# Sprint U - Ursaring 10/1/2024 - 10/20/2024 ![[manorbear.png]] ## Summary [[#Pretty Pictures]] It's been about a year now since my last Lab Update. I'm transitioning away from the old page and am integrating both the logging and publication into Obsidian. It has the right tools for the job. My meta-systems for managing these projects and documenting them properly have slowly been growing more sophisticated over the last six months or so & Obsidian is the cornerstone of that change. I'm pretty happy with how things are coming together. This first Update covers 10/1-10/20 and is intended to get the new ball rolling. The Next Update will cover through 11/4. I've been on this for a while. ### Breadbox I built a new flavor of integrated breadboarding workstation. For the very few people who have been watching me since I first picked up a soldering iron, it should look familiar. Just significantly less stupid. I tried gluing silicone with non-specialty adhesives though and that basically didn't work at all, so still a little stupid. This is definitely the strongest version of it so far and it's really putting a twinkle in my eye. The new Breadbox is basically `Large Breadboard + Parts Storage + Helper Modules` all placed onto a stackup of a few layers: Aluminum Extrusion > Wooden Board > Silicone Electronics Mat > Gridfinity Layout ### Rat Rig The Rat Rig is my personal sisyphusean effort. Especially as commercial printers get more integrated and automated, supercharged with AI and such. But we're here now and the machine works. The last two weeks I printed the first batch of orbs for SBR, and then did some needed repairs, upgrades, maintenance, etc. before the Rat Rig scheduled it for me. - Rebuild Exhaust System - Replace Motor Cages - Dress all Cables - Calibrate Flow/EM to ~0.25% About two months ago I finally figured out how to handle gantry bowing due to heat warp. This problem alone explains SO many of the problems I've been having. `Large Format + Enclosed + High Temperature` is a recipe for huge amounts of thermal distortion. I've been getting around it with a loooong (90 minute) heat soak and compensation tape. Since I've done that I haven't failed a print and everything has worked more or less effortlessly. What remains is slicer optimization. I am winning at Printer. ### Steel Ball Run I built a mobile workstation using a cheap toolcart from Harbor Freight. It really serves the purpose. I also tuned the rat rig and slicer settings to these orbs, and optimized their design in subtle ways. We are back in production for SBR. I've only produced the first batch of 50 so far. Yield was 53/53/58 (delivered/shipped/assembled), which is AWESOME. Waiting for the go-ahead to start the next batch. I've improved the production process at every step of the way. It's simple, effective, and consistent now. I'm getting near perfect orbs faster and with less focus and bother - higher yield and less labor. I'm also not painting anymore lol. ### Meta *Meta as in working on the systems for producing organizing my documentation.* I've been spending a LOT of time thinking ahead, building out systems I need to execute on this Lab Vault concept in a way that is simple, efficient, useful for me, and intelligible to others. What I was doing previously was not working and the system basically grinded itself to a halt after about 2 years. After doing an autopsy of all my processes and basically taking the time to get my shit together I'm ready for a relaunch & rebrand. I do sincerely hope whoever reads this finds value in it. It's ultimately for me first, and other engineers second. Friends and family are along for the ride and I take all these into consideration at a structural level. ## Contents ### Breadbox - [[#Power Module Enclosure]] - [[#Arduino Mega GF Module]] - [[#Arduino Mega Print Tests]] - [[#Arduino Mega Holder Pics]] - [[#Construction Plan]] - [[#Silicone Adhesive Tests]] - [[#Measure]] - [[#Shopping]] - [[#Painting]] - [[#Cutting Aluminum Extrusions]] - [[#Backboard Painting]] - [[#Backboard Construction]] - [[#Final Assembly]] ### Steel Ball Run - [[#Orb Calibration]] - [[#Quality Check-in]] - [[#Definite To Do]] - [[#U-Line Boxes & Shipping]] - [[#Garage and Workstation]] - [[#Mobile Orb Station]] - [[#Cook Test Batches]] - [[#Notes from Batch 1]] - [[#Production Tracking & Sheet]] ### Rat Rig - [[#Flow Rate Tests]] - [[#Overview Maintenance, Repairs, Upgrades, Calibrations, Issues]] - [[#Maintenance Routines]] - [[#Repairs]] - [[#Upgrades]] - [[#Future Calibrations]] - [[#To Do List]] - [[#Misc Notes]] - [[#Special Topics]] - [[#Shopping List]] - [[#Exhaust System]] - [[#Cable Dressing]] - [[#New To Do List]] ### Meta - [[#Meta]] - [[#Datasheet Thoughts]] - [[#Sprint & Log Structure - Separate Notes]] - [[#Header Logic]] - [[#Ex Formatting Tests]] - [[#GitHub Pages]] - [[#New Project Ideas & Proper Zettelkasten]] - [[#Personal Electronics Component Database]] --- ## Project Logs ### Breadbox #### Power Module Enclosure The last one I printed didn't work right. It's still wobbling and poking out. Unfortunately I have to go up to like 8u height so it's going to be like a pillar of power. 1u vertical is 6mm. 1u lateral is 42mm. This power module is about 50mm tall overall. I also need to model a snap plate ##### Faceplate Print Test 1 71.25mm x 39.00mm Need to expand the 39.00 -> 39.50mm, since the frame catches ever so slightly on the Green Terminal Blocks Need to expand the 71.25 -> 72.00mm, since the frame can't slide on very easily, and then it absolutely cannot pass the little frame snapper things. Need to add a 1/2" x 2mm bite out of the inner, center, shortsides Making it 1/2" + 1mm ![[Pasted image 20241003065312.png]] ##### Faceplate Print Test 2 ![[Pasted image 20241003072046.png]] That extra 1mm isn't necessary after all, and creates play. The tabs terminate so close to the face, the print would be really thin. I'm considering just making a long full shaft that wraps around the tabs instead of trying to snap into them. If I do that though, I'll end up obscuring the Terminal Labels. They really want me to make this thing like 1mm thick... I don't want to. Also, the shortside face of the module is longer than the frame's face, which makes it look stupid. I can make the trim even for both sides, but then it'll mushroom/pillow at the top. It's high enough off the ground that that shouldn't matter. It's only casting a shadow onto the arduino, after all. Let's try it Changes: - Increase top fillet to 3mm - Reduced tab slot to 1/2"+0mm - Increased border trim to 6mm on all sides (51.50mm x 84.00mm) ##### Faceplate Print Test 