Research

(Diary-entry post deleted in favor of more rambling about writing.)


So, research. I find research fun. Sometimes more fun than doing the book. There's at least two phases of it, and that first phase -- the background and familiarization phase -- is more like what you'd do to unwind than it is like work. That's the part of just looking at a lot of materials -- popular-press and fictionalized are just as good here -- to get a general sense of the material.

It's great because you can do this sort of stuff in the background, immersing yourself in the material while eating, at work, getting ready for bed, etc.

The final -- loosely put -- phase is the least fun, and that's when in the middle of typing up a scene you have to pause and go back over your collated notes and bookmarks and whatever to put in the correct spelling of the Pharaoh's name and the correct dates of her reign.

The latter is one of the reasons why it can be helpful to write up your research; to put together a monograph (or whatever you want to call a distillation of what you've found). Another big reason is that organizing it and writing an essay on it helps you to better grasp and remember it. Which is why students have to write so many of the things).



But what I really wanted to talk about is the pleasures of applying that research.

At the lowest level, there's knowing you are getting the names and dates right. I'm tempted to call this Wikipedia-level stuff, except that Wikipedia has gotten so crazy detailed over the last few years someone needs to step in and fork out a more entry-level approach for the general audience.

Better than this is four-senses stuff. It isn't quite as simple to find out what the colors or the sounds, much less the scents and tastes are (well, depending on what you are researching!) This is stuff that draws a reader into a setting much more than just naming the city.

More work yet -- and more payoff -- is a layer I'd call functionality. I could also call it the place where the nouns live. In your outline framework your character might sheath their weapon or hail a cab or button a shirt. When you've done the research, you know that you don't sheath an axe, the better casual transport option in Bangkok is the Tuk-Tuk, and there's no buttons on a wetsuit.

This isn't about errors, per se. This is about how the minutia of picking up and purchasing, strolling or hiding, or whatever functional things need to be done to get the character through the stages of the plot, unfold differently in different fields, with different tools, in different places. "Buy a burger" is a cheap way of covering, "in scene six the protagonist eats something"; good research will option up the option to have them "Haggle over a kebab at a street stall."

Better than that are the things writer Even S. Connel called "pickled plums." These are the bits that are so specific and yet so illuminating, things that could only have happened in the specific circumstances you are researching about. During one early Arctic expedition, it was so cold out that the moment the diarist (Cherry-Girard, I believe) stuck his head out of the heated tent his outer clothing froze "instantly," trapping his head in a screwed-around position, and he had to struggle through the whole morning like that.

My own pinpoint observation of extreme cold came on a glacier in Alaska when the moisture on my eyelashes froze. It was a quietly alien moment, gently astonishing as so far outside one's usual experience. I even thought I heard faint chimes as the bits of ice rang off each other with each blink (but I was almost certainly imagining that).

As delightful as the plums are, above that is where you are able to resynthesize. In all the other stuff, you are basically regurgitating. You've found a description that said "The color of fried tomatoes" or "a short walk from the pier." The best fun is when you've done enough research where you can come up with your own description from scratch. Where you can apply your own emotional response and parallels from your own experience to approach describing the thing.

And, also, when your understanding of it is so organic you are no longer dropping in the correct nouns and adjectives, but where the entire scene is shaped by the specifics of a thing or task or environment or character.

In short, 180 from the stereotypical "hollywood" approach (aka script first, then distort the real world to fit the existing plot).




Oh, yeah. diary stuff. I'm increasingly excited about composing again, getting in my instrument practice and music theory studies, but I need to set up, before I set up I need to clean, before I clean I need to put the holocron project to bed. This afternoon is about soldering up another set of boards, and I have hopes that I'll be able to keep focus long enough to get some detail painting in as well; two tasks my health woes and work schedule did not permit.

We'll make it your way

I work at a company that is essentially using Just-in-time manufacturing technique. Our product line is nimble and always changing, and our catalog is deep, so ww essentially build each order as it comes in. And as a necessary adjunct to this, we only keep enough stock on hand to fill known orders.

That's basically what I was doing. However, the lead time on acrylic, PCB boards, small parts, etc. is enough that I had to anticipate Holocron orders somewhat and purchase those materials ahead. Unfortunately, I guessed wrong. It looks like I'm going to be stuck with unused parts. Whilst, simultaneously, being short of what I need to finish the remaining orders.

The problem is largely one of changes.

Kanan in D

The Kanan Jarrus (otherwise known as the Caleb Dume) prototype is built. So far everything works -- in fact, the parts line up better than I had any expectation they would.

