+5V" from "TOOTHBRUSH_INPUT_PORT" because it's just a label, so you need to have some way to mark the net with metadata which indicates that it's participating in power sourcing. If you have a symbol with Power Output like babyeatingpsychopath described then you don't need to have PWR_FLAGs sprinkled around. Some of the built in symbols like USB ports do this for you. If you do need PWR_FLAGs then I find it convenient to keep them tucked away in a dedicated corner of the schematic that is only concerned with power sourcing. That way I don't need to search through the whole drat thing to find them.
|
|
#
¿
Apr 6, 2021 16:21
|
|