3 ![[Pasted image 20241003075722.png]] ![[Pasted image 20241003075750.png]] There's too much lip on the top and bottom. It turns out the backside of the face has uneven thickness. It's a simple fix though. New Border: (48mm x 84mm) I think it's ready to print the whole thing now. ![[Pasted image 20241003081601.png]] Came out great, fits great. We did it! #### Arduino Mega GF Module ![[Pasted image 20241003165621.png]] A Gridfinity Module to hold an Arduino Mega. It uses a 2x3 tray as a base. I used this image to get the placement and dimensions of the holes. In this case, I just imported this image as a canvas into Fusion and traced the outline and circles. ##### Arduino Mega Dimensions ![[Arduino-Uno-Mega-Dimensions-1.webp]] #### Arduino Mega Print Tests ##### Arduino Mega Mount Print Test 1,2,3 - Tried something funky like this with splines so I'm not printing an entire plate. - 3mm circles don't work. The holes on the arduino are 3.25mm. - It seems like 2.5mm diameter shafts will work better. The height seems fine. ##### Arduino Mega Mount Print Test 4 - Reducing the mount peg diameter to 2.5mm worked great. But since they're only 3mm tall, the backside solder and pins are actually blocking a flush connection. I might need to design the mount plate to route around them, which is something I hadn't considered before. - A quick look at the layout of all these pins means the routing would be pretty awkward looking. I plan on having this basically embedded into the floor of the Gridfinity Piece like the Transistor checker either way, so I'll just make the pegs taller for now. - *Little pegs poking out ready to hold an Arduino* - The taller pegs work fantastic. But since I have extra depth, I can see now that 2.5mm is just a bit too small. - I'll try 2.8mm next. I have a feeling 2.75mm won't be properly snug. ##### Arduino Mega Mount Print Test 5 - 2.8 did not do it. Checking the 3.0 from Print1 I can see it's loose as well. - Gonna full send and print 3.25. Let's see what happens - 3.25 is very snug. Hard to get in and Hard to get out. It’s ready for the full design now. - Came out great! #### Arduino Mega Holder Pics ![[Pasted image 20241003071521.png]] ![[Pasted image 20241003071500.png]] #### Construction Plan ##### Objectives Drill a hole into the MDF, and mount the aluminum extrusion to it Cut up the old silicone mat, and use the adhesive to glue the silicone to the MDF ##### Notes - The silver extrusions I had laying around (from the Kobra Enclosure) are actually 1", not 20mm. So I just drill some holes 1/2" in from the side. - M5x20mm seems to fit perfectly fine with the tnuts I have laying around. - 4.5mm Drill bit to make the hole. - I cut up the old silicone mat into about 9 pieces of 2" strips - The adhesive is like, super toxic, so latex gloves, ventilation, etc. - It recommends sanding the surfaces first, then wiping them free of dust. It also suggests wiping the surface with a damp rag before applying. - After applying the adhesive, I used a plastic scraper to smear it around. It's pretty compliant. It's about the same consistency as my silicone sealer. - I laid down the silicone pieces and then clamped it to the wood. - It says the cure time is 12-24 hours. I threw it in the closet and we'll see what it looks like tomorrow. Also need to check on the Silicone Bonding. It cured overnight. I believe all of them were wet surfaces. Three pieces total to check. ![[Pasted image 20241009090541.png]] #### Silicone Adhesive Tests ##### Silicone Adhesive Test 1 I peeled this one off and restuck it after about 10 minutes yesterday. It seems like it still bonded fairly well. This one peeled right off. There's definitely nontrivial resistance, but it's not something I would trust mechanically. I think it's appropriate for the job of just, holding a silicone mat down. The adhesive bonds to itself and to the MDF better than the MDF bonds to itself, so I can basically skin it off. The silicone test piece itself is completely unscathed. This is probably for the best, so the mats can be re-used How do you clean silicone mats? ![[Pasted image 20241009090558.png]] ##### Silicone Adhesive Test 2 This one I used more adhesive, smeared it around carefully, applied carefully, and then clamped down and left undisturbed for ~22 hours. At first glance it seems much more bonded. Same story as Test 1, peels right off. ![[IMG_8474.jpg]] ##### Silicone Adhesive Test 3 For this one I glued it, screwed it, then clamped it down. Screwing it down causes the silicone to bow upwards, which I consider a big problem and is antagonistic to the flattening/clamping I need out of the adhesive. I think it might be best to get a custom size circular hole punch and literally just skip the silicone layer for the bolts. Either that or countersink them properly. Both are kind of a pain. I've been leaning towards a principle of silicon-mat-reusability. This one peeled off effortlessly as well ![[IMG_8475.jpg]] ##### Conclusions The adhesive I purchased is no good for binding silicone. I think that's okay though, since as I've put more effort into this, I don't think I want to permanently bind the silicone mat to the board. Especially if I slightly misalign it or something I'd be mad. I'll still use it to bind the silicone since it's by far the strongest adhesive I've tried, but the results are underwhelming. I need to decide if I'm going to holepunch the silicone or if I'm going to countersink the bolts. Are there even M5 countersink bolts? I think I'd need custom router bit and custom bolts for that. #### Measure The MDF is 13mm thick The M5 Cap Head is 5mm thick The M5 Cap Head Diameter is 8.44mm #### Shopping #shopping-list - Countersunk Screws & Drillbits ![[Pasted image 20241010172000.png]] #### Painting I'm not sure how I want to go about painting it. My first thought was to use the truck bed paint I used on the desk legs, but I have a feeling that will be too rough and won't be very nice. My second thought was to use the paint bucket I used on the BJJ sign. I don't think I've tested that on MDF before though, so I don't know how the finish will go. I believe I also have some of the black enamel spray paint leftover from Orbs. So I can probably just do test pieces with each of those paints and make a decision from there. But to do that I'll have to move the car, clean up the garage and lay down a tarp for painting. I have to clean the garage to get the orbs in order either way though, so it's fine I suppose. #### Cutting Aluminum Extrusions I bought the scroll saw or whatever for the BJJ sign last year, and I've read that you can use most woodworking