Tetent TimerSpy [gd0136]

Since I wasn't able to print overnight, I thought I'd at least clean up the model (white oval) and cut away unoccupied material so that TimerSpy has a better chance of going over a sleeve (green oval). This has the added benefit of looking sleeker and curveless from the wearers perspective:

Talking about this angle, the diamond design idea stemmed from this japanese shrine paper thing:

I thought it would make the design look more calming yet modern with a paper-inspired look, and I believe it worked.

Render:

This is the design I've come up with. I haven't got any chain covers because Finger5 essentially determines were all the other fingers land anyway, and seeing shiny ball-chains sliding as I'm typing probably looks cooler too. I should also note that Finger2 would've actually failed if I didn't take that diamond chamfer across the entire length.

I've rotated the Tetrinsics 180 degrees and added a ball-chain-zone solid to make sure that it all stays within the case, which is not something that could be said for Attempt9.0. 

In other news, I wore Attempt9.0 with my coat and that's when the reality of moving the screen up came into play. Basically, the top left corner doesn't allow the coat sleeve to go over, and since I need a battery wire exit point, I can't just remove the rectangular diamond and leave the 3 square diamonds so that the sleeve can go under the screen. I then looked into what my battery and screen options are, and it's back to a 480px 60Hz screen and only a 3Ah battery (2 x 1.5Ah). I imagine it'll certainly feel more portable though. It would essentially be the highlighted section:

Since I'm mainly using timers and writing project logs on this device, I could probably manage the screen downgrade fine, but losing 7Ah for a device that has many reasons to use quite a bit of power -- just for an "I'd probably get used to it" inconvinience isn't something I'm willing to stomach. As expected, 10Ah has a lot of bargaining power.

The main issue is, once again, Finger5 and only Finger5. Perhaps 5mm is too low. Perhaps it's that further-back square causing issues. I'll see if a value of 3.3mm is any good, since that also works geometric-aesthetically.

It could also be that Tetrinsic 2.0 has too much of a "bottom bezel".

The good news is that's more or less the only issue. Other findings from the print:

• Moving the screen section up now means that my arm can lay flat on the side of my body when standing.
• The white part goes right over my forearm-end bone lump.
• The print certainly looks bulkier than Attempt8. 
  • A 2.x mm thickness increase in the white edges, 18mm longer total length and a suprisingly noticable bezel size increase are the reasons.
  • The bezel increase is due to the 70mm battey length. If I had concluded my battery search early and determined that the 6x67x89mm or 5.5x67x90mm 4.5Ah was the best available, I probably would opt for the 6x66x67mm 4Ah to keep things looking compact.
  • However, the battery seller I found sounds totally onboard with sending me a 7.3x70x95mm 10Ah battery for the same cost as 2 3.5mm ones. I won't miss a difference of 500mAh too much, but I'm not writing off the potential for an extra 6Ah without a fight. 
  • I am considering dropping down to 8Ah for a 70x90mm battery as the difference in CAD is noticable (mentioned further in the log) but that's a 2Ah cut for a 5mm total reduction. That's like an entire standard density 18650 there! 
    • Those "DIY yourself" sellers were the ones with the 5Ah 3.5x70x90mm batteries. 
    • The Bihuade 2 store only seems to have a 4Ah one.
    • Looking at my long list of project tasks, I'd rather not DIY everything here, and I feel like I'd have more issues with 2Ah less battery than 5mm (which I'd probably only really notice if put side-by-side)
• I can do push-ups without issue, but the edge does lightly tap on my skin occasionally. 
  • This is another reason for the 5mm reduction consideration. The total length in cad was set to 97mm when it should be 98mm (95mm battery + 2(1.2mm wall + 0.3mm tolerance))

Difference between 8Ah and 10Ah options:

I like the look of [yellow] when looking from the perspective seen in the above image, [red] is where my finger actually is (at least the angle is fine) and [green] is what's preventing motion of said finger. 

I hate it when I've got this nice looking design but then one of the fingers has an issue, so then I've got to think of another nice looking design. I'm wondering if I should move Finger4 down to the same level and angle as Finger5, but then I'd lose the ordered look of Finger2-4 seen above.

