Tuesday, September 30, 2014

Maybe I should get some bitcoins now

I have $56 credit in my Renderosity store, $19 credit in my Shapeways store, and $144 in my PayPal account courtesy of M40 grenade sales at the RPF. But I also just dropped $50 at Amazon for Pachmyers and a bottle of gun blue.

At some point I need to put some bucks down for a couple of larger chunks of aluminium. Slabs for the Jubal Early. And a big chunk of cylinder if I get the go-ahead to lathe up the flash hider for a Commando Cody.


I haven't entirely given up on the KP/-31; if I can just grind enough slag off the lugs to where they will function again, it might just be possible to make the welding alternative work. Especially since I discovered I have access to a wire welder (one of the el cheapo boxes, without even the external reels and gas fitting).



Saturday, September 27, 2014

Scrivenings



I could probably model the Jubal Early in Carrara. I've figured out how to work in scale in Carrara, how to use reference images, and how to export in a format useable by a 3d printer. It should be possible to follow the same pipeline towards g code the CNC milling machine can use.

But Carrara is useless for figuring out mechanics. The Jubal has a few interactive parts. Other projects down the pipeline will require even more. And perhaps more importantly, Carrara is not a parametric modeler. The control I get from it is due to working with mesh directly. I'm quite good at building compound curves and subtle rounded edges with it, but a proper parametric modeler makes this much more painless.

So I'm struggling to learn a little Autodesk now. Aside from their flagship, Inventor, they have a nice-looking student-level package called Fusion 360. Which licenses for a mere $40 a year. With a free trial period of up to a year!




The software I wanted to mention today, though, is Scrivener. This is another piece of extremely affordable shareware. It was built to manage a novel, but it is good software for organizing all sorts of documents. And, I've found, decent for organizing projects as well. It isn't project management software; there's no calendar and milestones, inventory control, Gantt Charts and dependencies, etc.

But it is a very good way of organizing a large number of document scraps and stubs, allowing you to sort them, view them in a hierarchal order or in the ad-hoc associations of a cork board or as selections strung together in a single free-reading text. These scraps and fragments, outlines and research findings and full chapters can include images, links, and various sorts of internal links and commentaries and side notes within the Scrivener project itself.

I've been using it to plan shows lately. I throw individual documents for each meeting into a folder, establish other folders for show-specific issues like pull lists or inventories, and I can look over the project as a whole in the outliner at any time, and open notes in split view while working up cue lists.

For the writer, Scrivener has the ability to generate word counts, set page goals, and track revisions. When a project is complete, it can strip out the inline notes and compile the selected final drafts into a formatted document ready for use in various online publishing systems.



Pity I can't make better use of the ability to arbitrarily re-arrange scenes and chapters. I tend to write too densely, with too many threads within each scene. It would be extremely difficult to re-arrange scenes when so many of their internal parts only work in one order.

Scrivener has a slew of other tools which I am quite behind in making use of, though. In-line annotations, global referencing (helps a lot to keep your character names consistent!) and so forth. Sigh. One more piece of software I need to find more time to learn properly.

Friday, September 26, 2014

More Blow-Back

I've been scribbling ideas for a new prop. But I also learned that one of my theaters is doing "Mockingbird" this season, and are already having questions about how they are going to do the gun.

I've done this once; for "Tis Pity..." the actor had a plastic gun and a keyfob transmitter in his pocket. He'd press the button while miming firing the gun, and the sound played out the sound system. Which is how I ended up cleaning stage blood out of a radio transmitter but anyhow.

The idea I meant to throw together was basically a cut-down Nintendo Power Glove. One flex sensor in a fingerless glove (or a flesh-colored finger cot) leading back to a XBee or similar radio transmitter.

This does have the advantage of simplicity. XBee modules have onboard analog to digital converters, and can stream a steady set of samples from the sensor. On the other side, a simple carrier allows you to bring the serial data stream into software via a USB cable. This is what I did to interpret the accelerometer data when I was messing around with a spell-casting wrist band. Since the parsing is done on the host computer, you can display the real-time data from the sensor, log it, look for patterns, adjust limits on the fly.

The downside is cost and battery life. The XBee pro is $50, and to make the sensor really responsive you have to be transmitting almost constantly.



But I realized yesterday, as I sat in a brush-up rehearsal of someone else's show scribbling away on notes on my laptop, the trigger finger sensor may not work.

The problem is recoil.

