#42

Some random things I learned making the TELEX (hope they are helpful):

  • When you are done with your design and have validated your prototypes, you are about a third of the way done.

  • When you have volume sourced all of your parts and had things like panels and PCBs custom fabbed, you are nearly halfway there.

  • When you have built and validated all of your units, you are roughly three-quarters of the way there.

  • When you have written your manuals, boxed your units, collected funds, shipped, and validated safe arrival, you are totally 99% of the way there.

  • As you support and manage the lifecycle of your unit, you are asymptotically approaching “done”.

–

  • If you change manufacturing technique, figure out how to make a proof before committing in volume. I 3D printed, then laser cut my prototype panels, and machined the final ones. Due to differences in tolerances and techniques, I ended up having to drill out a small shaving from every LED and jack on the finals so that things would fit. Had I only known, two minutes in illustrator would have saved endless hours and rescued my hands from serious torture.

  • Expect to spend some time buying and/or building tools to help you build in bulk. My reflow oven (Control LEO) and desoldering gun (Hakko FR300) were essential! And, while the assembly and stencil frames that I made with my 3D printer were invaluable, I also saved a ton of time by creating PDFs with a page for each part that clearly highlighted all the places it went. My resulting build success rate was near 100% - which more than justified the time and cost.

  • Allocate into your costs and build plan units for public testing and units to give to those who helped you along the way. Also, allocate extra units for in-field swap-outs. While I haven’t needed any of the latter, I sleep easier knowing that they are on hand.

  • When costing, don’t underestimate how much the little things contribute (like anti-static bags, packing materials, rack screws, plastic bags to hold the rack screws, boxes, shipping labels, printing documentation, power connectors, …). Adding this stuff in (especially when not purchased at serious bulk) put a lot of cost into my project that I hadn’t accounted for in earlier estimates.

  • If you have a Teletype, don’t just think of it as a music-making module, think of it as a scriptable testing machine! I saved an enormous amount of time by writing a few small testing scripts that allowed me to fully QA my builds. I dedicated a few LEDs on the TT to indicate PASS/FAIL conditions. It was amazing. I’ve seen others engineer and build complex test rigs that only accomplish a small percentage of what a few lines of TT scripting can.

  • I want to second comments made by @TomWhitwell in suggesting starting “small and simple” - but I didn’t do that either. For some of us, probably the hardest advice to follow. When you don’t follow it, the follow-up advice is to simply keep at it. I hit several points of pure despair that could have ended my project. But, with a little time off and the help of others, they were ultimately surpassed and mostly forgotten.

  • As for the utility and aesthetics of your product, this is really up to you. The relevance of various pieces of advice wholly depends on a shared context with that advice. How much are the modules an extension of your art? Are they the beginning of your new corporate empire or simply a break-even labor of love? What audience are you going after? Are they designed to fit in with an existing series from another manufacturer? Why are you doing this in the first place? Etc.

  • Seriously consider an open source approach. Not only does the practice of documenting and sharing your work product help the overall quality of your product (something that I’ve heard Oliver at Mutable say in interviews as well), it opens you up for help and collaboration that wouldn’t have been available otherwise. I know that this isn’t right for everyone or every situation - but it was a godsend for me. Plus, I had personally learned so much from the openness of others that it seemed wrong to do it any other way.

  • Allocate time in your schedule for unplanned breaks along the way. Whether it is a work/family emergency or you just need to take a few and let your brain repair itself, give yourself license to do it. This helps more than you can imagine.

Good luck!!! :slight_smile:

19 Likes
Telex: Teletype expanders
#43

Thanks for sharing this. And while I’m here: I’ve re-read the Telex thread several times, and I don’t even have a Teletype; just a sucker for design and process. Although…

… maybe I should get one.

2 Likes
#44

i have to kindly disagree with this point!

support only grows with the more things you build. my “legacy” support load is not insubstantial, to the point where it actually inhibits that sense of freedom to just design new things constantly. i am incredibly aware of the perpetual support ramifications that come with any new design-- and if i’d known this sooner, i would likely have made some different decisions as to what got made and what did not.

also note that this community has been phenomenal and completely enabling, given the amount of support provided by kind people here that i would otherwise be directly fielding. and then doubly amazing that everyone has been patient and understanding during my lapses in quality support.

beyond that, outstanding post!

17 Likes
#45

Just got the box art back from Cath, and amazing illustrator who happens to work across the office from me.

jackalope greyscale
jackalope greyscale.jpg2472×3272 1.31 MB

Things are coming together.

9 Likes
#46

20 characters of beautiful!

#47

FWIW, the “jackalope” is real: rabbits that contract a certain sort of papillomavirus are known to develop keratinaceous growths on their heads that strongly resemble antlers. See esp. this very well-written extract from Carl Zimmer’s book A Planet of Viruses.

3 Likes
#48

@papernoise just out of curiosity, how does width come about when your designing panels and layout? Do you have like varying versions (10hp vs 12hp) and then weigh the pros and cons of each?

#49

Yes that is totally the case.
Sometimes we start big and then realize we can make it smaller. Or we start small and then it just won’t fit in there.
Depends a lot on the project and at what stage I get in there.
There’s also cases when I – or who designs the module – already know what size it should be. There’s modules where you know exactly that’s got to be a 10HP one, 2HP more and it will feel to bulky. 2Hp less and it will feel crammed.
Then sometimes there’s modules that feel like they should be that big, but the controls just won’t fit .
It’s sometimes a long and windy road to find the right balance.

1 Like