• 04/22/22 - More Space Woes, and USB Awesome

    Daniel Grace • 04/22/2022 at 17:26 • 0 comments

    I printed a PLA version of the mini afterburner just to inspect the dimensionality in my hand (PIF parts are still coming, I won't USE the PLA version), and I realized there's a different place I can mount the board that is both better and larger.

    ...

    But it's still ridiculously tiny. Pictured is the board, the main chips that are needed on the board, the connectors needed (more on that in a moment). Not pictured are any capacitors, resistors, actual routing lines, crystals, etc.

    It's not going to fit. There's no avoiding that, I have to find a different way. Or only support the larger vorons, but that makes me sad.

    In better news, I have been lamenting that I need power delivery (of either 48 or 24 volts, more on that in a future log) and USB. That's at least four cables, and it'd have to be either a custom cable or two cables. Ick!

    I just ran across a newer USB spec that is apparently used for things like MacBooks. I knew that devices negotiated with hosts already for things like speed and current up to 500mA (from the original default of 100). I didn't know that the laptop charging stuff had gone so insane. You can negotiate all the way up to 48V, in 100mV increments. You can also draw up to 240W. Now, that adds cost with very specialized USB chips, and I'm not even sure how many chips like that are available to someone as small as me.

    But the dream of having just a USB cable to connect the hot end to the main board, AND having it not immediately blow up your device if you're dumb enough to plug your phone in ... is feasible??

    https://en.wikipedia.org/wiki/USB_hardware#USB_Power_Delivery

  • 04/21/22 - Space is TIGHT

    Daniel Grace • 04/21/2022 at 18:22 • 0 comments

    I knew the v0.1 was small. I'm starting to try and consolidate my experiments down into a single board, and finalize the aesthetic parts and ... woof.

    These are the mounting holes for the front of the v0.1 shroud. The fancy Voron logo is 20mm from side to side. That's what I consider to be the minimum to still be readable. And it takes up almost the entire space I would have if this is where I mounted the board.

    ...

    Time to figure out a different location to mount. That's difficult since I ideally want it to be usable by someone who did PIF and cannot modify the printed parts before using the board.

  • 04/19/22 - Components And Minor Progress

    Daniel Grace • 04/20/2022 at 03:51 • 0 comments

    Minor progress. I've been doing competitive research, figuring out what I want to include in the board, how many SKUs I want, etc. I am mostly waiting on things to arrive before I can make any hard progress.

    To try and update more often, and make the updates less dense on occasion, here's a photo of a reel of 27 ohm resistors, and my current progress on making a spool holder for them to store them efficiently. 27 ohm resistors are a very common value for USB connectors, so this is one of the components that I jumped the gun and went straight to buying a full reel (10,000 pieces!).

  • 04/12/2022 - Pricing Rant

    Daniel Grace • 04/12/2022 at 21:25 • 0 comments

    I just sent off revision one of what may become this board, and I wanted to take the opportunity to talk about some of my goals, especially as it relates to pricing and customization.

    Goal A: Quality

    I will never compete with China playing their game. For that reason, I am uninterested in trying to compete at being the lowest price. Even trying to play the game of "I'm not the lowest price period, but I'm the lowest price with THIS niche feature!" isn't a thing I want to do. If I ever got successful, China would come in and make it cheaper. The world is just set up to give them an advantage in that game.

    I CAN, in theory, compete on quality.

    That said, I have been poor before and am not the type of person to spend money foolishly. I don't expect my customers to be that type of person, either. And I realize that in this pandemic, there are lot of people who have had pay cuts (actual, or effective), and are having to spend money on far more important things than a niche within a niche.

    All that to say, I want to find a sweet spot in price vs quality, rather than be the absolute cheapest or the absolute highest quality possible no matter the price.

    Goal B: Flexibility / Customizability

    My first product will (probably) be a control board that sits on your gantry. There are benefits to this setup (wiring being a big one), but also downsides. Not everyone has the same gantry setup. Do you have a BL Touch? Some other type of sensor? Klicky probe? How many fans do you have? Do you have a Mosquito Magnum+ with TWO thermistors and TWO heater elements?

    So naturally, the idea of customization comes in. But customization actually costs a LOT of money. Not necessarily in parts costs, but in process costs. I find that most people underestimate process costs.

    Goal C: Price

    While price is at the bottom of my list, it's still there. If I can make a decision that doesn't impact quality or flexibility, but is cheaper, I will do that 100% of the time. Really, I think most companies would as well, but my intention is to keep as consistent of a profit margin as I can, and as I find more ways to save money over time (economies of scale, for instance, if this is successful), the price will naturally come down.

    Interesting Aside: Economies of Scale

    To hopefully add some interesting meet to this log, I want to drive home how important economies of scale are. Many components need power filtering capacitors. Often multiple of them. Often there's a range as to what value is appropriate to use, but 0.1uf is often an acceptable value. It can make a huge difference to standardize on a value.

    If I were to buy my chosen brand of capacitor at that value one at a time, they would cost $0.10 each. Currently, I have 13 on a board (this number will likely go up), so just that capacitor would cost $1.30 per board if I bought them one at a time.

    If I buy 10 at a time, they cost $0.011 per, so 13 would be $0.143. That's already a HUGE difference, and we're not even buying a whole board at a time.

    If I buy 50 at a time, it's $0.078 per board.

    If I buy 100 at a time, it's $0.0637 per board.

    There are quite a few more price breaks, but if I buy an entire manufacturer box worth, it's $0.02704 per board.

    From $1.10 to about $0.03 per board for this part simply by buying in bulk (~97% savings!). The prices can even get smaller if I buy multiple boxes at a time, but a manfucturer-box is a convenient stopping point.

    You better believe that I'm trying to standardize component sizes as much as possible.

    Unfortunately, things don't scale quite as crazy for the more expensive components that could be on the board. The TMC5160 that MIGHT be on the board costs $6.81 for a single, and a manufacturer box only gets that down to $4.05. (Also since they are so expensive to begin with, a manufacturer box costs $8,100 to order, and I eat that cost if they don't sell. The capacitor is more broadly applicable to other designs, and the entire manufacturer box is only $20.80,...

    Read more »