There's been a steady effort to simulate recoil in game controllers and laser tag guns. I'd been looking at those options that very day, towards the prop I mentioned above. On stage, it is frequently done that the actor will jerk the gun in their hands to try to simulate the kick of a real round. It isn't that realistic, but it is better than nothing (blanks, of course, have very little kick either).

This is the way in which the sounds don't coordinate. You can't exactly see the actor's trigger finger from forty feet away, but you do notice if they jerk the gun; auditory reaction time is on the order of .2 seconds, plus they do actually have to pull -- in a 1980's study, it was a full .5 seconds before a forefinger could be brought down on a button. Using a vibration motor instead of an alerting sound brought it down to .4 seconds.

Obviously, if it is too disruptive to tear open a rented prop gun and insert a sensor (the reason why I've been designing a sensor that is worn by the actor instead), it is way too much effort to put a recoil device on. Especially since the best commercial recoil devices are just about up to delivering a satisfying "kick" to the user -- they don't jerk the gun enough so anyone would notice it from stage distance!

Which leaves us back with the poor actor having to coordinate jerking the weapon back at the same time they operate a button. And because simulating a firearm recoil requires such a strong muscular effort, it probably precludes doing anything subtle and controlled with the same hand (like operating a button, or even like pulling a trigger). Which means you might as well have a key fob in the pocket.



But I may be overthinking this. The problem is that I shoot. The kinesthetics are clear in my mind; you grip the weapon securely, and you ease down the trigger. After you've taken out the slack, that last bit of squeeze is timed to happen just as you are on a resting breath, and the bobbing sights are properly aligned.

As an actor, though, what could I make work? Jerking seems natural, and controlled. Perhaps an accelerometer is what is called for. Assuming I can discriminate the desired motion. Pressing any kind of button or trigger feels wrong to me...but at the moment I'm trying to work this out with a hot glue gun.

Maybe I should build that prop after all. Or run across the street for one of those NERF guns I've had an eye on for my next repaint...

Thursday, September 25, 2014

Hope

I've been looking for a while for a better option than the XBee for theatrical use. The XBee modules are fantastically easy to use -- basically, they are the Arduino of the wireless world. But they are also expensive. Even eBay sellers can't get you below $20 a unit.

The other downside is indoors range. The higher the frequency, the worse the reflections, and the 1mw standard XBee can't reliably get from one room of my apartment to the other. And you can't count on having good line-of-sight in theater. The XBee "pro" model increases this output power substantially, and can brute-force a signal all the way from booth, across the stage, and into the dressing rooms (which I've verified by using it in actual performances). But this option is also costly, raising the price per module to nearly $50.

The trouble with finding wireless options for theater is the combination of range and necessary reliability. We can't afford dropped signal if that is triggering a key effect. And the majority of bluetooth data links down to cheap keyfob transmitters are designed to operate at, maybe, 25 feet.

Well, there is a series of modules out there. The RFM69 series from HopeRF are under ten bucks each, even for the 100mw RFM69HW model. There are even (nascent) Arduino libraries for these increasingly popular modules. They do lack the ease of use and existing infrastructure of the XBee, but this seem the better choice.

The combination of high power and low frequency makes for better indoors range. And the low unit cost means I can incorporate remote control as a viable option to the DuckLite.



I've been looking a lot at parts costs, and with Chinese RGB modules, the cheaper MOSFETs for power switching, and ATtiny chips -- or, at least, using ICSP headers instead of incorporating a USB option -- I can get the unit price down to not much over $10 a light module. This makes it plausible to use them in multiples in a stage production; to have a set of lanterns or party lights or tea lights or costume lights or whatever.

The downside was going to be that the cost of controlling them remotely was going to be prohibitive, enough it was almost not worth adding the option even as an unpopulated set of headers. But with these nodes, this might return to being a viable option. With the realities of interface, power management, CPU support, physical headers, and so on they should roughly double the price of the modules, but this still seems plausible.



It also means all the work I've done so far in laying out the new PCB is a waste. Back to the drawing board on that one!

Wednesday, September 24, 2014

Successful Failure

I went in to Tech Shop today in an effort to determine if grinding and welding is an efficient way to proceed on the KP/-31 rebuild.

The effort was successful. I determined it is not.

Grinding and filing was tedious, the results marginal, and it would probably tech less time to machine new parts from scratch. Casting is looking better and better as an option.

