![[Suneater Labs Update 8-5.png]] # Lab Update *2025-05-08 -> 2025-08-05 | [[#Pretty Pictures|Skip to Pics]]* **Contents:** - [[#Joule Thief Complete]] - [[#Highlights]] - [[#Summary]] - [[#Timeline]] - [[#Project Notes]] - [[#Joule Thief]] - [[#Plasma Toolhead]] - [[#PCB Book Club]] - [[#Rat Rig]] - [[#Breadbox]] - [[#Pretty Pictures]] ## Joule Thief Complete Joule Thief is finally complete! Check it out at: - [[!Joule Thief]] - https://github.com/fishPointer/Joule-Thief ## Highlights - Have been exploring Plasma Toolhead/PCB Maker with VCs and many engineers - Jumped from 4k to 8k followers since last update - Started PCB Book Club, the premiere electronics design community on x dot com - Finally "finished" the Rat Rig to my standards, mods to increase chamber temp to 75C - Did a heroic sprint to bring Joule Thief to the finish line just in time, great execution - Exhibited at Open Sauce, met a bunch of oomfs, had a great time - Rebuilt and reorganized both my Lab & Garage - More opportunities than I've ever had in my life show up in my DMs - Got trapped in the Obsidian Mines producing Documentation Quotas ## Summary I have been EXTREMELY active the last three months. The previous Lab Update was shortly after I was approved to exhibit at Open Sauce, which was a huge milestone for me, and since then I have "drawn the rest of the owl". Honestly, my life looks and feels completely different since then, and it's pretty much exclusively because of Twitter. I mean, it's a combination of me working like a dog day in and day out just to post more technical OC, but still. It's been exciting and overwhelming to say the least. My account started repeatedly breaching containment in mid-June while I was heads-down trying to get Joule Thief ready for demo day and I went from 4k->8k during this period. Around 6k is when it got to be too much and I could no longer maintain 100% coverage & continuity over all my notifications. I have been making a very deliberate effort to read every reply on every post, and reply in turn as appropriate. But it's just not possible anymore. There's been a tone shift since then as well. People have started talking about me a bit differently, lots of people reach out and ask for advice of all flavors, or just to say hi now. Some people hit me up for electronics design contracts, or just to build whatever with them. I've even had a few job offers and VCs reach out interested in me and what I'm doing. (If you are one of these people I am losing sleep because of you!) That part is shocking to me because I'm just so used to the HR timeloop of waiting weeks and weeks to get an interview for a position that's already closed, etc. etc. Genuinely, I'm not used to having Options. Nor am I used to getting treated like a niche internet microcelebrity. None of this was really on my list for this year, but it's been a fun adventure so far and I intend to keep doing what I'm doing. All the stuff that I do for Lab is all the exact same stuff I've been doing the last ~5 years or so. The only difference is there are people looking at the pictures I take and laughing at the jokes I usually have to keep to myself. The hardware life isn't as lonely anymore :) My life hasn't materially changed just yet, but it seems like things are coming together more and more each day. It's troublesome figuring out what to do when so many doors open at once. If the money fell from the sky with no strings attached, I would simply quit my job and build and post all day every day. Unfortunately a patronized sabbatical to hide in my cave and do plasma R&D for 12 months doesn't seem to be one of the doors, so I'll need to figure something else out. I'm a pretty all or nothing guy though so the next best alternative to being a financially independent turboautist househusband is building a billion+ dollar tech company. That sounds hyperbolic because it is & because I am. It's called Suneater Labs for a reason, after all. Anyway something big I've learned (confirmed) over the last few months working on Joule Thief is that I really can learn new things and upskill extremely quickly & efficiently. Genuinely faster than almost anyone, especially if you drop me in-situ with people that have the knowledge and experience so I can start mining them for it. Especially now after this Joule Thief arc I have developed so much more cognitive stamina. I can sit down and do novel knowledge work for 12+ hours without fatiguing. Motivation