tools for cutting aluminum as well since it's fairly soft. Or something like that. I don't really want to hit it with a hacksaw. Also, I need a new hacksaw. I could always ask Bill if he has one, or if he knows someone in the neighborhood. Worst case I can consult OlyMEGA or Lacey Makerspace or Steven #### Backboard Painting I finally took a few minutes to sand the edges, prime, and then paint the board with an extra can of black enamel I had laying around. Hoping it looks decent when it's finished. ![[Pasted image 20241017205336.png]] Gonna go ahead and send it and buy the Countersunk Screws and Drillbits.... After I set up the shopping list meta-system too. It's like a reward. #### Lamp-Mounted Camera I've been playing with the idea of mounting a camera to the lamp for a while now. I rigged up a demo by taping my webcam to it, but it sucks! I'd probably need to get myself a dedicated camera for it. ![[Pasted image 20241019061749.png]] The main problem is my phone is far too heavy for this lamp to support. So I either have to come up with a separate jig to hold the phone, or I find some high-res, lightweight camera. It doesn't need to be mobile or battery powered or anything, since it'll be installed into the desk essentially. This is just a webcam taped to a lamp. The quality is so bad though. Logitech C922 I'm pretty sure. ![[Pasted image 20241019062713.png]] #### Backboard Construction The paintjob on this piece of wood is not great. The enamel isn't fully cured and it's coming off a bit onto the desk. I can clean it with Acetone later. It's not really meant to be the once and for all. More of a prototype so I can get some organic work hours in on it and redesign it properly from there. I need to cut the aluminum sooner than later. I'm pretty sure I'll have to remove everything from the silicone mat in order to glue it on properly as well. It does look decent so far though. ![[Pasted image 20241019063818.png]] ##### Cutting Aluminum Extrusion As much as I'd like to go buy a Miter Saw, I don't really have the budget to start setting up a workshop right now. A proper workbench, toolchest, drill press, miter saw, and shop vac are first on my list in the future. For now I'll just throw this together using my hacksaw Aluminum Extrusions are 1" square The actual dimensions of the wooden board are 22" x 17.5" Almost exactly 22", just barely under 17.5" I'm doing long-side out, so I need Cut List: - 2x 22" - 2x 15.5" Look right after measuring as well, so time to cut. Cuts went well, then I did the drill hits and used the router to carve out room for the caphead screw, since it they would've partially overlapped with the Silicone. This doesn't need to be super polished because the work I'm doing isn't super polished. I'll build another one in the future appropriate to the precision of my designs. No brackets or anything holding the aluminum frame together - just positioned against the wooden board. #### Final Assembly It took a bit longer than expected, but it's all put together now. ![[Pasted image 20241019122056.png]] The board is painted, drilled, countersunk, and the aluminum is cut and installed. Then I used the caulk that only kind of works to attach the silicone mat, and ensured that no corners hang off when flipped upside down Then I reinstalled the gridfinity base - stretched it into place using a screwdriver. I also dug up some adhesive foam strips I had laying around and attached them to the bottom so the edges of the aluminum can't scratch and scuff the table below. ![[Pasted image 20241019122112.png]] It's a bit thicker overall than I was expecting, but it's very sturdy, and can slide smoothly but not accidentally. I'm happy with how it came out. ![[Pasted image 20241019122134.png]] ### Steel Ball Run #### Orb Calibration Printing 2x half orbs of the badge reels and tweaking settings until quality/speed is perfect. ##### Run A - External Perimeters: 60mm/s & 3000mm/s2 - Top Layers: 8 - Bottom Layers 4 - Perimeters: 4 - Infill: 15% - Variable Layer Height: 0.60, Smoothed - First Layer: 0.20mm ###### Outside: They look good, but it was too slow. Going to speed it up dramatically, since we're testing. ![[Pasted image 20241007124209.png]] ###### Inside: I'm seeing a small metal ring at the top where the ring comes out. It's slight thickness is crumpling against the extremely tight circle. There's some rotational shifting causing that metal ring to be misaligned. Further, the steel string does not exit that ring centered. ![[Pasted image 20241007124227.png]] ###### Solution: I added a domed top so the metal top-ring has room to sit, and smooth, curved guiding walls so the string isn't brushing against a harsh edge as it is led to the channel. I also sunk the inner floors another 0.5mm each. ![[Pasted image 20241007124358.png]] ![[Pasted image 20241007124512.png]] ##### Run B ![[Pasted image 20241007125240.png]] ###### Changes: - Combine Infill Every: 1 -> 2 Layers - External Perimeters: 180mm/s, Default Accel (20k) - Top Layers: 8 -> 4 New Variable Layer Height Profile: ![[Pasted image 20241007124800.png]] ###### Notes: 4 top layers probably isn't enough. I assume the top shell will be too fragile. I'll need to run a Crush Test between A and B (8 vs. 4) ETA is 1h44m for this one. Very fast. Thick 0.3mm layers like this may not smooth properly. ###### Review: Climbing layers seem good, very consistent and little ghosting/ringing Top Circular layers are also good. I think it's worth seeing what 0.05mm layers looks like, but 0.1mm seems pretty sufficient. The new top-ring void/guide looks pretty good as well. The badges are honestly kind of hard to get in. I may need to make it a bit less snug. I'm not sure yet. Once it's in there, it's in there for good. The extra 0.5mm means that it's not perfectly dead center between both orbs, but will be slightly offset towards the piece the badge is originally inserted into. The top ring has plenty of room now and the width of the guide funnel means the asymmetrical badge placement doesn't matter. Excellent seam once combined as well. Assuming this cooks well, I can keep the speed. If it crushes poorly, I increase the top layers It's worth making slightly less snug as well, at least for a test. ![[Pasted image 20241007172432.png]] ![[Pasted image 20241007172352.png]] ![[Pasted image 20241007172359.png]] ![[Pasted image 20241007172408.png]] ![[Pasted image 20241007172445.png]] ##### Run C ###### Changes - Top layers: 4 -> 8 - Variable Layer Height Profile: 0.3mm, then 0.1mm - Design: +0.25mm Offset on circle, and on Rectangular tab slot ![[Pasted image 20241007173211.png]] ###### Review: Badge doesn't fit. Trying +0.1mm instead ##### Run D ###### Changes: - Design: +0.1mm Offset on circle, and on Rectangular tab slot ###### Review: It fits very easily and then is very difficult to remove. I think this might be the one. I'm still under 2h as well. Everything snaps together effortlessly and holds firmly. It functions without issue assembled without any cooking or processing. It stands on its own. ![[Pasted image 20241008182950.png]] #### Quality Check-in Rat Rig is going to run out of filament today. Starting an 8h print for some more orbs. It'll be good to see how the spool transfer goes down. If it kills prints that's going to be a problem and I'll need to fix it. Launched print around 59.8C on Toolboard, Bed Mesh was 0.393mm. Sub-400 is great! Rat Rig is still having that power-throttling issue. I'm not sure if I should replace the wires or if I need to dedicate a separate 5V supply. If I go in to add that supply I'm going to have to overhaul ALL the wiring. So fingers crossed it's not an issue. ##### Major Print Test: 16 Orbs Started 2024-10-10 - 10:04 I weighed the spool and it has about 1.5kg of filament left, so I should be able to print a 970g print with 16 orb halves on it today. This is the most I've attempted in one go so far. If the Pi Undervolts and Crashes I'll dig into it. #### Definite To Do: - Need to clean garage - Need to set up print-farm books #### U-Line Boxes & Shipping It's premature to spend $120 on 50 boxes when there's only one batch order out right now, so I'll just reuse the filament spool boxes to send this first batch out. #### Garage and Workstation I need to clean the garage and set up some kind of workstation for Orb Assembly, the current AVC, etc. Junking is definitely not in my budget rn. The 27 gallon bins at Home Depot are $10 each, and I bought four of them. I currently have something like this: Some change I might make: - Clamp mechanism to hold orbs together so I don't have to? - Put the screw hook further back, use a taller screw/nail/hook - Add some sort of wall to clearly separate and protect the Plastic Bits on the hook from acetone exposure - Add some sort of retaining wall that the string can thread through to hold the orb and prevent it from snapping back - More Acetone Syringes, Bigger Acetone Syringes Some Processes: - Prep the badges - Take out of plastic bag - Peel off protective plastic - Snap together ring bit - Prep the Orbs - Insert Keys beforehand Other Ideas - Brainstorming - New drying/curing rack? - Fan/Ventilator/Air Purifier so I can work in the Garage? New Gas Mask? - New Lab Coat? $185, no sooner than Thanksgiving, Damn! - More Bins to hold garage junk - Bin rack for Garage to free up more space - Wheeled Chair in Garage? - Wheeled Table for Workstation? - Build a proper workbench? - Better Lighting? - Bluetooth Speaker? - Camera/Stream Setup? - Something to prevent cork bits from falling and making contact with the orb? - New Glass Container? - Hot Sand Box for even heat distribution? - Use the new hot plate? - Acetone Syringe Rack so they're not free floating? - Heavy Duty Shelf, for printers, storage, etc.? - Label Printer? plsplspls - Buy U-Line Boxes! - Can use a U-Line box for Badge Bucket. Probably too big for keys though. #### Mobile Orb Station Today I built out the new Orb Processing Cart, it took about two hours to lay everything out and get going. Afterwards, I cooked about 30 orbs. Everything is running smoothly ![[Pasted image 20241017201004.png]] ![[Pasted image 20241017201018.png]] #### Cook Test Batches Each set of 10 I tried slightly different parameters: 1. 90 second cook, no sanding 2. 90 second cook, light sanding 3. 120 second cook, heavy sanding #### Notes from Batch 1 - A few badges were too tight and collided with the hook - One badge was bad out of the box - I'm still recovering from my wisdom teeth extraction, so I got tired early. I need to do another 20-30 orbs tonight or tomorrow. - I'm considering buying a new gas mask - I'm washing the lab coat today for the first time since painting and such from the previous orb run. - The metal cart is already super worthwhile - There are 8 badge hemispheres that have a very fat band on them from a filament swap, before I figured out the Z-Offset trick. I don't have the energy to sand them right now, so I just threw them into a bucket for later ![[Pasted image 20241017201059.png]] - I finished cooking the orbs last night, and had a total of 58. - Let them cure overnight. QA testing in the morning. - 5 Broke, bad strings. I noticed most of the breaks were before I took a nap and came back. I think they come out better when I don't have a chair because I have to lean in a way that prevents acetone from getting into the badge. - There's definitely a visual difference after cooking between sanded and unsanded orbs. Asked Pancho to confirm if that difference is still clear after painting. - Shipped the 53 good ones out this morning. No plan with the scrapped 5. - Stopping Orb Production until Pancho pays me for the next batch. - Pancho did receive the Injection Mold samples in the mail today. They look amazing! ![[Pasted image 20241017201112.png]] ![[Pasted image 20241017201138.png]] #### Production Tracking & Sheet Still need to set up the production tracking sheet & books with Angel. Something like this: - Batch 1 - 116 Hemispheres Printed - 58 Orbs Assembled & Cooked - 5 Badge Reel Breaks in QA - 53 Finished Orbs Shipped - 0 Failed Prints #shopping-list - More Bubble Envelopes for Orbs ### Rat Rig #### Flow Rate Tests ##### Flow Cube 2 - Bridge Flow Ratio: 0.8 -> 1.0 - Thick Bridges OFF - 0.2mm Layers - EM: 1.000 - Fan: 40% - 250/110 Review: Looks mostly the same at first glance. Under harsh light, you can tell there are fewer gaps and the gaps are a smidge smaller. - Could try increasing EM to 1.03 - Could try increasing Extruder Temp to 260 - Could try Ironing - Could try lowering Max Flow Rate to 20/16/12 ##### Flow ube 3 - Extruder Temp: 250 -> 265 Review: This one looks better, a bit cleaner. Under harsh light there are still slight gaps between the top solid infill lines of the final layer. I'm sure this could be filled with ironing, but it indicates something is a little off. - Could try throttling speed/flow, but don't want to - Could try increasing temp further, but don't see the need - Could try printing with 0.4mm line basis instead of 0.6mm The patchiness only occurs in the middle area. So I'm assuming it's an issue with the power output of the hotend not being able to heat the filament fully during those fast, long, high flow stretches. I can recalibrate the PID, I