Well, actually...since the outer edge is opaque, I could maybe work out a way to open the center to where it really should be (essentially this style of holocron looks like six corner pieces just touching. Which is hard to do and have any structure, and harder yet when I'm trying to hide the overlap of the inner acrylic pieces.) But I've committed to working with the material this way, the designs are already tested and approved and the first six shipped out to customers.


I also had time to re-do the "temple" shell design for the new USB and other tweaks I've made over the development of the "stolen" shells. Cut one out already and will glue it up and see how it works.

(I need to run off a set of "Guardian" diffusion, and maybe a new set of "Gallifrey" cutout pieces as well, to properly make a new "Temple" prototype. I'm also temped to paint it silver and use green for the internal layers).




So I'm one step closer to finishing off the Holocron project and moving on to the next. And apropos of next projects, I made a discovery that might get my keyboard back up and running faster than expected:

SamplerBox

It's a software sampler wrapped in a custom-patched Raspberry Pi install. It's designed as exactly what I need from it; a no-frills Pi that simply boots up into a stand-alone sampler. Although...most people add an external DAC as the built-in one on the Pi is a bit noisy, and one enterprising programmer managed to get Freeverb running as well.

But the really unexpected and lovely thing was: I downloaded the disk image, threw it onto a microSD with the freeware utility program Etcher, and.....it ran. Right out of the box, first time, no tinkering.

For an acid test, I threw my Behringer controller and a pair of USB headphones at it, and it adapted without hesitation. Playing with low noise and latency and plenty of polyphony, too.

Brilliant

Another minimalist show with my two Elation dimmer packs and a $100 American DJ lighting board. One of the specials is the clip light from my work bench with a 75W halogen spot in it. It's like that. My main problem with the show is having to run the board myself, which adds another 10 hours to an already crowded week. Work was busy enough I came in on what was going to be a day off.

The calf muscle is healed enough to attempt the gym again. And I somehow found time for a checkout class on a more powerful laser cutter, and cut some more Holocron parts (on May the 4th, of course. How appropriate). So I'm pretty tired and needed the weekend.

I was worried about the 1.3 revision of the Holocron lighting circuit until I tracked the problem down to the USB detector loop I'd added. After omitting that one resistor from the board it passes all tests flawlessly.

Totally sensible, then, to stop everything while I work on the software. Cue wasting most of Saturday staring at pages of code trying to make sense of them, instead of putting Holocrons in boxes to ship out to my customers.

What is unexpected is I actually got it working. I have the usual scheme of a resistor ladder connected to a bunch of tiny buttons, so an analogRead() with the right constraints will detect which button is being depressed. The fun part, though, was working up a set of EEPROM.write() and .read() lines.

And it all works. Even the detect-on-startup. How it works is this; in setup() the code reads the value for one of the variables off the EEPROM. That variable is loaded once and only once by testing for a flag also written into EEPROM.

If the INC and DEC user buttons are pressed the variable is increased or decreased, and if SEL is pressed the new value is written back into EEPROM so the new value for the variable will be set each time the circuit boots.

As a last trick, if the user holds down SEL during the boot process, the software loads the "factory" value back into the variable.

This has all taken about a third of my remaining program memory, though (the Neopixel libraries are PRAM hogs), so the number of user variables I can make accessible is limited. Fortunately, the one I really need for this project is the ability to tweak the sensitivity of the capacitance sensor (a variable known as midiclorianCount).



My last laser session I spent mostly tweaking two of the parts sets, making them fit tighter and look nicer. The next big task that isn't putting finished kits in boxes, though, is to draft up a new vector engraved layer for a variation many of my customers have been calling for; what I've been calling the Kanan Jarrus holocron, as shown in Star Wars Rebels.

So the nightmare is almost over. One or two more shows, another 20 holocrons to laser out and ship, and I'll be able to concentrate on my day job, working on new creative stuff, and getting some sleep.

Oh, yes, and getting more practice time in on my 1967 Roth-adjusted E.R. Pfretzschner Model 301.

Shell Game

Well, I got my day off. Didn't feel up for work today.

Holocron serials #02 and 03 are moving along. I had to paint up a new shell in order to record the steps on my revamped mini-studio (a piece of blue paper tacked to the wall and running under the subject as a no-horizon backdrop, and a high-CRI fluorescent bulb in a clip light).

So I've posted up the first sheet of assembly instructions:

And the first payment is in on a successfully shipped Holocron. No word yet on the one I sent to Japan, though.

In any case, there's an annoyingly short list of steps before the first kits can ship. Annoying because it is just so difficult to work in enough constructive hours to actually do those last couple of steps. I really feel I have to finish one more of the "complete" Holocrons both to make sure all the current parts are cut to the right length, and to record the steps properly for the remainder of the assembly instructions.