and inspiration are definitely big factors here, but the bottom line is: Adaptability go brrrr. The primary downside to this I've found is the typical Deep Work problem. I can tell my mind has a very large time constant, and so interruptions early in the work session really mess with my ability to get the flywheel moving. Unfortunately, my current position means I'm essentially on-call at all times, so it's generally just easier to do all-nighters and sneak naps in than it is to sneak labs in. My mindset and approach are dramatically different now that I have access to so much expertise from the peanut gallery. Being able to just DM people with PhDs that know everything about lasers and nanofabrication and PCB supply chains and everything in between makes me want so badly to go NEET for like 6-12 months and "powerlevel in public". I am exponentially faster when the friction is gone and I'm not spread so thin - I still work a day job in between all this! Some people think I'm a seasoned electronics engineer, but I don't have any professional experience doing this and have tried and failed to get an entry-level electronics job probably like a dozen times LOL. I didn't even know what a soldering iron was until I was 19 and I barely graduated with my EE degree due to a confluence of extreme stressors in my personal life... I'm still playing catchup in a lot of ways, but it doesn't bother me as much now that I get to talk to people just like me every day. Joule Thief is my first time doing this from scratch. I referenced no prior art and did not fork from anyone else's existing design. I did the whole thing as cold and blind as possible. It's the first system I designed myself from arbitrary specs and the first time I've ever tried to route a board with any sort of engineering best practices in mind. I've never picked my own chips, read & reviewed the datasheets, sourced & implemented them, designed my own analog/discrete component solutions, done any SMD soldering, or worked with rechargeable batteries or any sort of custom power electronics application, precision instrumentation, etc. These were all firsts for me, but I specifically structured the project so that I would have to learn those things in order to get it done, because those were skills that I wanted to have. And now I have them! so mission accomplished. That's just kind of how I do things, and it's why there is usually very little overlap between my Main Projects. I can move on and go learn something else now that stacks on top of these skills. A few people noticed the early passes of Joule Thief showed how clearly inexperienced I was vs. the level of bench gear I had laying around. It's because I made the commitment to myself to focus single-mindedly on getting good at designing electronics and PCBs at the beginning of the year, and to use Joule Thief as a springboard to improve & to get into Open Sauce. EE isn't going anywhere for me, so it was an easy call to make to drop $2-3k to ensure I had all the precision & tooling I needed (and no excuses to stall or pivot!). This whole endeavor went really well, and I'm glad I committed to it, and I'm glad I started sharing my lab work on twitter when I did as well. I know a lot of people have been "along for the ride" and subscribed to the Joule Thief journey, and are now on board for some larger fish arc. It feels weird for me because I'm so used to having to teach myself everything in complete isolation. The increased pressure comes with community support, and it's a very clear net positive for me. I have no shortage of vision and ambition for these things so as long as I can find the resources and articulate the plan, I'm sure it'll work out. Open Sauce went great btw! It was the first time in a long time where I got to show up and show off (I haven't left the house to Do Something in years). It was really surprising standing around at my table presenting the project and maybe a few dozen people each day came up and asked for fish. The twitter followers thing has just been a number for me, like an online game where I dump all my quips and niche jokes I normally crack to myself in silence. Seeing all these people irl was an eye-opener because as it turns out they're not all bots some of than are real people! lol. Unfortunately I didn't get to spend very much time exploring exhibits (or sleeping), and even if I did, I definitely didn't have the energy to get nerdsniped