Tuesday, September 23, 2014

Random Failing Canons

The thing that's making my Tomb Raider/SG1 fanfic so much fun is that I've given myself implicit permission to be wrong. There's a lot of fun stuff to research, and my followers seem to enjoy all the random weird trivia, but it really takes the pressure off to straight up admit I'm not going to get all of it right.

And, yes, I just started writing from one end. Started her in Malta because introducing her on the sundeck of her yacht over sparkling turquoise water seemed like a good entrance. Malta of course makes it obvious to go blathering on about the Knights of Malta, the Tribute of the Falcon, the Battle of Lepanto and so on.

Searching around for random interesting antiquities in that corner of the Mediterranean led me to Tripoli, and inherent in that was a clash with Gaddafi's Amazonian Guard. And the Tribute of the Falcon gave me an excuse to visit San Francisco as well. And so it went; pretty much following random connections around like an archaeological James Burke, until I needed to actively push her in the direction of the Stargate.

I've finally reached the chapter where Lara sneaks into the SGC. Filling 5,000 or more words with that would make it like playing Black Mesa in reverse; lots of crawling down tunnels filled with steam pipes and fiddling with doors, and that didn't interest me. Fortunately, the absurdities of having the Stargate buried under the old NORAD command center, in (according to the show) an abandoned silo, gives me implicit permission to be just as silly.

So I'm postulating a top-top secret continuation of the old Project Pluto (a remarkably insane Cold War creation that never quite flew, thank the gods -- it was, in essentials, a nuclear-powered cruise missile the size of a locomotive. A nuclear ramjet, in fact, spewing hot radioactive gas from an open-core reactor as it hurtled along low to the ground at Mach 3, destroying buildings from the sheer shock wave alone.)

And to find out about Project Nergal I've got her talking to a permanent resident of the gallant, foolish, and probably hugely claustrophobic and depressing Sealand, a would-be micro-nation and de-facto data haven off the coast of England. And while she is in Colorado, I've got her exploring some of the remnant dreams of Gilded Age philanthropist Spencer Penrose. All of which I discovered simply by reading as much as I could about what it looked like on the trails and summit of the mountain.

I may still manage some air vents, but I'm hopeful there's enough random material there to fill out my chapter. Then I can finally get to the place that first interested me; having Lara Croft and Daniel Jackson compare, contrast, and be confronted with the realities of their Indiana Jones school of archaeology approach to the ancient world.




I've been watching old episode of Stargate: Atlantis of late. Which depresses me a bit. The show makes a strong start but then becomes increasingly depressing. Like the much older Space, 1999 it seems to go out of its way to make the heroes ineffectual; no matter what they try, things usually turn out badly. This is not helped by the fact that this is a self-satisfied, moralizing bunch of bigots. There are numerous places where the "designated hero" rule is turned up to eleven; the Atlantis crowd giving the cold shoulder to someone doing or suggesting exactly the same thing they just did or suggested. Only they are the "heroes," so it's right when they do it.

Among the strange attitudes is their seeming certainty of what makes a "human." People who don't meet their narrow guidelines are not treated well. Wraith hybrid? Replicator? Well, it is just fine and dandy to lie to one, betray one, murder one without a qualm. It isn't as if they are people, after all.

As unsettling as this is, the mistake that most stands out for me is putting Atlantis back in contact with Earth before the first season is even over. Voyager did the same thing. Given the chance to set a series in a completely new world, the writers seemingly panic and immediately make every effort to bring back all the elements of the last show. Voyager couldn't travel a parsec without running into something from the Alpha Quadrant, whether it was yet more Borg, or Amelia Earhart.

Enterprise took this one step further. The very precept of the show was "Star Trek before the Federation, before transporters and phasers, when the Vulcans weren't friends, our ships were primitive, and we hadn't even heard of Klingons." And yet, by the second or third episode everything that had been in the franchise before was back. They just had to have the Borg again, even though it violated continuity so bad they ended up ending the series with a giant reset button.

It is tempting to blame lazy writers, but most of these shows don't have writers per se. They have giant teams of script doctors. And they also have marketing departments, and layers and layers of producer oversight. And, yes, the fans want more of the damned Borg. Or at least they say they do.

The error is giving the fans what they say they want, instead of giving them something they need. The first moment someone sat down to create a Voyager script they should have been saying, "Here we are about to cross our entire galaxy. This is a different scale of exploration than we've ever dealt with before. How does galactic structure change things? Are we going to end up in rifts between the spiral arms? How close can we get to the core, and how many episodes can we get out of flirting with the galactic-scale black hole there?"