As far as I can tell there's just a few wires left to solder (USB breakout board and the tail for the sensor lead). And maybe include a few pre-cut pieces of double-sided tape. And figure out how to wrap everything so it is properly protected in shipment. And I think I have enough parts cut for four kits.

Oh, and make sure I'm uploading the latest version of the software. And doing a complete function check of the assembled electronics module.


Unfortunately I still have rent paperwork to do today. And I should really finish those taxes. And this cough is distracting me enough it is hard to concentrate on anything technical. After all, if I felt well enough to program and solder and sand and paint and record shipping data I'd be at work being paid for my time. I mean, more than $90 a kit.

Vibrations

I'm on show schedule again. Non-profit theater that does free morning performances at local grade schools. So that means a different school every Tuesday and Thursday, show up at 7:30 AM, unload the van, set up the gear, one hour show with no intermission, then break it all down again.

This was the setup as of Tuesday. Today I patched in four more channels of wireless microphone. Which needed a second mixer as submixer. See, that's the thing about these small boards. It may say "16 channels" but what that means is 8 or less mic-level inputs, then the rest of the channels are line level and shared in pairs.

Some mixers take this even further (looking at you, Behringer). Digital input channels are counted, "tape" inputs (RCA jacks that patch directly to the main bus with no trim or EQ) are counted. They even count channels that share controls, meaning you can use "5 and 6" or "7 and 8" -- just not both at the same time.

Well, it didn't go well. There were massive RF issues. Drop-out as well as interference. Some of that may have been the environment (different school every show, and not enough time to do a scan of the RF environment or adjust frequencies). But some of that may have been IM; intermodulation interference between the microphones themselves. So I'm running Shure's wonderful freeware program Wireless Workbench before the next performance to see if it can identify any intermods and find me new frequencies.



This has also been a week for progress on the violin.

Here it is being all tacticool with the clip-on Snark tuner, new shoulder rest, and new (black) carbon-fibre bow.

Anyhow, had a neuromuscular breakthrough; stumbled upon the right mental signal to do the proper arm vibrato gesture. Only for a second, but that was enough to find it again, and again, until I could find it reliably. Soon I'll be able to practice it and start refining it until it becomes instinctual. At the moment, though, it is still a ton of concentration to hit that exact gesture.

There's nothing exactly "vibrating" about the Holocrons. Aside from the "talking" one -- I got fairly decent results clamping a surface transducer against the outer shell and using the Holocron itself as a speaker cone.

There is PWM, of course. Unlike my other LED projects, I'm not using my hand-rolled software PWM library, or the on-chip hardware PWM the multiple internal timers of the AVR series makes possible. Instead the Holocrons are leveraging the WS2812 chip, a chip that combines stand-alone PWM, a constant-current limiter, and serial bus. The "neopixel" sold from Adafruit combines that chip with an RGB LED. More costly at a couple of bucks each, but really makes wiring simpler. All you need is a single signal lead and a 5V source. And they are free-running, which means your code can be busy doing other tasks while the LEDs work away in the background.

(My Holocron code is calling them about 500x a second, though; a slow dim up and down, and superimposed on that, an even slower color shift from blue-red to blue-green and back again).

Anyhow, I'm finally over the 'flu, show is open, so I'm hoping to start shipping within the next few days.

Tell them we are from France

I suspect it is a bug, a botnet, or some peculiarity in the Blogger reporting system, but my site stats still claim 95% of my visitors are located in France. Landing on a random scattering of posts from around the middle of last year. I dunno.

I'm currently broke and sick. Neither will last long; payday is Friday and this feels like a 2-day bug. And if it were to get worse, I actually have coverage now. Extremely good coverage, which makes it worth going to work even when I feel like staying in bed.

And I am this close to finishing the first production-run holocron. More minor details that aren't right, but I'm done with iteration. Time to start putting them in boxes.

(I've got about 15 minutes worth of glueing before Serial Number 00 is complete. But I'm sick, and distracted by being broke, and tired because I worked anyhow, so it is safer to wait rather than spoil the thing by combining plastic welding glues with my current lack of coordination).

Next post will probably be on those last steps of making the holocrons. (The most head-scratchingly complicated and annoying one being trying to figure out current postal regulations on lithium polymer batteries).



Oh, yeah. And I'm still averaging a mere twenty minutes of practice a day with the violin, but I'm now scratching away at "Katyusha." Good exercise to work with different scales once in a while (actually, if feels just like the same old first-position/Gmaj with a flatted second finger there on the A string).

Whadaya mean, "too sensitive?"

Monday I did a long workday, went to the gym, worked on the Holocron, went out for dinner, and still had time for a few minutes of fiddle practice.

