Here’s the update for February!
Note: Audrina will be providing her own update & take on current events as well, so I won’t go into much more on part status and stuff. We continue to receive parts as we order them, and Audrina will be sharing those updates.
Kickstarter Update - This has some duplicate info, but just in case you care about Pika specifically.
Here’s a video that describes alot of what we’re going through and why this month is a bit different than the last few, Including why the update is late vs before.
This machining journey I started at the beginning of 2025 was not taken lightly, nor really was it what I thought I’d be doing so soon. In Q3 2024, I placed PO’s and put deposits down for material for what would be chube compact prototypes, getting them made at the same location as the regular chubes. I enjoyed working with the people, had excellent transparency on pricing, and could visit for fitment tuning during beta programs.
Amusingly, they had a history of being quite late with delivery (does anyone remember gen1 Chube launch?), so instead of getting parts in november, it extended to december, and then January 2025, when I got a call from what apparently was their new owners. This phone call (and subsequent calls) informed me that my costs were going to 2-3x across the board, sometimes citing “material costs” (I was the one who priced the material, it hadn’t changed, and I had *already put money down and the material was purchased*) and citing labor costs going up (they had decided that 1 person who normally operates 3 machines at once would now garner a 3 person rate) and a few other awful things. More importantly for the time being, they decided that they weren’t going to even complete out the previously booked work, and that I would need to agree to the new pricing or not get any parts.
Well, this is not really what I wanted to hear, and spent the next month re-quoting, calling in favors, and otherwise wasting 100+ hours just trying to keep our beta program on track! Very cool behavior, unnamed company. You definitely deserve “business of the year” for that.
Next I went through long term pricing, getting competitive quotes, and, upon discovering that competitive quotes were a joke (Again, I’m told that the material cost for the material and supplier I selected for them was 3-4x the cost that I had gotten quoted the week before), went and looked at other options to produce our products in the USA. I then dove in to in-housing our own manufacturing, to be started with me learning how to make parts and then eventually handing off the process to a dedicated machinist, once our volumes got great enough. After searching for a bunch of different options, new, used, Mazak, Haas, Tsugami, Citizen, I eventually landed on the Tornos XT32, what I didn’t realize was a brand-new platform with limited installs around the world, but otherwise my course was set, I was under contract, the machine was already a “dealer buy” and had been in production for a month or two already, and delivery would be in May.
By this time, it was April, and it was time to finalize a loan for this machine, yet again spending 100’s of hours filling out loan documentation for SBA loans, large banks, small banks, local banks, all to wind up with a single loan approved for the amount I needed at a rate that was pretty spooky. I grit my teeth, and wound up signing my life away for the opportunity to make parts ourselves and to control our own destiny. As I’ve said before, this was purchased solely with Chube Compact in mind, but would also allow us to make Chube60 (we’ve already made heatsinks here) and any future hotends I could come up with (Pika!). This is honestly what allowed us to even produce Pika in the first place, as we would wind up paying external manufacturers in the USA more per part than we currently charge, with similar issues surrounding Compact’s current pricing. It was literally do it in house or close up shop, which wasn’t really an option.
That brings us to production in house, and why this month’s update is a bit more sparse on the production side than the last few months.
Since September, I’ve been sprinting hard with a sole focus on making parts to fulfill both the Pika Kickstarter, our reseller orders for Pika (Produced same batch/time), and then Chube Compact. I planned for this, and added in a tripling of time I predicted from review with local machining educators & professionals in the field for production.
Seriously. The total machine run time is just over a month for Pika of continuous operation, in steady state. When I manufactured the Pika blocks, after resolving the pilot drill+countersink situation, I was able to let it run continuously for over 24hrs in a single sprint, producing ~1440 blocks in that time, without me needing to touch it, only reviewing tool life. This is what I planned for, steady state production after a few days each part of setup, using tools that were selected for me by tooling experts based on the engineering documentation provided for these parts (models and eng drawings) and a properly configured machine from the get-go that has all of the proprietary parts from the machine tool builder that supports my applications and parts.
This was not what I was given, and unfortunately that’s meant that I’ve been unable to manufacture these components in the predicted timeframes, particularly when it comes to copper, as there is extremely limited knowledge in general in the machining spaces (as seen by instant breakage of turning tools, pilot drills, reamers, chipbreaker inserts, and custom reamers) despite some of these tools being specifically “designed” by the tool manufacturer to mitigate and address the peculiarities of the material & its specific alloy.
Here’s a few small examples of what I mean, and why this can completely destroy my predictions, with some of these being what I’ve mentioned before, starting with the “simplest” and starkest example of just how things can get off kilter
Expectation:
Pilot Drill is able to both create pilot hole for 20xD drill & countersink for filament lead in.