can increase the temp, or I can throttle the flow. If I recalibrate PID, it would be best to reprint #3. But I'm not going to do that. I'll recalibrate and then throttle the flow to 12mm3/s and see if it proves anything. After that, I'll try thinner lines, then increasing flow rate, and then ironing. I should probably test Ironing either way, so I have it in my pocket when I want it. ##### Flow Cube 4 - Flow throttled 24mm3/s -> 12mm3/s - PID Calibrated to 265C Review: This is definitely the best one yet. The streaking pattern is still obviously visible, but the gaps are nonexistent. ASA Might just be a more glossy material than PLA. My first instinct is to try the 0.4mm extrusion width, but after reviewing the stuff on ellis3dp, it kind of seems like I'm just barely gapping. I'm going to try increasing the EM to 1.02 and reprint as is ##### Flow Cube 5 - Flow throttled to 12mm3/s - Temp 265C - EM: 1.020 Review: Definitely swung past the golden point. There are ridges now and the edges are smushed up and becoming rough. The EM I'm looking for is somewhere between 1.000 and 1.020. It's probably closer to 1.005. I'm going to try 1.0075, no other changes. ##### Flow Cube 6 - EM: 1.0075 Review: Still too much. it’s overextruding and producing ridges. The near-perfect surface I saw on Flow Cube 4 means the difference is very small. Trying 1.0025 next. Even if it comes out totally perfect on the next pass though, I’ll still need to tune it to 24mm3/s. I’ll be sure to save these numbers for a slow, max quality setting too. There’s so many forks in the road between each print! I’m developing a slow, ironed, quality preset in parallel with a fast, well-tuned, production preset. I’m still thinking I should look into CNC Kitchen’s microgram scale or whatever for this. ##### Flow Cube 7 - EM: 1.0025 Review: Just a bit over, trying 1.001 next, and 60mm/s top solid infill Looks fantastic ##### Flow Cube 8 - EM: 1.001 - TOP SOLID INFILL: 60 MM/S Review: Best one yet. Just baaaarely underextruding. There's like one or two areas with gaps between the lines. At this point I think I'm satisfied. It may be worth trying 1.00125 or something to that effect. At this point I can either keep the speed and throttle, and work on Ironing, or I can work on increasing speed and maintaining this quality level. I'll go for speed now. ##### Flow Cube 9 - EM: 1.0015 - Top Solid Infill: 60mm/s Review: It's perfect. But can it go faster? Trying 180 now that it's dialed ##### Flow Cube 10 - Top Solid Infill: 180mm/s Review: That definitely reintroduced some slight gapping. It's still very good. I'd put it around #8 in terms of quality. My first layer speed is 120mm/s and that splits the difference between 60 and 180, so that's the natural next test. 60mm/s is my bridge speed. It's not the end of the world if my top layer needs to be that slow, but I will be validating that assumption. It's probably also worthwhile to look into testing bridging specs after this. ##### Flow Cube 11 - Top Solid Infill: 120mm/s Review: Again, extremely slight gaps. I have a feeling I can keep 120 or 180 and set EM to 1.00175 or maybe 1.002 ##### Flow Cube 12 - EM: 1.00175 - Top Solid Infill: 180mm/s Review: I tried increasing the flow and the speed at the same time. But now that I've tweaked the settings in a few directions, I think I'd rather just set Top Solid Infill to 60mm/s. It's literally just one layer, and slowing it down is worth the extremely consistent surface finish. ##### Flow Cube 13 - EM 1.002 - Top Solid Infill: 60mm/s Review: I have doubts this will look better than the winning #9, but I think it's worth going over to see how it looks, just in case. It's incredible. ##### Limit Test 250x250x5 20 top layers #### Overview: Maintenance, Repairs, Upgrades, Calibrations, Issues Now that the Rat Rig is 1. Finely Tuned 2. Paid off 3. Successfully meeting production quotas I can look at some high level decision making. I factory reset my laptop trying to fix Satisfactory the other day and thought I might've lost my slicer settings. But it turns out PrusaSlicer saves a copy of all the settings profiles when you save the .3mf Project File. Thank god! I've been printing in batches of 16 so far. I tried printing a 5x5 but the mcu undervolted and crashed, which led me to make a list of organic maintenance routines I've arrived at, and some future repairs/upgrades. I've halted production for now and can make some of these repairs. ![[Pasted image 20241017201708.png]] #### Maintenance Routines ##### Regular Maintenance Routines - Every 3 days of printing or so, run cleaning filament through the nozzle, and use the brass brush to clean the nozzle tip. Also clean up any stringy bits that accumulate around the nozzle and fan - Before every print, re-oil the X-Rail and check both Y-Rails lubrication level - Before every print, alcohol wipe the bed - Every 3 days of printing or so, save config and shut down, power down, reset pi. #### Repairs ##### Pi Power Problems The Pi keeps undervolting, which slows down processing speed. I have a feeling this is why I sometimes just lose Mainsail and the printer has to fly blind until I reset it. Some possible sources of the problem: - Motherboard is overloaded and can't provide the necessary 5V current - The spliced jumpers I have bridging the mobo to the pi have bad connections - The Webcam's 6' USB extender is somehow overloading the pi Some possible solutions: - Rewire the power jumpers between pi and mobo - Install a dedicated 5V supply just for the Pi (reminder to connect grounds) I've had the mcu timeout because the pi was going to slow when I tried to launch a 25x orb 5x5 print the other day, so this is probably the biggest issue right now. If it fixes the mainsail issue, that's double good. ##### Motor Heat Motors are getting pretty hot during longer prints. No symptoms so far, but something to keep an eye on. ##### Rail Lubrication The long periods of heat-saturation and consistent airflow cause the oil I put on the X rail to dry up quickly. I've been oiling it in between every print. I'm considering this part of the pre-heat/prep routine for now. It might be worth switching to a more durable oil or grease or something in the future if this is an issue. I don't want it running 20k accel dry. ##### Cable Management Cables are on the floor in the back, it's a mess. I don't really have any good ideas for this. I feel like the spinal cord print that holds all the cables in place should have it's dimensions at least doubled. I'm pretty sure I can't replace those noninvasively though. ##### Filament Path The filament path is a little dumb