Rotary Un-Smartphone Shipping Update

[Justine Haupt]

I’ve been doing electronics production (printed circuit board 
assembly/soldering/programming/quality checks) since the beginning of 
the month. All other parts are in hand and ready to be sent out. In 
parallel, I’ve been doing test-builds and quality checking along the way.

Okaaaaaay. So.

*
The rotary dials aren’t dialing consistently. *I thought (was sure) I 
had this problem worked out, but now about 10% of the time, when I use 
the dial, it returns a wrong number. At times its very reliable (owing 
to software improvements, which I previously thought made good dialing 
the rule rather than the exception), but on the two kits I’ve assembled 
from production parts so far, the reliability is, frankly, terrible. 
It’s unusable for real-world use. I can’t send these out.

This is a a knots-in-my-stomach-oh-my-god-what-the-hell-am-I-going-to-do 
type of situation, considering that with both old and recent pre-orders, 
about 770 kits are promised to be sent out this month (or at least by 
Christmas TBH).

After more tweaking, both mechanically and in software, I’ve resigned 
myself that a hardware design change is the only solution, meaning all 
the phone motherboards I’ve produced so far need to be scrapped. As of 
two days ago (Nov. 12^th ) I redesigned the motherboards and will not be 
using any that I have on hand. The daughterboards are also affected, and 
those too are being scrapped for a fixed version.

TECHNICAL PARAGRAPH, OK TO SKIP. The way the I made the rotary 
mechanism, pulses (numbers) are counted both as the dial is wound and on 
the return, so the final count is divided by 2 to get the intended 
number. This allowed a significant mechanical simplification, but, on 
the windup, those pulses often go too fast to get counted accurately. 
The software fix I had in place used the fact that the total number of 
pulses always needs to be even to be valid, as well as incorporating 
timing for debounce detection, but this just isn’t good enough. My fix 
is to add a hall sensor to detect when the finger stop is pushed to its 
limit (while dialing), so that the software knows only to count pulses 
after the finger stop is released.

I’m trying to figure out if I can still get 700+ *working* kits in the 
mail in time to arrive (let’s say) by December 18^th . If I overnight 
them and take the loss on shipping, which I will absolutely do, then 
that gives me about 31 days from nowto finish the electronics for all 
orders accepted until I moved the promised ship-by date on the website 
(which was Sept. 14 at 6:15PM EDT).Producing the new boards should take 
18 days end to end. A few days will be needed for programming, kitting 
and mailing of course, so if I have all new blank PCBs from my 
fabrication house in hand by November 28^th , this can still work.

To air the rest of the rest of the dirty laundry, there are a couple 
other (less critical) problems I’d like to make public:

-The headset jack: The auxiliary TRRS jack only work to hear, not to 
speak through a headset microphone. The microphone built into the phone 
works as a speakerphone mic, so this could be used with headphones in 
that sense, but proper “talk/listen” headset functionality isn’t working 
on this version. This will NOT be fixed until a future production batch.

-Battery life: It seems that right now the real-world battery life is 
only about 12 hours. The issue isn’t the battery running out. It’s that 
the system browns out when the cellular modem goes into call-receive 
mode while other call-alert features are also working. My temporary 
workaround is to have it consider itself out of battery when the battery 
is down to about 50% capacity. The battery capacity is there but the 
instantaneous current draw is too high. I may be able to figure out a 
software fix for this before any units ship.

IFyou’re done with all this and would like a refund, please email 
info at skysedge.com with “refund” in the subject.

IF you don’t particularly care if your order arrives for the holidays, 
please email info at skysedge.com with “no rush” in the subject to let me 
know. This would help me meet the holiday demand.

IF you /absolutely/need it for the holidays, please email 
info at skysedge.com with “RUSH” in the subject. Please don’t hesitate to 
let me know if this is your situation.

*A paragraph I was on the fence about including:*

Lastly, I just want to say to anyone who might start to feel like I’ve 
taken all these pre-orders (many from about 2 years ago now!) on 
anything but honest pretenses, I want to say that this hasn’t been a 
process I’ve taken lightly or that has gone without impact to my 
personal life, or my family’s. I’ve wanted to have a stable and 
successful company for some time and in that sense this is a dream come 
true. However, to make it work I’ve had to reduce my time at a career I 
was previously passionate about as much as they would let me without 
losing health insurance, while also balancing caring for an ongoing 
critical health issue for a close family member that arose last year. 
And then there’s the chip shortage, which is still not resolved for many 
of my parts. I’m not one to lean on excuses, but, it hasn’t been cake a 
walk. No one, /no one/, wants to have phones in customers hands more 
than me. The idea that we’ll be at the point in the next couple months 
where every single day isn’t overwhelmingly stressful for me… it almost 
sounds to nice to be real. But I’m confident about it.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://skysedge.com/pipermail/rotaryun-smartphone_skysedge.com/attachments/20221115/37e3ce82/attachment.html>