touring the booths, but the experience was still really positive. I've been running on fumes for a while, is all. I'm glad that it's done now so I can finally provide the final writeup and consolidate all of the Joule Thief notes and put a bow on this whole thing. The most impressive part of the Joule Thief by far, is simply the fact that it is a finished project. It went better than I could've ever hoped. It's a tight six (well, seven now) month arc so at this point I'm just tidying things up and preparing for whatever comes next. I'm not really sure what that looks like just yet though so stay tuned I guess :p thanks for reading all this btw! -fish ## Timeline - 5/21 - Made an Axe - https://x.com/fishPointer/status/1925344078772019229 - 5/25 - Started PCB Book Club - https://x.com/fishPointer/status/1926891838448849083 - 5/29 - Set up New Lab - https://x.com/fishPointer/status/1928091496667840982 - 5/29 - PCB Book Club to 100 - https://x.com/fishPointer/status/1927993358217269495 - 6/6 - First oomf Meetup - https://x.com/fishPointer/status/1931185350623535164 - 6/7 - Started Joule Chief - https://x.com/fishPointer/status/1931573454031274104 - 6/9 - Redid the Garage - https://x.com/fishPointer/status/1932147588565999725 - 6/11 - Feedback on Breadbox - https://x.com/fishPointer/status/1932972065700864179 - 6/14 - Full Siglent PC Integration - https://x.com/fishPointer/status/1934095696485044609 - 6/22 - Start Joule Thief Sprint - https://x.com/fishPointer/status/1927993358217269495 - 6/27 - Full Breadbox Layout - https://x.com/fishPointer/status/1938502116257239156 - 6/30 - Joule Chief Schematics - https://x.com/fishPointer/status/1939919885087515073 - 7/1 - Joule Chief Board - https://x.com/fishPointer/status/1940285457138618474 - 7/3 - Joule Thief Board - https://x.com/fishPointer/status/1940748187712782778 - 7/12 - Breakout Soldering - https://x.com/fishPointer/status/1944273011206680752 - 7/13 - Firmware Development - https://x.com/fishPointer/status/1944558508520931493 - 7/14 - Functional Joule Chief - https://x.com/fishPointer/status/1944789496941400525 - 7/16 - Soldering Day - https://x.com/fishPointer/status/1945405643722281139 - 7/16 - Joule Thief Finished - https://x.com/fishPointer/status/1945881833822171627 - 7/18 - Open Sauce Setup - https://x.com/fishPointer/status/1946285111844249756 - 7/20 - PCB Book Club Stickers - https://x.com/fishPointer/status/1947003232955642290 - 7/20 - Open Sauce Day - https://x.com/fishPointer/status/1947012477260300589 - 7/24 - Open Sauce Closeout - https://x.com/fishPointer/status/1948474043851759809 - 7/27 - Ping Pong Table - https://x.com/fishPointer/status/1949705817302716686 ### Account Growth Curve: - 2/6 - 1k - 2/21 - 2k - 3/21 - 3k - 4/23 - 4k - 6/12 - 5k - 6/23 - 6k - 6/27 - 7k - 7/14 - 8k - 8/6 - 8,867 ## Project Notes ### Joule Thief - Sketched out parallel architecture & bootstrapping circuits - Stalled for a few weeks working on rat rig, setting up new lab, setting up new garage - Returned around 6/14 by setting up PC connection for all my Siglent devices - Sketched Joule Thief N-N Architecture based on a mainboard called the Joule Chief - Designed a breakout board for the LTC3105 - Slowly pieced together a complete breakout version of the Joule Thief's parallel extraction architecture on Breadbox to validate the decision and develop the final Schematic - Got a bit better and faster at throwing together boards, schematics, as well as producing quick breakouts & modules for breadboarding - Wired up the busiest and most hierarchically complex breadboard I've ever made (previous was MFOS Noise Toaster Analog Synth) - Did many all-nighters over the course of ~2 weeks in order to move the project from Breadboard into Schematics & Layout and ultimately order the boards - Cooled off for a few days, then began developing more robust firmware for the system, worked on misc. lab stuff - Identified a good handful of errors in the board layout, thankfully none were critical to kill the project on short timeline - Soldered 1 Chief and 4 Thief Modules over another few all-nighters and confirmed everything worked first try - Touched up firmware and confirmed all core functionality was correct & consistent across modules - Printed a frame to mount all the boards to, and then packed it up for Open Sauce - Exhibited Joule Thief, Breadbox, and Big Benchy as the main events at Open Sauce - Project worked on site first try, and continued working fine