Tuesday I was sore, nor did it help that I had to climb the walls of the theater (pretty much literally, using rock-climbing moves up the unistrut to get to the things I needed to adjust). But still got the Holocron work and the fiddle practice in.

But I am so glad Wednesday was a short day and the next four days are off work.

I needed to dial in the sensitivity on the touch sensor before I moved on to the last parts of the software. And that took four days. And the solution is a software cludge. 

The theory is simple. Let current leak through a high-value resistor into the parasitic capacitance around an input pin and associated antenna until it charges and trips the pin. Measure the time; that's the total capacitance. 

A human body in contact with, or even close to, the circuit changes that time and you can thus detect it. I can get the circuit to detect a hand moving six inches away.

But not in a stable way. The values fluctuate too much, including the sensitivity climbing up steadily over time. The touch I'm looking for is getting lost in the noise.

And two other problems. The sensor wire is inside a sealed acrylic cube. So the ground floats when it is used in battery mode. And the sensitivity goes way up when it is tethered via USB. So my software cludge looks at the detect events and if they are way, way above the default threshold it assumes the Holocron is on USB and flips to the higher threshold setting. And I'm forcing auto-calibration at frequent intervals. And I'll probably write in a masking routing to prevent trips from coming too close together.

(That's more RTOS stuff there, of course. The Holocron is constantly feeding illumination level and color updates to the neo-pixels, meaning you have to use counters and/or interrupts instead of relying on delay() to tweak the timing.)

On to a new "Talk" animation, and take at least a weak swing at the user programming buttons. I've got the first six boards soldered up now and I have at least that many orders waiting to be filled. I'm glad I'm smart about this and have refused to accept payment until I'm ready to ship. (Because I still have to run out to the lasers to cut more pieces, too. At least the designing is finally done on that.)

Oh, and I just ran into the cutest hack on Hackaday. I'm already leveraging the fact that you can reboot the Holocron merely by tipping it far enough to activate the tilt sensor. But with a capacitor on an I/O pin, you can detect if the sleep interval was short or long. Meaning different behaviors depending on whether it has just been turned on, or if the user tips and restores.

And, yes, I was tempted to give up on the capacitive sensor and use this to trigger the sound effects on the "Talking" Holocron. But it made it a little too much like a Moo Can.

Cold

It's cold out. Rainy also.

Good day to spend inside prop-making. Where I'm chasing down cold solder joints. Perhaps I need to scrub my PCB's before I start soldering on them. Or on the revised board (which I thought was complete until I thought of adding pins for a "talking" version), increase the annular rings on some of the critical pads.

No failure yet on any surface-mount component. But the hand soldering has thrown up two USB jack errors and even a couple of loose joints on the LEDs.

Oh, yeah. Talking. There's an Adafruit board that will trigger 10 consecutive (or random) sound files when a single pin is held low. Unfortunately they discontinued the version with an on-board power amp, which means I need to find space on the Holocron for two additional circuit boards. This, and even more ribbon cable, is so aesthetically displeasing I don't really feel like making this a standard option. After all, this is why I went through all the effort to make a single stand-alone board that combines charge management, USB hosting, and LED controller:

Also aesthetically displeasing is the sound. I experimented with the drivers I had lying around, and I needed to pump up to three watts into a quality driver before it could push its way out of the acrylic box and sound half-way decent. I've gone and ordered another of the surface mount transducers that failed me on the raygun project --

-- because anywhere else would create at the least a nasty shadow and possibly require a hole in the box as well. Besides, sound coming from a hole sounds like sound coming from a hole. It is a better effect if it sounds like the box itself is making the sound.

Pity I happen to know a little acoustics (and cabinet design). Because what this really calls for is a ported enclosure:

But that totally changes the look of the Holocron. And this is not a time to think outside the box. I've spent way too long refining on this particular box. It is time to brave the rain, visit the lasers and cut enough pieces to start shipping complete Holocron kits. (I have just enough to make one complete kit now, and I might assemble it myself just as a final check of the latest round of adjustments).

Or solder up the last of the current run of circuit boards. I stopped dead when I ran into odd USB behavior, but now I know what was going wrong and can fix it. I should probably solder up a complete "talking" Holocron before I order the version 1.1 boards, though. There's just barely space to put a couple more terminals for that...



(Oh, yeah; there's alternates. Best is go with Teensy and the Adafruit audio adaptor daughterboard for it. Skip USB hosting and in-place recharge and just put a string of Neopixels in for lighting. The Teensy does the heavy lifting of not just sound playback but DSP to equalize it. But that's costlier, still messy, takes up even more space, and requires I spend time programming.)

Help me, Obi-Wan

I'm in the middle of filling those Holocron orders. I'm already sick of them...been sick of Holocrons for months now.