Reality:
Pilot Drill breaks after 1-5pcs, taking out $300+ dollar drill (brand dependent) that follows it
Attempted mitigations, in order:
- Install New Drills & verify offsets, & verify datasheets for speeds&feeds
- Verify Gcode for depth
- Slow down Feedrate
- Email Manufacturer’s rep of tooling with gcode, images of setup & problem
- Attempt to use 3x other pilot drills from different “series” from same manufacturer, some more easily adapted to copper than the others (like a titanium-series drill, which has faster/better speed&feeds for copper, since they’re somewhat similar for chipping behavior)
- Attempt recommended speed&feed adjustments from Manf rep
- Attempt recommended relocation of pilot drill to live tool
- Manufacturer of tooling visits, inspects setup. Confirms that my gcode, previous operations are correct, that I’m doing everything from a setup side correct (no novice errors). Suggests that my *brand new machining center* is broken in some way. (seriously)
- Write a custom program to drill 6 holes per part and just cut off a 5mm puck to test durability of pilot drills
- Attempt range of feed&speed adjustments ranging +/-50% from recommended, which is 2 axis, not one, burning through pilot drills like madness.
- Attempt to use alternate drills from different brands, but these do not provide a “perfect” pilot hole for guiding the long drill (as the pilot drill & long drill are matched, and there isn’t time for getting custom drills from china)
- Finally get frustrated, and decide to shorten the depth of the drill/forgo the countersink, and poof. Suddenly, no broken drills!!!!!!
Cost to timeline:
2.5wks for *just* this singular issue.
This isn’t the only situation I had run on this part, I had also addressed endmills not performing correctly & the boring bars instantly snapping when running anywhere from +/- 20% recommended speed&feed. Turns out, after pressing the manufacturer for 2 weeks for any help while also running our own tests, You need to run inconel speeds with our copper, which is ~12% of the speed and feed recommended for “copper alloy”, meaning there was essentially no reasonable way for me to have gotten to that point without sacrificing dozens of hours and parts spreading out the speed&feed from the starting copper alloy reference point on the datasheet.
Another simple example is getting custom reamers ordered for the internal geometry in the nozzles! -
Expectation:
Using provided Engineering drawings, including material choices (Even specific alloys!) Custom reamers will be designed and provided with supporting accessory tooling that will create the inner geometry for nozzles
Reality:
Despite only running reamers after confirming other parts of the process (confirming no crashes would occur, starting the reamer with a “high” offset to walk it into full depth, making sure previous operations were correct depth) both reamers broke after *first part* despite using highly conservative feeds&speeds provided by toolmaker.
Attempted Mitigations:
- Reviewing all possible countersinking options available on market, even with less aggressive taper angles, considering metric and imperial standard options
- Purchased wide range of parts and began testing of options, which includes gcode rewrites, cross-sectioning of each part produced and analyzed
- Cost to timeline: 3+ weeks, 100’s of hours of testing & iteration to get an industry-standard inner taper.
Cost to Timeline
Over a month due to holidays preventing tool arrival, testing out of options, and waiting for tool shipping.
Other fun issues
- $200/ea Small Diameter Drills were ordered by "experts" that didn't support usage in copper -manufacturer’s documentation recommends different drill series (also manf’d by same company) for both steel+copper use, meaning I had to A) learn and source my own B) source a new high speed spindle to drive them, driving extra lead time and learning how to run a coolant powered one!
- Machine wasn't configured with proper spindle adapters, meaning I couldn't load the tooling i was told to buy w/out getting new ones in! At one point, I was told that the parts would be ready in September, 4! months after machine delivery. They definitely improved their delivery time frame, but there were also key misses on required equipment in the sales process, meaning I had to wait for parts and bought them after!
- Threading tool wasn't provided with correct anvil for proper thread helix for metric thread
- High pressure lines weren't configured correctly for machine, even after a visit to the physical machine and review of the spec sheet, resulting in both incorrect fittings & dual female lines when the distribution block was already female.
So, despite my working pace including a 7 day workweek, I can’t work any faster than parts can arrive or vendors respond to me. This means that calendar delays add up!
However: I also have an existing business. At the beginning of 2025, I had recently hired a new full time employee (when I launched the kickstarter) and unfortunately, due to issues I highlight somewhat in the video, I’ve been unable to fully train and integrate as well as I had planned when first hiring. Despite this, He’s done admirably, but there are some deeper insights that I had been unable to truly provide support for due to my 2025 being sidetracked out the wazoo, resulting in several delayed other projects that were booked planning for XYZ schedule at ABC interval of work.
Here’s some graphics that hopefully help illustrate the problem.
The Plan, with some buffer time and slack to be flexed into problem areas:
The reality, after all slack time is devoured because of unknown unknowns
All that is said to say that this last month was primarily focused on catching up projects that had fallen by the wayside with my regular business, namely some printer projects, farm maintenance, organizational updates & training, and some components to help serve our existing B2B clients that lined up correctly with machine setups. I could no longer ignore these projects in favor of machining as they’re also needed to keep our lights on and employees paid. In addition, there has been time needed to train our newest employee, who has already been contributing heavily to operations here, bringing in skillsets previously unique to me in our organization & additional manpower to accomplish tasks. Here’s a video of nozzle inspections that was setup by him, reducing tasks that previously only I was able to do.