right now. I'm considering flipping the tote around and letting it drop straight down through the center of the acrylic top-plate. This would shorten it a LOT but it would separate it from the Umbilical. I think it would be wise either way to get the 5kg filament spools off of the top of the printer. I don't know how it affects resonances but I'm sure it's not good. ##### Umbilical Stiffness I currently have a 3mm stainless steel shaft ziptied like crazy against the umbilical to give it the little bit of vertical stiffness it needs to ensure it doesn't sag and brush against the prints at any point along the bed. This works surprisingly well and has gotten at least 50 hours of print time without issue but it's still a crazy liability lol. ![[Pasted image 20241017203206.png]] ##### Squeaky Belt The left CoreXY belt is squeaky. I have an extra set of belts, and I don't think the belts are necessarily the problem. I think the toothed pulley inside the motor cage is the issue. The belts may be too tight. It's super invasive to get in there and replace the motor cage and possibly the pulley as well, so I've been putting it off. It makes squeaky noises all the time though which is neither mindful nor demure. To fix this, I'd need to - print a new set of motor assembly cages, - verify I have the necessary hardware for the motors and prints - reinstall both motor cages so they match color - undo & redo the belts - tune belt tension & gantry squaring And if I do that, that would be a good time to revisit some cable management inside and out, as well on the hotend & umbilical. #### Upgrades ##### Filament Dryer The only upgrade I would really consider investing in right now would be a 5kg filament spool dryer. I'm not sure how much that would really do for me. The big ones are about $200, so it's not a priority at the moment. #### Future Calibrations ##### Retraction I'm noticing some consistent stringing towards the end of the orb prints when it's printing around 0.05mm to 0.1mm layer heights. This melts away without issue during processing, but it's not great for other prints, so I could probably afford to spend a day or two getting autistic about stringing. #### To Do List Setting up for a pretty invasive work session Summarizing the list above: - [x] Rewire the Pi - [x] Install SSR for Part Cooling Fan - can't find - [x] Fit new Fan - too small, nvm - [x] Fit new Window Seal - [x] Print and Install new XY Motor Cages - [x] Cable Management - [x] Interior - [x] Hotend - [x] Filament Path - [x] Exterior - [x] New Filament Rack ✅ 2024-10-24 - [x] Write Preheat Macro ✅ 2024-10-24 - [x] Sweep and Mop under Area - [x] Compress Working Space - [ ] Space Divider - [x] Clean Hotend ✅ 2024-10-24 - [x] Clean Tartarus - [x] Clean Acrylic Panels - [ ] Install a sticker? - [x] Design & Print an expanded Toolboard holder with more Vert - gimp suit - [x] Design & Print a Power Strip Platform - no need #### Misc Notes Definitely seeing some issues - LED Panel Mounts are burnt, but that's expected - Weird white dust on upper umbilical holder - Lower belt is frayed - Rightside XY Motor Cage is grinding the belt and producing belt dust - Exhaust Fan is tangled up with cobwebs - Gap space above beacon and behind hotend is stuffed up with cobwebs I can create a floor plate to put inside the bottom of the electronics zone and mount the Power Strip there so it has a permanent home and can't fall. Should probably replace this power strip, maybe get one with some diagnostics. I may be able to redesign the cable snap collars to enclose a larger volume, which would fix a few problems I'm definitely going to have to button it back up in order to print the motor cages, and some other stuff too. Considering Mounting the LED Panels to the topside. But I don't have a good means of extracting and rewiring them. If I can get the cables to actually clean up and fit in that space, I can mount a magnetic clear acrylic plate to box them in and then put a filament box behind the printer and have it run straight up. That would be ideal. The Motor Cages I printed are showing definite signs of overextrusion... I guess I don't print anything but orbs and cubes so real parts show up a little differently. I'll probably dial it down later #### Special Topics ##### Rolling Casters https://www.modix3d.com/product/casters/?srsltid=AfmBOopJb1YRM6oLXpy3A019VBtUanmYod9Z6R2YpXfO8xBEiPwxPuwi Something like this might be nice, but I'm skeptical. Not sure if there are casters that will mount into the Extrusions, of if that's a terrible idea and I should be putting it onto a dedicated wooden palette or something and mounting the casters into that. ##### 4028 Part Cooling Fan This thing seriously takes 20W? It does produce 6x the cfm. Costs about $20. I don't think it's in my best interest to replace this one though. It's really powerful and there's no sense in limiting my capacity for cooling. I'll have to find some other way to get it below 64dB... ``` Fan - 4028 Axial - NIDEC W40S12BS2E5-57T04HF9 (12V 1.61A 25000RPM) **Fan specs:** - Voltage: DC 12V - Current: 1.61A - Input Power:19.3W - Speed:25000 RPM - Air Flow:30.3cfm - Noise:64 dBA - Bearing:Ball bearing - Feature:Speed Sensor (Tach Output), PWM Control **Dimensions:** Height: 40 mm Width: 40 mm Depth: 28 mm **Wire Termination:** - Black - 0V - Red - 12V - Blue - PWM (Control) - White - Sensor (Speed-Tachometer) ``` ##### Cobwebs If I tune retracting and stringing I should be able to get rid of this issue. I'm scared to turn off Z-Hop, but as long as I'm diligent about managing the preheat cycle I think it should be okay. We'll see I guess. I imagine tuning retraction is going to be a huge pain in the ass. ##### Thermal Mass This would also be a good time to look into the thermal salt thing Claire sent me ##### Pi Undervoltage Problem (Fixed) I removed the USB extender from the webcam, and spliced together new supply wires - neither fixed the undervoltage error. I figure there are too many USB devices plugged into the Pi. - Beacon - Toolboard - Webcam - Motherboard (Data & Control) Either way, what I found was unplugging the webcam removed that issue. Not sure what to do with this information. After sanity checking with ChatGPT (Which did not make me feel much better) I removed the 5V GPIO pins bridging the Mobo to the Pi, and then plugged in my 5V 3.1A USB-C into the POWER IN USB-C of the Pi. This clears the undervoltage issue. I'm guessing there was some limit to how much power the Pi could manage when sourcing from GPIO. ##### Extruder Dust I found some rust-colored dust around the extruder the other day. It cleaned off