Oh, but while I was restocking parts at Adafruit (I'm dangerously low on some of the Holocron components already) I rediscovered a very cute little OGG-based sound board they have there. It's got flash built-in, is visible via USB as a standard mass storage device, and can be as easily triggered from a free microcomputer pin as it can be from buttons.

So I actually could make a Talking Holocron without that much more equipment. Since I've kind of overtasked the CPU of the current Holocircuit already, this would be a re-purpose of my previous Cree-based board (or the next generation Cree board...long discussion there about what is going to go on it, though.)

And it would also be a good excuse to test Cree lighting of a Holocron; now that I've incorporated the internal diffusor I previously thought was going to be too complicated to do...

HoloBOM

I finished the Stolen B prototype today and opened up the thread at the RPF to start taking orders.

(Top image is the completed "Stolen B," including a prototype assembly of the final lighting diffusor, support structure, and circuit with USB jack. Bottom image is a mock-up using borrowed "Temple" shell and the lighting module from the "Stolen B" to show off the combination of "Guardian" diffusion and "Gallifrey" circuit layers.)

The above is also why I spent a few minutes today developing a BOM with parts numbers just so I could keep track of all the pieces properly. Here's the BOM for the design I've been showing off in earlier posts:

20.1 “Stolen A," assembled

0.1.31 “Stolen” shell set

0.1.31.1 Top

0.1.31.2 Side (3 pieces)

0.1.31.3 USB Side

0.1.31.4 Bottom

0.1.41 “Counselor” diffusion set

0.1.41.1 Diffuse top

0.1.41.2 Diffuse side (3 pieces)

0.1.41.3 Diffuse side USB

0.1.41.4 Diffuse bottom

0.1.51 “Circuit 2” set (6 pieces)

0.1.61 “Standard” diffusion cube set

0.1.61.1 diffusion cube top

0.1.61.2 diffusion cube side (4 pieces)

0.1.62 Support set

0.1.62.1 support top

0.1.62.2 support side (4 pieces)

0.1.71 “Revision 3” electronics package

0.1.71.11 “Revision 2” neopixel board

0.1.71.21 USB jack

0.1.71.31 Standard LiPo

0.1.71.41 Capsense wire

0.1.81.1 Magnet (4 pieces)

Tomorrow I'll probably solder up another couple boards, and scrounge and adapt from my discards pile to complete the "Temple," "Stolen A," and "Imperial Archives" prototypes. I need two Holocron gifts so at least two of those are going away (I can't see those particular ones, as they don't quite meet my standards for shippable product.)

Yes...working at a company that makes precision audio equipment has tainted me. I think in terms of QA and Reliability Testing and BOMs and MAI's now. But I need to; despite my original intention of making the cheapest possible kit that was also as smooth and simple to assemble as I could make it, the realities of the core design elements of the "Stolen" fork has produced a design that has a lot of individual parts and requires a fair amount of finicky work to assemble.

Now all I need is to add a proper tracking system for revisions...

Out with the old, on with the new...

...soon.

Yesterday was productive. Checks have arrived, bounced back from sick (after getting even sicker, that is), I finally did the "Gallifrey" circuit design for the Holocron and drew up my big attempt at the "Guardian" diffusion layer. Old-school; worked with ink (and lots of opaque white) on a drafting board because working through a graphics program was constraining my choices in the wrong ways.

While I was shopping for a new technical pen to complete the above I got an idea how to make the Wraith Stone work.

I've been pondering for months various schemes of multiple castings or fills to get the green inclusions. Well, after seeing some "ebony" Rub n' buff at the art store, I finally made the paradigm shift to accept casting the whole thing in translucent green, painting out the rest of the model, and dealing with the less-than-perfect way paint will interact with the internal lighting.

And actually, if I go clear instead of translucent color (using the LEDs to provide the color) I may be able to omit a casting stage and use the printed model straight.

So, yeah, I'm all inspired on this one now. I also came around on the qualities of the raised edge; a hand-worked look and resulting variation of thickness is fine. So cut the basic form from MDF then build up the details in Apoxie Sculpt. It's going to be built at about 3-up, scanned, and then printed to the actual smaller scale so not that painful to get the details clean enough.

Holocron history

This prop took forever. But there is reason for it. Basically, it isn't "a" prop. There wasn't a straight-forward design process from a base idea through a directed iterative exploration.

I was handed a kit to assemble and paint. I'd just been introduced to laser cutting and engraving, though, and I thought I could pimp it up a little.

The experiments worked. Well enough I ended up documenting the project for Instructables. And that's where the trouble started.