But, here’s the important part - Hardened Steel nozzles were set up in less than 10 iterations from beginning to end. An iteration in this case is a toolpath change, an offset adjustment, or a feed/speed adjustment to get a better cut. This is *faster* than I expected these nozzles to be setup, but its been aided largely in part by the work that I’ve done on previous nozzles, including writing macros and using variables instead of more hard-coding items, having pre-written programs for centering, applications written (http://machininghelp.app) for easy calculations, offset centering, and tool interpolation, and mainly steel being a well understood, well defined material that actually plays nicely and chips correctly instead of being a stringy gummy mess that breaks tools for breakfast.
If this result had been different, I’d be feeling worse than I already feel about this entire adventure, but this *validates* everything that I had been thinking when I started the month - The worst is over, and now time to go back to “easy” production with normally-running materials!
Based on my success with these nozzles, which still have a considerable amount of tricky-to-execute features like an internal taper, microholes, and high centering requirements, I expect the heatbreaks hardest feature, thin-wall “break” section, to be very simple and straightforward to execute, as on my machinery I’m able to do “pinch” turning as featured here by Titan’s of CNC. resulting in a very stable cut that will result in a very consistent process. From a timing perspective, this part has at most 3 min of cutting per part, and with a 12hr shift (I don’t want to run titanium unattended due to fire risk) this will only require *4* days of runtime before completion! This is why I’ve been so aggressive on timelines - When working with well understood materials (which I thought copper would be, based on what the tooling manufacturers & others were telling me) things work pretty much as they’re described. I even have drills that are designed for *only* this alloy of Titanium as its the most commonly used titanium in the world!
While this is being machined, all of our other components will be out for plating, and we are already fully through our inspection process for blocks & nozzles on Pika.
I’m happy to report in addition that we’ve installed an early-edition Pika that had internally-produced parts (Shout out to Coex3D for testing!) and its been merrily producing parts at 50% increased flowrates in their print farm! We take part quality and performance seriously, and we are still confident you will greatly enjoy your hardware once you receive it!
In conclusion - I'm sorry this has taken so long.
Now that the machine is happily humming behind me and I have a short opportunity to write this update and reflect and after recording several versions of the videos: over 600 hours of additional work on my end trace back directly to recommendations from industry experts, tool dealers, and machine manufacturers that turned out to be wrong - wrong tooling, wrong process flows, wrong machine configuration, which had to be corrected in order to proceed. To be clear, these aren't hours spent learning new skills (which I expected and planned for). These are hours spent undoing and reworking things that should have been right from the start, because I trusted the people whose literal careers are knowing what they’re saying. Several of these representatives have had more time in this career than I’ve been *alive* and yet couldn’t provide useful advice or solutions, despite my willingness to do whatever asked, buy new tooling (just sell me the tool!), buy new parts for machine (I’ll buy a high-speed spindle/software feature), etc. I didn’t expect this egregious of a mismatch between expected and delivered performance (eg, I expected to have to calibrate/tune feed/speed, not *entirely throw out the recipe and start over*)
That 600 hours also doesn't include calendar delays like the machine arriving months late, or waiting on tooling shipments after breaking yet another set of tools & their spares. The above is purely my time at the machine fixing problems I had no reason to anticipate.
Have I made my own mistakes? Absolutely, more than I predicted. But for context: I built my schedule with triple my expected learning curve hours and added weeks of buffer time beyond actual machine runtime. I bought spare parts & extra tool holders ahead of time to make sure any of my “oops” would be handled. The expert-driven overruns have dwarfed all of that, with the additional “bonus” of exhausting any kind of buffer I may have had in my personal energy levels, reserve funds (I have $10k in tooling that cannot be returned “per policy” despite it being sold to me as compatible), and time that I had with other projects that I’ve been forced to put off completion and not accept new projects. Sales through resellers also do not go directly to me, there is a certain percentage paid for preorders & the rest on delivery of Compacts.
I'll also share some fun kickstarter background as my last paragraphs here - I was backer #7934 on the Coolest Cooler, at the time the largest Kickstarter ever. I never got my cooler, which sucked, I was in college and that would have been baller for the summer. But I watched that project collapse because the creator kept expanding scope and chasing scale until the money ran out. That experience stuck with me.
I'm not scaling beyond what I can control. I'm doing this work myself, in-house (the only way for any of this to be possible), making sure every backer&preorder gets what they paid for, even if it takes longer than any of us wanted.
I also know that Kickstarter has changed since then; most big projects now are basically pre-funded product launches by established companies with millions already invested&spent before the campaign even goes live. That's shifted expectations around timelines in ways that don't really apply to a project like this. I did my best to set reasonable, safe timelines with what I knew, and promised I'd work my ass off to fulfill every backer while sharing the journey along the way.
Well, my timelines haven't been the best. But as I write this at 0424 on Sunday morning, I'm not going to stop working until every order is fulfilled, and excited for the future once ya’ll come back to get more cool tech. Thank you for those who've been with us since the beginning, and I can't wait for more in 2026 to show what we're truly capable of.
Other Cool thing I made this month, to catch up on some orders for Chube 60's that were due
Copper Meltzone adapters:
I learned two more things rare for machinists with these, segmented threading & broaching!