easily, but I'm not sure where it came from. The extruder runs very hot during printing, since it is just one little gear yanking a 5kg spool through ~1.5m of PTFE. I've been wondering if it'll burn out, or grind itself to dust and I'll end up replacing it. I think I'll wait for that issue to present itself though. #### Shopping List - New Head Lamp - New X_Endstop - Noctua Brand Part Cooling Fan? - New PTFE Run - 4" to 2" adapter for Ducting - Air Quality Monitor - New LED Panels - Dedicated Android Tablet? - Long, Small Zipties - New 5V/3.1A USB-C Pi Power Prick - New Power Strip - Custom Clear Acrylic Plate for Electronics Protection #### Exhaust System ##### Noctua Fan It's uh, really small. I thought it was bigger. It's a 40x20. Is this what I meant to use to replace the exhaust fan? It's too small to replace the part cooling fan, which is the louder of the two anyway It's rated at 9.4 m^3/h for airflow 1 m^3/h = 0.588 cfm, so 5.52 cfm I guess it'll fit inside the same profile as the current exhaust fan, it's just so tiny. The Inline Duct Fan I'm using now is 12V/6W and has 62cfm max, but I usually run it near min and it works fine. ##### Recirculation I'm considering doing a [recirculating fan with a charcoal filter](https://clearviewplastic.com/products/carbon-air-cleaner-3d-printer-carbon-air-filter?variant=44083080921326&currency=USD&utm_medium=product_sync&utm_source=google&utm_content=sag_organic&utm_campaign=sag_organic&gad_source=1&gclid=Cj0KCQjwsc24BhDPARIsAFXqAB1PglPiSC_ZZjz1K353iH1ZJmRYQw2tsV0_7oJViF8fAcZyZHok7m4aAnsFEALw_wcB) like new the voron has. That product is like $170 though and I don't know why. [Reddit Thread Discussing Printer Ventilation](https://www.reddit.com/r/AskEngineers/comments/154fjc5/comment/jsqqsxw/) This guy suggests you just throw the whole printer inside a tent and give it its own extraction system. It's something I've considered below but I think that's just way too overkill. I'm still trying to find ways to keep noise down too. I suppose I can run some tests during extraction tuning with the cooling fan low or off and see if that gets me anywhere. I still have the resonances issue when moving around 100mm/s but that might go away with a new filament rack [Prusa Forums Post Discussing Printer Ventilation](https://forum.prusa3d.com/forum/original-prusa-i3-mk3s-mk3-improvements-archive/filtering-abs-fumes-options-for-enclosure/) ![[Pasted image 20241019141210.png]] [http://www.cfpdirect.co.uk/products/hepa-filters/](http://www.cfpdirect.co.uk/products/hepa-filters/) . [http://www.fabbaloo.com/blog/2015/3/1/finally-a-3d-printer-filter-accessory](http://www.fabbaloo.com/blog/2015/3/1/finally-a-3d-printer-filter-accessory) ``` Yes, an enclosure is required, Ultrafine Particles are very difficult to capture, HEPA filters are limited to capturing a small percentage, and carbon filters only suppress the odor. The only cost effective way to reduce the level of Ultrafine Particles is to repeatedly scrub the contaminated air, reducing the level with each pass until the total concentration has been reduced to safe levels. ``` ##### Sketch 1 - 2x120mm fans and some HEPA Filters I can get 2 PC fans and print an enclosure to strap them to a 120x240mm HEPA+Carbon Filter and then mount that to the ceiling of the enclosure. That's all doable and there's nothing competing for the real estate really. If something like that can work, and works well, I can stop exhausting and just recirculate, which will probably be good for heat retention. I don't really have any heat retention complaints at the moment though. It may be worth looking into a high heat lubricant... Fan Option 1: [Noctua NF-P12](https://noctua.at/en/nf-p12-redux-1700-pwm/specification) - $16 - 25 dB - 120.2 m3/h or 69cfm Fan Option 2: [Antec P12 Series](https://www.amazon.com/gp/product/B08XQ6X86C/ref=ewc_pr_img_1?smid=ADT2TM5FCS88L&th=1) - 5 for $30 - 35 dB - 60 cfm [HEPA & Carbon Filter Combo Kit](https://www.amazon.com/gp/product/B01M5CX59R/ref=ewc_pr_img_2?smid=A2BYTFJZOUPNVA&psc=1) - $23 for 4 of each I'd get 120cfm or ~200m3/h from the parallel fans in ~1/2m^3 of space, so about 400 air exchanges per hour. If I mount it hung from the ceiling it'll help recirculate the hot air as well, which is huge. ![[Pasted image 20241019143138.png]] ##### A Sketch from Claire ![[Pasted image 20241020175015.png]] #### Cable Dressing Started again around 9am. Dressed all the cables, cleaned off dust and paint residue. I went through like, 200 zipties I'm guessing, but the before and after is fantastic ##### Before ![[Pasted image 20241020175156.png]] ##### After So yeah that took about 9 hours on its own, including cleanup. Overall, I've been working on this printer since around 1pm yesterday, and slept for 6 hours - so about 23 hours of work overall. #### New To Do List So here's what's left: - [x] New Filament Rack ✅ 2024-10-24 - [x] Write Preheat Macro ✅ 2024-10-24 - [x] Consider SSR for Part Fan ✅ 2024-10-24 - [x] Consider SSR for Exhaust Fan ✅ 2024-10-24 - [x] Clean Hotend ✅ 2024-10-24 - [ ] Space Divider - [ ] Sticker Install - [x] New Flow Tests ✅ 2024-10-24 - [x] Retraction/Stringing Tests ✅ 2024-10-24 - [ ] Do some shopping ### Meta #### Meta Systems ok time to come up with a project and then annotate the process of developing it using the schemas i spent like 3 months coming up with. well, that's what i will do, once my brain works again. i also need to consider how to manage sprint documentation in parallel. this note is probably the beginning of a similar organic synthesis to what i just finished. this is pretty hard! #system-need Sprint-Based Notetaking #### Sprint Sheet 2030 The calendar is ajar, the heavenly bodies are not spherical, and all the lights in the sky are askew. Also I don't actually know anything about Scrum lol. Let's just cut a week off of Z and A since it's holiday anyway, and I end up spending that time reviewing and planning. So now instead of 364 days it's 350=25x14. Essentially just merging Z and A, or deleting Z. I have the sprints for 2024 mapped out on my other calendar. 