Enough people at Instructables showed an interest that I made my files available. Since some of the parts weren't originally mine, I had to come up with designs for those, as well.


It was through Instructables that I was contacted by the master of a Jedi Temple, wanting a custom design made for his students. I agreed to work on it. Many emails and iterations and a full free kit shipped out no charge and I stopped being able to shake the feeling that he wasn't actually going to be good for the cost of the final kits. So I parted ways with that customer.

Since I now had a new and tested shell design I tried for a while to generate a new holocron based around it. But I didn't like (and still don't like) and of the results.

The holocron does not yet appear in any movie. It appears in some games and animations; one appearance being the best documented appearance I've been able to find. This natural goal was blocked, however, by the seeming impossibility of achieving it with the materials at hand. So I continued tinkering with other alternate designs, trying to fold in various motifs from the Star Wars universe.

It was at that juncture that I opened an interest thread at the Replica Props Forum. I got strong interest there, but still couldn't satisfy myself with the design.

Took a break to work on other projects. Did the Retro Raygun, some other things. My Croft necklace was also a hit, and I gave it away on long-term loan, which led me to starting the Wraith Stone project, and that looked to require some advanced electronics, so I picked up the holocron project again just to be able to work out the charge circuit and load sharing and surface-mount issues on a simpler board than what I intended for the Wraith Stone.

And when I returned to the holocron, what I had seen as an unsolvable problem turned out to be trivial. The critical insight might have been a function introduced on the new lasers just installed at TechShop; vector engraving. In any case, I immediately moved to front position a design much more closely based on that one animation.

It is just different enough from my first holocron, though, that the lighting didn't look right anymore. So back to some very basic development to rethink how the lighting circuit interacts.

And, of course, late in the day I realized there were possible ways to get it to look even more like my selected source. The very first holocron was a three-layer model; painted shell, solid diffusion layer, then the vector-cut "circuit" layer. I finally broke through this paradigm -- first through having to add a diffusor cube, then through realizing an inner "hypercube" might be an even closer match to what was seen on screen.

And that's where I am right now; cutting out yet more test pieces to see if this idea works out, while my growing list of confirmed customers are demanding I let them give me money...

Which is of course the absolutely perfect time for a major change at my day job.

Way Hot

The delightful image above is from a lightweight but not completely horrible 1993 version of the venerable Journey to the Center of the Earth; another failed series opener released as a stand-alone TV special. It is much, more more Pelucidar than it is anything from Verne, but anyhow.

Weather is starting to cool but I'm walking to work. I'll see how long I can take this. Come the first of October, my boss is going to talk to his bosses about trying to get me a raise to something a wee bit higher than the new minimum wage.

The holocircuit has passed all tests.

Even with a mere 150 mAh lithium polymer cell I got seven and a half hours at the nominal/bright setting. That's plenty long enough. Unfortunately I'm down to scraps of acrylic...and Tap Plastics is inconvenient to public transit.

(No inconvenient to get to, per se. But inconvenient to get back from while holding unwieldy sheets of material that is quite capable of giving you nasty cuts. Bad enough riding BART with the things. Walking a mile in the heat with them is an experience I'm not longing to repeat).

Stall

The warm weather is back. Allergy season is upon us. And I'm feeling it.

I'm in one of those places where I'm not that interested in what I'm doing at the moment but feel obligated to finish those things off before I can start anything new. So everything else is in a holding pattern, but frustratingly, the current stuff just won't stick the landing.

I'm hours away from being able to start sales on the Holocron. But there's a logjam. The new lighting circuit doesn't arrive until the end of the month, and without it I don't really feel comfortable with spending the laser time cutting more parts. I could start selling the prototype version right now, and problem-solving in production is the most cost effective at this juncture, but I don't want anyone to have buyer's remorse when they see a different prototype -- so I kinda have to complete and take pictures of everything before I can sell anything!

There's at least two laser sessions needed here, so there's no sense in waiting to do it all at once. But I'm thinking so slow I can't seem to figure out what parts I need to complete the other mock-ups. Adding to the problem, I saved in the wrong format and I can't even access the most recent version of the laser files until I go to the shop!

(The parts I feel most comfortable in working on right now exist -- of course! -- only in the version of the file I can't currently access).

And the new diffuse and circuit layers are not going well. The one is just slow, and the other just doesn't look right. Oh, and just to add to my dissatisfaction; the elements that are currently on the prototype and look good are those that are not canonical. So it is makes a good looking prop at the cost of making an inaccurate prop, and that's deadly.

Ah...there we go. A quick ad-hoc lighting test, and the idea I had for a more canonical Prototype looks good. Which means I need to either make two trips or schedule enough time in SF to revise the files before I laser them out. (I need a Windows PC with the latest version of Illustrator before I can save back down to Inkscape compatible files I can work on at home again).