But, no. Their first question was, "So I want the Borg to attack like they did on every other episode of the last show I was writing for. So what do I have to change to make that happen?"

Star Trek, the original Star Trek, confronted expectations. It told a story that hadn't been told before, and trusted the audience would be intrigued enough to pick up the language and absorb the mythos. It is a very difficult chore to toss any of that mythos, especially when you know you have a ready-made audience. But when the very basic premise of the show is, "This is going to be a little different," you murder your show by quickly erasing and retconning every aspect of that difference away.

Atlantis should have stayed out of contact, fighting their lonely battle with the Wraith and dependent on the contacts and alliances they could make in the Pegasus Galaxy. They should have been always constrained, and forced to new improvisations and new acts of heroism, based on not having access to infinite equipment, re-supply, power, and new personnel from Earth.

They should have confronted the reality of being stranded, and grown in attachment to Atlantis, and adopted the Pegasus Galaxy as the home they would fight for. And all the choices made in the show, from being force to abandon the city to deciding whether they could afford to keep someone like Cavenaugh on, should have been shaped by those realities.




Sunday, September 14, 2014

Permission

There's something I left out of my attempt to explain TechShop. And that is something about the place and the experience that is strongly emphasized in the Maker Movement; a thing that has gotten interest in Make as a way to rebuild interest in STEM skills in students in our schools.

The idea is permission. And that's something I've been conscious of as I've worked with my niece on growing her skills.



From the point of view of someone who grew up around tools and crafts (who was, among other positions of default permission, male), the significance of TechShop is it removes the barrier between the kinds of tools you can own or rent, and the kinds you need to be a company to have. Certain things were just plain out of reach to the hobbyist. You could chisel wood. You could cast plastic. But to mill it out of metal -- well, you had to find someone willing to do it, and they weren't interested in small one-off projects.

I went through this on the Fury Gun. Just getting a weld done was a horrible problem. I had an underpowered welding rig that couldn't handle the steel of the barrels. I talked to a bunch of local shops -- one even had me come over to discuss it, but he bailed on me when he discovered he just couldn't talk me around to giving him a full eight hours of work on the project (and billing me more than my total project budget).

So for someone like me, TechShop is about being able to finally get access to the machines.



But here's the trick. Not everyone has given themselves the necessary permission. The TechShop attitude is very close to the Maker attitude. The basic assumption is that anyone can learn, and that everyone who becomes a member will proceed with care, being aware at all times of the limitations of their knowledge but always pushing to expand it.

It is a polar opposite of the idea of training -- what I think of as the Dojo Method. This is the idea that there is only one way to do the thing, and you will train and practice with paper and simulations until you can reproduce exactly that sequence. Then you are given access to the real thing, where you continue to copy exactly what you were shown in training.

Maker attitude is about grokking the subject. Makers always ask "why." Even something as simple as...

The other day I was grinding a lathe tool, and one of the TechShop staff came up to tell me I was using the wrong side of the wheel. But he didn't give me a rote instruction. It wasn't "The wheel by the lathe uses the side, the wheel by the cold saw uses the front." Instead he said "Look at the tool rest."

The idea was, you build the necessary rule (use the face of the grinding wheel) out of an understanding of the underlying process (place a tool against the wheel using the tool rest to set the desired angle). Knowing the process, means you apply what is necessary even if you are -- as I was -- free-handing the grind.



So this is what happens on every tool in TechShop. You only get access after you have passed the SBU. The "Safety and Basic Use" class is two to six hours and familiarizes you with the controls and the process and gives you basic safety tips. It also includes hand's on, always in the context of an actual project.

In the case of the Epilog Laser class, we were told verbally how to arrive at the correct power and speed settings for a material, how to look for burn-through and dirty mirrors, what to do in case of fire. And we had hand's on in focusing the beam and setting zero. And then we all engraved a dog tag for ourselves, using templates and materials provided by the shop.

This gives you, as some call it, just enough to get in trouble. You really don't know when you finished the class how to get a clean cut in acrylic, much less a nice burn on leather. That comes with experience, experiment, and, yes, failure.

But what it does give you is confidence. Even though you are on your own -- expected to ask for help from staff or fellow members, expected to read up online or even study books, expected to experiment on test scraps and jot down your findings in a notebook as you grew your laser skills, but still facing a machine without an instructor over your shoulder -- you've already done it once.

You know it can be done. You know you can do it. So even when your next piece, you forget how to set zero, fumble around trying to remember how to start a burn, and end up ruining your test piece: well, you don't take those as evidence that you can't do it. You just take them as part of the process of getting skilled.