throughout the entire con with no hiccups (!) - Came up with a lot of interesting ideas for a Joule Thief V3 that I will probably not pursue - Retired the Joule Thief to the Finished Projects Shelf - DUSTED! - Spent like 2 weeks combing through the copious amounts of documentation and curating it into something coherent ### Plasma Toolhead - This idea has been explored and discussed extensively over the last several months, with little direct action. - As mentioned in the previous post, this whole investigation began as an inquiry for solutions to "The Via Problem" - **List of Via Techniques:** Rivets, Spot Welding, Chem-Cartridge Electroplating, CoreXY Electroplating, Powder Press&Sinter, Copper PVD, Thermal Cure Conductive Paste, Laser Carbonization, Pre-via'd Board Stock - I've talked to a lot of very interested engineers, or people who work in PCB Fabrication with strong opinions. - I've discussed the market & opportunities for different niches and approaches for this sort of thing with a TON of people - Team and idea have both revolved several times, and is currently stable. - Like any good prospective founders, we've pivoted about a dozen times in the process of refining the idea & our approach - Building a MVP for plasma deposition toolhead to seek R&D funding - Purchasing and reverse engineering a sputtering machine - Purchasing and reverse engineering the LPKF catalog and integrating prior art into a monolithic machine - Purchasing the existing toolchain as discrete machine steps and working towards automation/integration - Pivoting to developing a framework for retrofitting for & designing fully automated "Dark Factory" Lines - Pivoting to setting up a single high-spec high-throughput PCB Fab & Assembly Line - Pivoting to chopping aluminum CNC precursor blocks with a bandsaw using conveyor belts to dump them into the back of a waymo for full automation - Pivoting to a 32-Layer, Max Spec ASML-Class PCB Fabricator & Assembler capable of producing boards at the edge of hardware development that fits into a shipping container to deploy containerized dark factory lines as a product - Pivoting to building high kVA transformers and becoming a T&D Infra supplier - Because the type of person who will buy Breadbox is also the type of person who will buy our commercial PCB Maker, we've also considered using the Breadbox as a bootstrapping mechanism - kickstart with breadbox and offer discount for early backers when PCB Maker comes out. Breadbox will happen either way, regardless of association to the PCB Maker initiative. - The primary mission is Stiffening the US Electronics Manufacturing ecosystem. - Whether that's bottom-up (Distributing 2-4L Board Fabricators) or top-down (ASML-Class Computer Printer) remains to be seen. We're mostly in the process of calibrating our approach to maximize TAM and minimize Technical Risk. - The **ASML-Class PCB Machine** is currently preferred as it sparks autistic joy for everyone, and serves big fish clients like Defense, Medical, Aero, Smartphones, HPC, etc. - It also enables a wider technical moat, and will make the deployment of a Consumer 4-Layer machine that much easier down the line. - Main drawback is the tremendous R&D lift. I believe the level of talent I have access to is sufficient. Technical, funding & timeline are the main risks, and so a technical-funding milestone sequence is probably necessary. But really the most important step is figuring out what the very first MVP/demo needs to look like in order to continue building investor confidence & trust in the mission. - Developing a framework for Containerized Dark Lines like this will be extremely valuable in a world where labor is scarce/expensive, IP protections matter, and adaptability > scale. - My ultimate goal is to develop advanced manufacturing & energy infrastructure for space. To that end, I consider the development of highly compact and reliable autonomous factory units a near absolute prerequisite. ### PCB Book Club - I've had a vision for a while now of getting a handful of electronics engineers together to make passes at the same design, share them, review eachother's work, and exchange tips and tricks to produce a community of mutual improvement. - I saw Yacine & Stocko running Robotics Journeymen with the singular hard rule of "post updates weekly or get banned" and thought that was just perfect. - So I launched PCB Book Club and warned everyone they had until July 1st to post