I also like the look of this face:

The edge to that face is also stopping Finger5.

Unfortunately, I can't continue with Tetrinsic until I've got a viable solution. I can get away with TestCut because it's larger and I'm much more likely to have a solution, especially if there's one for TimeSpy.

I was supposed to do this tomorow or something, but Tetrinsic and TimerSpy are things I've needed since 5 years ago! I'm currently printng this concept:

I've added covers to the unused side (color model shown further down) and hopefully made a printable version of the chain:

Still, that's a lot of retractions in there. 

On the product design side, I had a slope that went over Finger5. I think it looks waterfall-esque:

I then tried like 3 different cover designs and went with the first one:

I was thinking of having the wrist chamfer be 45 degrees so that it and the black lines were parallel.

I admit, from the 3rd person perspective, this looks saaaaatisfyingly ordered, almost like I found this concept on Pinterest. However, I have concerns about the forearm bone lump if the TimerSpy slides down. The main reason is because it doesn't look as good from a 1st person perspective:

I will admit that my opinion of one or the other is largely in favor for 3rd perspective and ever so slightly against for 1st, thus it's likely that the larger chamfer will return. I've already got a plan for the bone lump should it become an issue.

Oh right, I haven't mentioned why the chamfer looks steeper than in the previous log. This is why:

I'd also like to mention that the diamond colum strat isn't the first time I've used it. I first used it in #Teti [gd0022] 

[13:30] So I was going through the model timeline and then hit this hurdle: Finger5 is now lower than the main white surface. The main issue is that I need to ergonomically reach that Tetrinsic, but this face is also the second most seen behind the screen face.

Right now, I'm experimenting with using diamond columns instead of that half ( curve.

As I hoped, the columns look open and fresh. The connection to the front screen is iffy though. I can't just have a 4th square-diamond because I'm planning to route the power cable out of that small amount of space where the 2 white planes meet. 

This looked nice when looking from left to right up until the wider horizontal section meets the thinner vertical section.

Not sure about the look from this side, but I very much like how it looks from the top:

Going back to the first variation, I could fully connect the rectangles instead of stopping when I hit the top of the horz pane/slate:

Or perhaps a very large square is what's needed?

Ok and then a shallow cut inside?

Triangle?

Maybe it's a bit much compared to the ultra calm other side?:

This wall is also not helping with visuals:

Ok. How does the large square (no indent) and the southeast rectangle look on a desk?

Okay okay... Would be nicer if the large square lined up with the top face but it's promising

Mn. It looks unbalanced. Looks like I'm moving ahead with the large square or northeast rectangle.

[15:10] So I've gone in and redefined the main sketch so that everything lines up.

I'm thinking of a sweep here... a fillit there... some lofts too. It's kind of hard to 3D sketch with only 2 dimensions.

Or maybe, a simple slope?

Nah. This makes the Timespy look bigger than the first but more complex idea.

[15:40] I've hit another hurdle. I wanted to curve that chamfered edge so that I could have the rectangular ended face shown earlier in the log, but I also want to keep these two edges parallel:

There's nothing that actually needs to be in that chamfer area. It's just that this point (below) looks kind of sharp:

I guess I could chamfer it:

 One would rarely see Timespy from this angle anyway. But now I'm wondering if I should double down on the sharp+crisp edge look?

So close.

Hm that hasn't worked like I hoped it would.

[17:40] I've had a bit of a break, looked back and have done this for the design:

I'm wondering if I should drop the draft on the 2 diamonds that cross through Finger5:

I think a middle ground (that is, a smaller draft cut) is the best choice.

I'll do the Tetrinsic top covers (to prevent sliding the wrong chain) at another time. 

I just wanted to make sure that an aesthetic solution existed before going back to developing Tetrinsic, which I'm not making in a vacuum but for use in Tetent projects. Therefore, I want to know everything works for a concept before spending days working on a solution.