And TechShop, just by existing, without you taking a single class, sits as one big case of the whole thing. Just that TechShop is there, and members are there cutting steel and wiring circuits and pulling plastic and serging edges; it says, "We believe you can do it."

It is implicit permission for anyone, including those who never held a tool in their life, to walk in and take the controls of a serious, massive, high-tech piece of machinery capable of doing the kind of professional-looking work we tend to think you have to be a name-brand company to do.

Saturday, September 13, 2014

+++OUT OF CHEESE ERROR, REDO FROM START

Spent yesterday with my youngest niece assembling a 4x4x4 LED cube from Freetronics. That's 64 LEDs to solder. They are RGB LED's, also; four leads each, meaning 512 discrete solder joints to make.

Tough soldering, too. The pads are small, LEDs are temperature-sensitive, and the board had overly generous fluxing. But she really leaned down on it and got good joints and she blew through a good half of the LEDs on her own.

(I had to chip in because we only had a few hours in which to work and I wanted to be able to light it up before we finished for the day).




So I fired up the Arduino IDE to run the terminal monitor program and show her command-line interfacing to the completed cube. At least that IDE was looking somewhat familiar to me, but earlier in the day I was looking at pulling an accelerometer from my DuckNode proof of concept, and ended up first looking through Processing sketches to see if I could see the data coming through the XBee link, then running my old terminal ap to talk to one of the XBee modules directly.

Which, yeah; the "+++" is the escape code for the Hayes Command Set. Which is to say; "+++" alerts the machine that the following serial data are to be treated as instructions to the device itself.

And, no, I didn't remember most of my AT commands. So I was trying to code in a language I'd forgotten, running on an application I barely remembered writing, which is written in a language I'd also gotten a little hazy on.

I'm having this more and more. Especially since a lot of applications don't adhere to underlying GUI standards. It gets really annoying when spacebar drag or control-V does something different in the application de jour than what it does across a dozen other graphics applications. Especially when you are flipping from one to the other.

I'm still trying to get used to The Gimp, which goes out of its way (for legal reasons) to do everything differently than PhotoShop did it. And I'm still far behind on working out Blender. But with the software landscape changing so rapidly, and my own interests also spreading so widely, I'm having to on any ordinary day confront at least two pieces of software with unfamiliar interfaces. And worse -- trying to do something productive on them when time is short and the pressure is on!

I had to quickly set up the Arduino IDE and get it talking to the LEDs we'd just finished soldering before I lost the attention of a sleepy kid. At some point in that same day, I was also diving into Reason to show off a piece of music I'd written for a recent production -- and I haven't had that software open for a month, at least.

Over most of this past week I've had Scrivener, Carrara, Fritzing, The Gimp, and Wineskin Winery open as the projects I've been prioritizing (the Jubal Early gun, and the DuckNode), require I jump back and forth between several of these aps as I plan and develop.

This stuff makes my brain hurt. I need food.


(I'm taking a short break from microphone repairs -- have one of those dratted TA4F connectors open now and you need a clear head and steady hand for those.)

Thursday, September 11, 2014

TechShop

I've been mentioning TechShop a lot of late. So what is it?

TechShop is like a cross between a tool lending library and rented space in a warehouse.

In one sense, this is nothing new. People have been setting up in garages and lofts and building things like the first Apple computers there. And there have always been ways to network around people in the same or a similar business to trade a little time on the tools you didn't have.

What TechShop does is bring this down to where anyone can do it. You don't need to pay rent and utilities on a workshop or loft space. You don't have to have twenty years in the industry and a stack of contacts and favors to trade in order to make a few welds. You just have to have a membership fee.

The modern term for places like this is a Hackerspace. Which starts to imply other things; that it is also a bit like a club (as in, a collection of like-minded individuals who pool their resources to get access to machines no individual could afford, as well as aggregate and share knowledge and experience). And it is strongly connected to the Maker culture, which implies several things as well, prominent among them universal access, learning as a goal in itself, and an emphasis on new technology.

But this is way too theoretical.

What TechShop is, is a for-profit organization that operates several facilities around the country. At each facility is a bit of space to work, so you don't have to try to grind metal in your apartment and end up with metal shavings in your soup. Or try to stitch up a full-circle cloak when you don't have enough floor space to lay the whole thing out flat.