Revised version 3.1.4 being modified diffusion "hypercube" with "Stolen" shell pattern over it, hand-sanded diffuse layer, and no circuitry layer. The only other achievable canon look I could do is if I could attach a little theatrical gel or something to make the cut-out look:

As a reminder, the current version, with diffusion hypercube and circuitry layer, looks like this:

The thickness of the outer shell pieces is a necessary flaw of the materials and assembly method, by the way. To keep it under my $65 target the kit by necessity needs to be not entirely accurate.


Oh, and pricing is still boggling me. Including the commute cost for a laser run, the shell pieces look to cost me about $35. The electronics parts look to be $25, of which most is in the PCB itself, the battery, and the neopixels -- the latter, even though I could substitute cheaper with my new diffusion design, provide more flexibility.

Also looks like about two hours for the latter, four to six for the former (depending on if I factor in time spent in prepping for shipping). Assembly of a non-kit version adds up to ten bucks of additional materials and 4-8 hours itself (depending on how you count waiting between coats for stuff to dry).

I'm aiming for near-cost for the kit, so I think I can do the shell kit for $45 and the electronics module for $35. Which adds up to an unpleasant $80 for a total kit and since I'm at cost already I really can't take anything off for getting both.

For fully assembled, the electronics doesn't actually cost me any more that way, but attempting to factor in materials and some approximation of cost of labor (rather, opportunity cost) brings the total to $95 or so. Since I want to deprecate the complete model and emphasize the kit, I should go at up to the full recommended markup of 2.5 - 3x cost; a minimum of $180, preferably slightly north of $200.

Yeah...this stuff doesn't really belong in the blog, but I need to think it out somewhere before I post up at the RPF.

Sigh. Too many options. Even worse if I acknowledge that some options are more or less costly to me. But here's what I'm looking at:

Fully Assembled Holocron: $190

          - $5 if omitting the diffusor hypercube
          - $30 if omitting the circuitry layer (will assemble with unconnected USB jack, cord notch, or plain wall as requested).
           + $5 for alternate colors beyond uncolored/dark blue/light blue options.
           Please discuss for custom diffusion or shell elements

Holocron Kit:

            $45 shell kit
            $35 lighting circuit

Actually, when I think about it I will really have to set them up as "Stolen A, Stolen B, Imperial Archives, Temple A...." with "Please detail for mix-and-match options." With the make-up of each of the above described and illustrated with an example.

And, oh, sigh, now I really can't start that sales thread until I've assembled a complete "Stolen A" version with the new combination hypercube I just dreamed up today.

Holo....whatever

I've got at least five orders now, but all of them are for completed holos. Maybe when I put the price list up I'll get a few kit orders.

Circuit board is at the fabber and expected back in two weeks.

I've got a napkin sketch now for the "Gallifrey" circuit layer. Shouldn't take that long to make the Illustrator file for that but...

At least everything else seems to be working. The diffusor cube was the perfect idea. The new support is much simpler than the old "spider" support. The USB jack is finally solved "good enough." It just took a whole bunch of iteration to get this far.

Assuming the revised board doesn't have any problems, about the only test I can think of is of the capacitance sense circuit, and of the programming buttons (and of course the software routine that leverages those buttons). So basically I can put up my price list now.

Only reason I'm holding off is I want to make as many of the options I intend available from the start. I don't want any buyer's remorse happening because I posted a design they wished they had ordered instead.

Holoback

The hypercube works:

I created a 2" cube of white translucent acrylic and propped the lighting circuit inside that. (Actually, the picture shows the cube being illuminated by one of my 3 watt Cree). It looks very nice now. More, the diffusion works so well I'm no longer forced to bend the LEDs around to try to get 360' coverage (although the construction of the cube and support means the underside of the circuit board, where the holocron lid is anyway, is going to be less than properly illuminated).

While I was lasering that, and doing the final tweaks-and-tests for the revised circuit board supports and inner layer pieces, I also ran off a test of an idea I'm calling the Imperial Archives Holocron:

This is actually done with 1/16" translucent white mated up with identically-outlined 1/16" opaque black acrylic. Unfortunately I don't have a holocron-sized white light source, and this photograph is using a Maglight Solitaire propped up on a roll of solder.

The circuit is routed as well. Just needs the silkscreen cleaned up.

Is funny; I felt like hell all week (longest four-day week ever) but after getting some sleep managed to crawl out to the shop for the lasering above, plus put down the next thousand words on my Tomb Raider/SG1 crossover.