As you might have seen in the previous log, I've started experimenting with variations of a slightly new name that orbits around "Timer Spy". This is because I'd kind-of want to avoid sharing a name with a very popular benchmark, but mainly because this device would be used to plan and organise my life (I think that's what PDA's did back in the day) and I'd likely be using timers such as #Interval Provisional [gd0097] to get things done.

Current candidates:

• Timerspy / TimerSpy
• Timerspi / TimerSpi / TimerSPI / Timer·spi
• TimrSpy / TimrSpi / TymrSpi / TymrSpy (cutting out the e like in "SecSavr")
• TymerSpy (like "TyMist")
• Tym-r-spy / Tym·r·spy / Tym·r·spi / Tim·r·spi Tyme·r·spye
• Ty(mer)spy / Ty·mer·spy / Ti·mer·spi

Grrr I can't seem to find a solution that looks nice in text. Maybe TySpy -> Tymer Spy the same that TyMist -> Transparency Mist?

[Jan 01, 2023] I might just refer to it as TiSpi or TiSpy.

[28 Dec, 00:15] Ok. I've got a plan. With the shorter Tetrinsics, I can move the CB1 under them. Then, if I increase the screen thickness to 10mm, I have 6.7mm of space (not considering the strap):

Now to dig through AliExpress to find a--

Starts playing celebratory music.

Ok, can we take things even further?

It's going to be a tight fit, but it seems I could get back to parity with the 21700 cell. However, considering that I also need to route out the cables, I think the 4.5Ah is the best I can do.

There's actually a 5.5mm thick battery that has 4.5Ah too, so I could stick with 9mm thickness.

Then again, the shortest length of the front screen panel is only 80mm. I'd likely need to go to 93mm to fit these. If I can't manage that 13mm size increase, I'd likely be limited to 4Ah.

I did notice this 3-wire battery that was only 4mm thick. Makes me wonder why I haven't seen 8mm, 10Ah batteries:

There's an even thinner variant:

And there are some dual-pouch batteries:

I could always ask the seller if he can stick 2 together vertically.

Still though, it's looking like I'd need a longer battery pack area. But that would hinder ergonomics and aesthetic.

Or will it?

I could make the white thickness 11mm and extrude it forward by 18mm. Now the actual thickness is better hidden from this angle, saving aesthetics, and the screen appears higher up on the arm, improving ergonomics.

I wonder why this battery size is so popular, or has comparitively higher density than others. Actually, it might be because it's so popular that more recent battery chemistry has been used. At any case it totally sounds like, with the correct protection circuitry and some marginally thicker bezels, I could have an 8 or 10Ah power capacity, doubling what I could've achieved with 21700 cells. 

Worst-case seems to be 4Ah, which is decent. It should mean that the overal Timerspy dimensions are primarily display-size determined.

Thankfully, it seems that Tetrinsic 2.0 is GO.

[08:15] So far, I've asked 3 LiPo battery sellers on AliExpress and 2 said "Please DIY yourself". The first seller I found in this project log said that the total thickness will be 7.3mm for a 70x95mm. That's in my space budget! 

10Ah could actually be in the cards! I think that's a worthwhile compromise for a non-removeable battery. For reference, considering the rather large size of TimerSpy, this is what I've had in mind regarding battery capacity:

• 2.7Ah: Poor (same size as the LEM-T, the largest capacity Android smartwatch)
• 3.4Ah: Acceptable (highest capacity 18650)
• 4Ah: Decent
• 4.8Ah: Great! (also around the highest capacity 21700)
• 6Ah: Amazing! (unlikely, but if the opportunity presents itself, take it!)

I need to get this concept printed ASAP to see if a 10Ah solution could actually succeed ergonomically.

I also know that a 3.5x70x90mm cell exists if a 98mm TimerSPI is too much (but a 93mm one is fine). However, I'd have to take 8Ah (which is stll way more than I thought I could ask for) or DIY.

[20:15] Now, as I feared, the Tetrinsics are sticking out quite a bit. The good news though is that I can have 19mm spacing for Finger2-4 and a 20mm spacing with 4 degree tilt on Finger5. Finger5 is also 4mm lower than the others.