At each space are also machines. They do have some of the smaller stuff you could find in a well-equipped garage. I mean: I live in a small apartment, and I still own a bench grinder, drill press, and scrollsaw. They also have larger and more expensive tools, from laser engravers to engine lathes to such monsters as the 4x8 CNC water jet that can cut elegant scrollwork out of 1/2" steel.

So, yeah. If my apartment was slightly larger, I might be able to cram a benchtop lathe in here. Metal shavings in the soup aside. But with my membership fee at TechShop, I can put a few hours in on a 14x40 inch engine lathe. And this isn't just a "bigger hammer." I could probably do those M40 grenades on a benchtop lathe running $600/800 or so. But such a machine could never handle something like, say, a lightsaber hilt out of stainless steel. And cutting threads would be near-impossible.

The emphasis is on high-tech; at the San Francisco shop are six 40-80 watt laser cutters, a CNC milling machine, three CNC "2d+" routers, the CNC water jet, four 3d printers, a CNC vinyl cutter and CNC embroidery machine. The emphasis is also more-or-less on machines you probably can't afford yourself; no home sewing machines I'm aware of, but at least one industrial and a serger. The metal shop has the usual grinders and brakes and cold saws, but all of them are larger sized than you'd typically find in a garage workshop; the metal-cutting bandsaw can chuck something eight inches in diameter (and the machine alone is about eight feet long!)

This does lead to some oddities.

TechShop is not really a shop cooperative in the model of each member having an area they can work on their project. It has a small membership fee (less than cable and internet) and a large membership and it operates more like a gym. It is drop-in based. Members don't spend all day there, and they don't leave their projects there.

It isn't project-oriented, it is process oriented. You don't go into your local TechShop to build something, you go in to use a machine or do a layout or otherwise complete the parts of a build that are difficult to do at home or at your own shop.

My M40 grenades are about as close as you can get to a "pure" TechShop project, in that 2/3 of the hours I spend on each are spent there. However, I pack in my materials and tools, use a machine for a a four-hour slot I reserve ahead of time, and pack up and clean up after the reserved period ends.

More typical for members' projects would be the Jedi Holocron I built; I prepped the Illustrator files at home, came into TechShop just long enough to use the laser for an hour or two, then went home with the pieces to glue and paint and install electronics.



Would I like the ability to stay set up longer? Everyone would. Having a four-hour slot is really a pain when it comes to something like the milling machines, where you can blow an hour easy just tramming the bed, aligning the work, and doing an air pass with your g code.

Would I like if I saw the same faces more often? Yes; sharing a shop means there is deeper sharing of knowledge and more potential collaboration. Often TechShop SF feels like a library; hours go by without anyone speaking to each other.

But at the same time, the reserved hours/large membership paradigm works. It makes it possible for someone like me, with only a peripatetic interest and not a lot of training, to, with very little monetary investment, do work on a 3 + 1 axis CNC milling machine.




Another wrinkle that comes out of the "come in, use tool, leave" is that the smaller tools and parts can be sparse and sometimes indifferently maintained. You can count on finding a screwdriver, because those come and go on a regular basis and they are practically impossible to ruin. But everyone I know brings in their own lathe bits, drill bits, calipers -- and some bring in their own mill collets as well.

The "reserve the machine, work the reserved period" clashes with projects that naturally span several machines. A sonic screwdriver, for instance, would be equally milling and lathing. And the short reserve period means you really don't want to spend three hours sanding and doing test-fits before finishing your turning. It emphasizes breaking a project up into operations that can be performed primarily on that one reserved tool, maximizing your access to it.

This agrees with some of the Maker paradigms while breaking others. What it breaks is the concepts of exploration and iterative method. The onus is on getting it right the first time. Well; the short reserve periods do make it more natural to iterate; you invest less with each pass, so it feels less onerous to give up for the day and come back tomorrow with a better plan. This is what I did with the Holocron; I made four different visits as I tried out different ideas and evolved the design.

The place where it hooks in closer is with design. So many of the machines are computerized, it becomes practically required to model everything on the computer -- and solve many of what would otherwise be trial-and-error fit-up issues there instead of in metal. This is of course aided by the partnership with Autodesk, meaning high-end software is available on the computers at TechShop (along with desk space and free coffee to tinker with your TinkerCAD.)



Is it a solution for everyone? Of course not. Even me -- if I were turning grenades for a larger community, as well as DL-44 flash hiders, Threepio linkages, and so on, I'd want a lathe I didn't have to commute to the City to use, and the ability to set up a jig and come back to the tool the next day to run off another dozen of the same thing.