Well, I've got some problem-solving to do before the latest chapter is complete. Because I've put Checkov's Concentrator Solar Power Plant up on the Mescalero Escarpment mantlepiece, and I need to figure out why it is about to go off...

I am not left-handed

I'm taking a day or two off work to let my injured thumb heal. I needed it anyhow -- got a cold to get over as well, and I could use some uninterrupted time for prop work.

Amazing, though, how many things one uses a right thumb for. I've even trained my left (I'm a self-taught touch-typist) to hit the spacebar instead, but it throws me off.

Anyhow.

I've three variant holocrons coming along. Still trying to problem-solve on the related lighting and USB issues.

I hacked up my prototype today to try several different arrangements of an external USB connector. The latest try is trying right now. I think the best lighting is going to come from a diffusion cube suspended inside, and the simplest way to attach the circuit board that maintains easy access to it is to come up with a platform that glues right to the bottom lid. That way, the USB connector travels with it, no dangling wires.

As clever as I've tried to be with the USB jack, it still seems to mean there is both a visible seam and an even more unsightly hole in the side. Hence the quick repair and paint job so I can put it all together and eyeball just how bad it really looks.

And...it works. The hole is a trifle ugly but small enough you can more-or-less ignore it. And there's no distressing light leak either.

The neopixels I want are still on back-order. Might end up hand-soldering raw "5050" LEDs (those are the RGB's with integrated WS2811 driver chip) -- they have a poor survival rate when reflowed. Would save $3-4 in parts, but also would make the diffusor cube mandatory.

The revised Eagle CAD is coming along. I've got everything in a 2" x 2" footprint now, even the user programming buttons, and it still looks routable (the auto-router digested it without issue). Pulling the USB jack off the board will make it even easier.



Drat. But that was two days off from work. Sick days, yes. And I did a little cleaning, laundry, etc. But there's still enough to do on the laser files I'm not sure I'll even have those done in time to do on the weekend. And I've a promise to run off more grenades and some 10mm caseless rounds as well...

Hol-oh-no

I stuck pictures up of my first production prototype and so far all the responses have been for a fully-assembled and wired Holocron.

This isn't what I got into the project for.

I wanted to make available a kit; a kit that had better aesthetics than the one kit I knew of, that assembled easier (not hard!) and that was if possible cheaper. The majority of work I've done on the mechanical design has been in aid of making it easier for other people to assemble one (if it was just me, I'd build a jig and slap glue over the pieces and there it would be).

Probably thing to do is to finish up the samples of the alternate designs and see if I can entice more people into getting the kit instead. Or deal; assembly is really not that onerous (after all, I did do all that work to make it easy!) and figure out how I'd have to price them to be worth it to me.




Oh, but the lighting circuit just isn't working. And there's worse. I just did an experiment. Faked up a box of translucent white 1/16" acrylic and stuck a 3W Cree in there. And it looks wonderful. I'm willing to deal with a dark spot on the bottom -- the holocron opens up there anyhow so the illusion suffers already at that angle -- so I can probably make this work with a circuit board lying on the bottom of a diffusor cube.

The cube shape is just barely glimpsed, but gives an incredible sense of depth and complexity..and almost tesseract-like effect, which can only be enhanced by detailing the cube with some black acrylic or vinyl decals.

The alternatives I have aren't wonderful. Assuming I can't come up with a simpler way to cover at least the 180 hemisphere, either through-hole Neopixels that can be bent outwards, or make my own mini circuit boards with right-angle headers.




Well, okay. I need to fix some issues with the files, laser off enough pieces to fix the magnet problem on the first production proto and complete the Imperial, Temple, and perhaps a second Stolen (mostly because I already cut most of the shell parts for one already). But I might not get to the laser this weekend; I also have to problem-solve this new idea of an inner cube, and how it gets attached.

Holonought

The lighting circuit doesn't work.

Basically, the circuit board is just too damn big. Also, the current neopixels are a pain to work with.

So my current idea is to break up the board into two parts; the upper part being the CPU and LEDs, the lower part containing the USB host and Lithium Polymer battery charge circuit.


Also, there's something odd going on with the magnets. The lid seems to want to hover just slightly ajar. I think I need to move the magnets so they are attracting straight in instead of sideways. And I finally figured out how to do that. And it should even allow me to shrink the border slightly for a nicer (and more canonical) look.

I need to cut a new diffusion shell anyhow (to improve the snap-fit) so for my alternate "stolen" holo (which has narrow edges but is otherwise design compatible) I will do one without laser-engraved diffusion, and experiment to see if judicious sandpaper application can get more of the canonical look. Plus be a cheaper kit offering.

Meanwhile three other shells are in various stages of assembly and painting. Given the time to run off more diffusion pieces, I should end up with four holocrons lying around...