2024-U is supposed to start on 10/7, and it's only 10/5, but it's not that serious. #### Datasheet Thoughts Components-First approach seems appropriate. The point is to develop deep understanding of a few components so it's easier to understand similar components. Basically, read more data sheets. I annotated and read out the LM7805, but I didn't actually go and summarize the useful information anywhere Also, It seems worthwhile to cautiously separate pure .md from extensions like excalidraw and pdf annotator that can't be read as plaintext. #### Sprint & Log Structure - Separate Notes I figure these are relevant to the current sprint, even though they're in different notes. This note should cluster everything that goes on during this timeframe. Then at the end when I write up a summary I can just link everything here and curate a slideshow. I could even manually redirect all the assets to a new folder each sprint so it's easy to collect. #### Header Logic Wondering if I should nest the Test Logs into a master like the below. If I do that, I feel like I'm just wasting a good heading layer. I only have 6 and that puts me at 5 already. #### Ex: Formatting Tests ##### Test 1 ##### Test 2 ##### Test 3 Like how do I go and talk after this? I need subsection headings within project log entries. That is: Log > Timestamp > Project > Test/Subsection Anyway. I need a Lab Shopping List, since I need a adj. screwdriver Decided against buying a webcam for now. If there's demand for a stream it'll come later. Not a good use of my money at this moment. Now, I'm seeing pretty clear rails being formed for Orbs and the Actual Design for the Breadbox, but not too much for how I intend to use the Breadbox. Setting up the shopping list and maybe a project specific BOM seems like a useful effort. I can collect the Logs One issue I'm already running into is, when do I Log Break? That is, make a new timestamp and new sections, etc? It's 7pm now, and it's been 12 hours since I worked on stuff earlier, I'm not going to work on the same stuff, so it seems appropriate to Log Break here. --- Some other meta systems to set up - system-need - Works pretty well, and can funnel into a meta note - shopping-list - Is a little confusing. There's stuff I want for the lab, and there's stuff I want for the projects specifically. - Then I should also be maintaining an expense log for the lab as a whole, with expenses allocated to specific projects or to overhead - folder-structure #system-need - Obsidian-Integrated File Folder Structure #system-need - Meta Lab Shopping List --- Okay, so where am I at right now - Multi-CAD File & Folder Structure - PDF Annotation & Component Database - Lab Bookkeeping, Project Ledgers, & Shopping List - Obsidian Toolchain & Meta-Systems - Sprint Log, Compilation, and Synthesis System - Project Management & Documentation Schema how about uhh you know Current Projects? and uh Project Status? Timeline? Daily Updates? To Do? Anyway, I spent the vast majority of my brainpower on this today. I'm gonna go play factory games now. #### GitHub Pages [# How to Host a Website On Github Pages](https://youtu.be/OltY8JIaP-4) #system-need - Github Pages Hosting Thinking about setting up some GitHub pages for this stuff. Probably the most expedient way to get the lab up and running again, at least tentatively. #### Logging & Chopping Sketch Noticed I was kind of avoiding adding onto the Log because it's long and from "before" and I don't have a "now". Which I think means I need somewhere to extract and compile to. But I'm not sure. Maybe I just do all that at the end of the Sprint and get over that feeling? The !Log note IS the "now" I've already run into the issue of appending a previous days' lab entry because the context is already established and it'd be redundant to re-establish it if I'm truly picking up where I left off. The appended entry can't be meaningfully timestamped with this format/structure, but it's really. not that big of a deal. Extracting chunks of the log hog into distinct notes and relating them back to project masters is a little more complicated. I imagine what will end up happening is: ##### Note 1 - Project1 Master Note - Timeline - Sprint U - Devlog 1 - Devlog 2 - Sprint V - Devlog 3 - Devlog 4 ##### Note 2 - Sprint Master Note - Projects - Project1 - Devlog 1 - Devlog 2 - Project2 - Devlog 3 - Devlog 4 ##### Note 3 - Master Hog Log - Sprint U Raw Log (Hog Log) - Sprint V Raw Log (Hog Log) #### New Project Ideas & Proper Zettelkasten - New Project Ideas or shallow dives should go into the Wisps folder, since they don't reflect committed work. !Log and Sprints and Projects are for committed work. They fit fine inside Ghosts, but they're just too ephemeral and ADHD to include here, so they get their own section. - The hotkey for a wisp is CTRL+SHIFT+NUMPADPLUS - I'll figure out a system for managing the Wisps later. #### Personal Electronics Component Database - I talked to Tyler and Angel both for a while about the problem of keeping a personal collection of annotated datasheets for electronic components I've reviewed and used before. - Octopart is the most accurate and comprehensive electronic parts database - https://octopart.com/ ##### Basically, there are a few problems: 1. Documentation is Hard 2. Precision is Hard 3. Characterization is Hard ##### Expanded: 1. Proper Documentation is hard 2. Precision Measurements & Instrumentation is hard 3. Thorough Component Characterization & Parameter Mapping is hard ##### Expanded further: - Datasheets and manufacturers make up new part numbers and then generalize and/or just lie in their datasheets - Personally verifying the specs listed in datasheets to any meaningful degree of precision is extremely time and labor intensive and balloons project complexity - This idea of a personal library is much more suited for a matrix-based database application than a linked-journal application like Obsidian. A simple spreadsheet, or Notion are options - There is a tremendously hairy documentation problem here - Developing a process for acquiring a database entry-characterization of any single component that is both verified and precise is a problem so hairy as to basically become the primary operation of the Lab. - Do you want to be a precision and instrumentation expert? - Precision, Characterization, and Documentation are intrinsic and fundamental frictions to working with electronics, but there doesn't seem to be a Best Practice approach, or a superior methodology for handling all this complexity and ambiguity. - Something worthy of being titled the Electronomicon would necessarily require a very firm understanding of the problem and present best practices for resolving them at both the library and project levels. --- ## Pretty Pictures [[#Sprint U - Ursaring]] - back to top ![[2024-10-21 09.37.10.jpg]] ![[2024-10-03 10.04.56.jpg]] ![[2024-10-04 15.32.43.jpg]] ![[2024-10-07 17.20.34.jpg]] ![[2024-10-08 16.16.57.jpg]] ![[2024-10-13 07.20.33.jpg]] ![[2024-10-16 13.08.04.jpg]] ![[2024-10-16 15.52.39.jpg]] ![[2024-10-16 15.55.05.jpg]] ![[2024-10-19 12.11.23.jpg]] ![[2024-10-20 16.54.26.jpg]] [[#Sprint U - Ursaring]] - back to top