The heavy users are prototypers; they are using the TechShop equipment while they tool up, experiment, and attract investors. As they grow their company, they purchase their own equipment and move out.

The other user that comes every day is cottage industry artists; especially leatherworkers, seemingly. They have designed their projects so they can come in, do the engraving and the cuts that would be time consuming to do manually, then return home to finish. The membership cost is significantly less than the amortization to own the tool themselves, so it is sustainable as part of a small-business model.

The last major group of people coming through are the students. These are people who just want to tinker around, either with some project they want to do, or to learn a new process or tool. These folks come in a lot for classes but appear to often let their membership run out without them visiting more than once or twice.

Although I have yet to speak to a user who is only interested in one thing (they all tell me they are looking forward to a class on some other machine or process), there appear to be relatively few generalists; people whose work or projects require they move frequently between wood area, metal area, fabric area, etc. It really looks like the laser people come in and walk upstairs to the lasers and are almost never seen anywhere else. And similarly for the machine shop users (the exception to both being time on the computers with their motherlode of high-end software).

And this strikes me as the oddest part of all. It would seem the most like both Make movement, and industrial art, and the capabilities of TechShop, for projects to appear that were equally on a range of very different processes. But so far, I haven't really observed that. Each area almost seems to have its own culture, as well as what are increasingly a set of familiar faces. And the cross-over is minimal.

Wednesday, September 10, 2014

Repairs

A large part of my gear is getting rented this weekend. I also looked over the french scene breakdown of the next show, and I need more channels than I have at the moment.

So far the only soldering I've finished today is fixing a Blue Snowflake I got as a gift. Which, I am sorry to have to say, has some piss-poor engineering inside it. I understand necessary compromise, and how process will make unforeseen changes, but who would feel good signing off on this one!

Here's the problem. Daughterboard holding the mini USB connector. That makes sense; put the stress there, use that to screw to the chassis, have the electronics snug into it. The main circuit board, though, is suspended on two 3-pin headers. Which means it wiggles, and you get poor USB connectivity. After I'd reheated all the solder joints from USB jack to main circuit board, I re-assembled, and then put a couple drops of hot melt glue in to keep the thing from rocking back and forth on the connectors and breaking connection.

The boards are notched to go around the swivel nut that connects chassis to frame. And somehow, the USB was put in at about a 17-degree angle to the notch. Which means, when the board is assembled and installed, the swivel nut misses the notch entirely, and is all but jammed into the connector to the actual mic capsule. It makes it extremely difficult to finish the assembly without knocking out the connector (or tapping the circuit board off it's precarious balance on its headers.) Broken? I'm surprised the thing actually survived being shipped in working condition!




A vaguely similar problem besets my Sennheiser belt packs. The pack usually ends up strapped close to an actor's body, often in the natural curve of the small of the back. The microphone element is connected to the transmitter with a 1/8th TRS locking connector (expensive little things; I'm paying almost $12 each). The length of the connector means lots of leverage on the jack, causing several problems; the connectors unscrew, the wires break inside the plug, and the jack itself snaps off inside.

The latter is a pain because the circuit board is crammed with surface-mount components. The last one, I repaired by carefully cutting the broken jack to ribbons and then individually de-soldering the remaining scraps of the legs. The one previous, I tore the traces off the board trying to lever it out, and that means that whole unit has to be shipping back to Sennheiser for repair.




Despite these directly job-related repairs, I also have many of the parts from my Duck Node experiments spread out on the work bench as well. I want to have something ready for whatever comes up on the next production, even if it is another hand-soldered prototype instead of a proper board. Tangentially related to this, I'm going to be doing a soldering project with my youngest niece this Friday; a 4x4x4 LED grid from Seeedstudio. Which is also intended as a gateway drug into programming.



From the top, this is my wireless EasyButton (which may or may not currently have a configured XBee in it -- I keep switching the things around). Then an Arduino Trinket from Adafruit; a more-or-less Arduino-compatible board based around an ATtiny. I've been thinking about it for the DuckNode because it offers built-in USB programming for under $10. But it is also severely limited in the number of I/O pins, plus the USB bootstrapper is fragile (hence the festoon of wires on it now; those are programming leads). So bumping up to over $20 for an Arduino Mini probably makes more sense.