That battery is currently on thin ice. On one hand, I don't really want to give up a 4.8Ah battery. However, I'd like the battery to be easily replacable anyway so I could always bring a spare with me.

If I can't find a suitable battery at all, the Forced Solution Solve I sent from Tetrinsic might be reflected back. This is a description of what I'm talking about between Me In The Past and Me Sometime Later:

1. Solve a solution for Tetrinsic 2.0 with TFT.
   1. Solution solved with 1 warning: It's almost 53mm long. The Timespy solution may fail.
2. Send a solution resolve request to Timespy.
   1. A solution to increase the length of Timespy... failed, due to exceeding a maximum solution limit of 70mm set by Me In The Past. Reason: Ergonomics.
   2. Attempting to discover a new battery solution. (<--I am here)
   3. A solution... fails. (<-- what happens if I can't find a decent battery and don't want to connect umbilical power cords so that Timespy doesn't run out after 3 hours)
   4. A forced solution solve... fails. The new solution limit for Tetrinsic is <43mm.
3. Send a solution request to Tetrinsic 2.0 with TFT.
   1. A solution... fails, due to geometric constraints.
   2. The solution... is rejected. (<-- I don't want this. Tetrinsic with a digitally adjustable aesthetic looks really nice, even if I'd only be able to see it when I'm not using it.)
4. Solve a solution for Tetrinsic 2.0 with white LED backlight. (<-- This is what's likely to happen next...)
   1. A solution... fails, due to geometric constraints:

      That wasn't supposed to happen.

5. Solve a solution for Tetrinsic 2.0 with backlight and single track. (<--Ok I've just skipped over and I'm here now. I need to see where this is going because it doesn't sound good!)
   1. A simulation produces a near-failing warning: Ergonomic and grip concerns issued. There is a very noticable difference in feel and confidence compared to 2 tracks.
   2. Thus, a solution... may fail.
6. Recompute a simulation with real world simulation metrics.
   1. A simulation... fails, in part due to walking vibrations causing loss of grip; this increases typos and thus reduces thought flow to a solution failing level. Additionally, finger pressure on a single ball became ergonomically uncomfortable after approximately 1 minute.
   2. (Auhh nouh say it isn't souh!)
7. [Insert more back and forth solution reflections and recomputations]
   1. Due to time constraints, 2 solutions cannot be persued. Additionally, due to Timespy having a higher solution coverage (how many current and future problems the solution mitigates or eliminates) compared to Teti, Tetrinsic 2.0... fails.

Wow. Who knew that this project, which started off as a "Mn... might as well see what I could do..." has higher priority than my main PC. I guess this is the Steam Deck Effect, where some users report using their £349 not-a-console over their £2,349 gaming PC.

Anyway, it seems that finding a suitable battery has much higher stakes than I originally thought. No pressure, Me In The Recent Future.

Every so often, I wear the concept to see if there are any ergonomic issues, and the only main one is that he bottom left edge of the screen occassionaly (< 3 times a day) scratches my other arm. I don't really want to give up the 90Hz, large, high definition screen just because of that, and reducing the visual apeal slightly to improve ergonomics. 

Then again, it does look a bit like a japanese wood sandal now, which I think look nice.

I don't think the new design would improve or impair the ability to get it through a sleeve.

If you're wondering what those Dynamic Islands on the Tetrinsics are, read this log.

Whilst I was doing research in HDMI -> MIPI, I wanted to see if there was a way to get USB-C Alternative Mode video out from HDMI too, since the CB1 has 2 HDMI ports. It wasn't long before I went through a hyperlink into this PDF about the ANX7688:

Cool. That sounds great. Is it on AliExpress?

Noh... It couldn't be... Is there another chip with the same name?

Unreal engine!

Anyway, it wasn't long before I found https://xnux.eu/devices/feature/anx7688.html, where it sounded like there was some linux drivers and firmware written for it. It's featured in a device called the PinePhone, and it turns out that there's a schematic for it! The chip in question is on page 13, so there could be a chance that I don't even need to read a datasheet to implement it! Truly exciting.