a board they worked on in some capacity. Entry to the club also requires at least a description of your board. - We got 100 members in basically the first day, and swelled up to around 300 members. Then when the culling came, I removed a little more than half of that. It was a little more work managing that spreadsheet than I was expecting LOL. - It's been fantastic though and everybody understood the assignment immediately. I didn't have to put much effort into spinning up the community flywheel, people just showed up and started posting designs to show off or solicit feedback and a lot people have reached out to say they're very satisfied with it so far. I'm happy it's gone well. - Personally, I'm farming all of your board designs to practice cold/no-context PCB visual analysis so I can get to Larry's level. It's also been tremendously helpful for me since I would not have been able to produce the Joule Thief/Chief without learning from everyone's feedback. I've upskilled dramatically over the last few months and it's primarily because I have access to people way smarter than me that are either willing to help or that I can bait into criticizing my work :) - Thanks again to everybody who joined and has been participating. It's been wonderful to see, and it's clearly made a lot of people happy. I'm looking forward to continuing to cultivate and serve our electronics community in many ways. - I'm not sure what the next culling criteria/assignment should be just yet. I haven't had a lot of bandwidth for it lately, but there definitely will be another culling soon. Probably 9/1 or 10/1. So just be ready! ### Rat Rig #### Developments in my 3D Printing Journey: - Fried an egg on the bed - Purchased Larry's P1S as a reliable ASA-compatible backup - Replaced Belts, instant improvement - Tuned EM/PA+SAFC very carefully, achieved new highs - Printed, Smoothed, Painted, and Shipped a giant cosplay axe in 1 day - Printed some stators for Parth - Began exploring the world of high temperature printing & PPA-CF - Printed additional Breadbox parts for redesign to kit/commercialize - Tuned Chamber Heater for higher temps, 55C -> 75C - Installed foam insulation, swapped high temp lubricant - Backed up Pi SD Card and added active heatsink aluminum case/fans to help with high temp CPU throttling issues - Rebuilt entire Z-axis system after part failure #### Current State: - Rat Rig can reach ~73C chamber temp in about 25 minutes now, drawing 2kW during preheat and 1kW during operation. - Raising the temperature from 55C -> 73C has resulted in dramatic improvements in the surface finish and strength of my ASA parts. Layer delamination (Snapping) was the primary failure mode of the prints before, at very low stresses. Now they can handle much, much more, and are living up to the potential of the material. - I'm interested in chasing 80C chamber and redoing the entire electronics panel as an exercise to get good at DIN Rails, but I'm going to save that for later I think.... - Rat Rig is good to run 1-click prints with little to no babying/oversight and can easily handle the entire build plate. I got like 5x 8x8 Gridfinity Baseplates in a row with no hiccups whatsoever, which was a huge accomplishment for me. The system is performant, fast, stable, and requires very little overhead now (until I start upgrading it again) ### Breadbox - Have been brainstorming ways to kit it and productize it - Have probably gotten ~100 verbal LOIs across x and irl feedback at Open Sauce - Came up with the idea to market it as moving from Terminal -> IDE for hardware development - Came up with the idea to put 20 of them on a Baking Rack for context/project-specific workstations - Came up with the idea to put 20 of them on a Baking Rack, 1 for each student in an electronics lab - Currently redesigning in 2 different sizes, a smaller, cheaper, simpler, economy version, and a full-size pro/IDE version. - TONS of ideas for custom tooling, PCB modules etc. to integrate into the Gridfinity Framework as well - Collected a ton of feedback from people at open sauce as well - Product is likely to be a big hit, and am looking to launch in August maybe? Likely limited run. - Purchased a Miter Saw for this purpose, but I also just generally wanted the ability to cut aluminum extrusions some more ## Pretty Pictures [[#Lab Update|Back to Top]] [[Block 1.png]] [[Block 2.png]] [[Block 3.png]] ![[Block 1.png]] ![[Block 2.png]] ![[Block 3.png]]