In the lower right is an abandoned proof-of-concept for an ATtiny-driven 3W RGB. It is being switched through power darlingtons, with hefty (1 watt) ballast resistors. Why I wanted to go for the more efficient constant-current drivers. And on the left is a proof-of-concept for the DuckNode concept, along with the wrist holster; an accelerometer paired with an XBee. For my tests, I parsed the output of the accelerometer in a Processing application, and fired off a sound cue when it crossed the right threshold.

The problem here...and why the project got suddenly so much more complicated...is very few accelerometers will output in analog that can be easily patched through to the ADCs on an XBee. I haven't seen a single one of the integrated accelerometer/compass units that wasn't digital (usually I2C). And that is such a pain to interface with XBee you are better off parsing through a microprocessor and handing that data off in a serial stream to the XBee. Which is also more power-efficient.

And, yeah; XBees are expensive, have a large footprint (relative to some other radios), and they (like a lot of electronics) are moving...from the Series 1 I own, to Series 2 -- both of which are deprecated in favor of the (much less useful to me) Zigbee protocol. But all the other options take time to learn and develop around. I already know the XBees. And I've used them in real stage conditions.



Sigh. And I also have to cut down the cords (they are too long and thin to be safe to use) on some chorus microphones before Saturday, where I'll be tacking them to the unistrut with zip ties. I filled several pages of notebook trying to come up with a combined unistrut bolt and shock-mount mic clamp for them that I could machine up myself, but no luck. Which means my first milling project may indeed be the Jubal Early.

Which is also going slowly. I'm finally getting a grip on how the parts fit together, but that meant throwing away the first attempt at a poly model. Arg. This would make so much more sense in a proper CAD, or at least in a modeling application that was more parametric. Carrara is not a NURBS modeler. What you build is what you get (aside from the option of tessellation -- but that tool has extremely limited options and tends to make a mess of models).

I really, really need to get comfortable with Blender, but out of all of the things I'm constantly having to learn and re-learn, software is the slowest and most frustrating of all of them.

Monday, September 8, 2014

I'm Back!

Was feeling awful for a good three weeks. Yesterday woke up like it was Saturday; like I'd been pulling long hours all week and finally had a chance to sleep in; that sort of "still sore and kind of weak-feeling, but basically feeling refreshed" thing.

Went out to the city and spent ten hours at the lathe. Was taking my time, but being efficient; cut fourteen new primers so I'd have a small stock of them, fitted the guts in two bodies, and then went and cut two more full-function grenades from scratch. Even with the loss of efficiency in doing them serially instead of in parallel, took less than two and a half hours each. 

And still had a spring in my step when I returned (with a heavy backpack filled with aluminium stock, metal working tools, laptop, and a thermos full of coffee.)

Somewhere in there I roughed out possible layouts for the Duck Light. That thing is still bugging me. For me, the platform works, because I can program and I have the ISP to access a minimal AVR. Well -- there's a potential solution via the XBee node, to include in the serial command set to the Duck Light commands that will re-write constants held in flash. 

Still, the very flexibility of the thing is my boon but the downfall in making it really useful to non-programmers. I'm building a tool that allows me to create a lighting look for each specific project as it comes up. I can really only think of one option that I can hard-program so it is a mere button-press or the quick soldering of a jumper away for the end-user who doesn't program, and that's making a lantern (and even then; the power of the complete platform is I can tweak the color, the depth and rate of the flicker, the length of the fade, even the brightness.)

Sigh. Even if I make up a remote programmer wrapped in a GUI that is friendly to the end-user, it wouldn't handle the hardware end. Even something as simple as adding a capacitance sensor, or a microphone input hard-coded to a DalekEye circuit, means jumpers, or dip switches, and of course adds to the number of components on the board.

(For that matter; I really need to allow space for an accelerometer/accelerometer breakout in the footprint. Which, like the XBee, is an expensive part so it doesn't want to be hard-soldered as part of the basic package.)

The Arduino Problem

There's probably another name for it. I think of it in terms of the Arduino shield, however.

When the first Arduino board was made, due to either mistake or some rationale which has long been forgotten, one of the headers was offset by about half a standard pin space.

This meant that protoboard doesn't fit on top of an Arduino with standard headers. And that meant that ever shield (aka the stackable daughterboards which fit on top of an Arduino and add functionality) had to echo this error in pin placement.

So what happens when the next edition of an Arduino board comes out, or a third-party wants to make a compatible? If they want to remain back-compatible with the existing shields, they have to reproduce that same pin spacing. The Arduino board is basically stuck with it, and the longer it goes uncorrected, the more difficult it would be to correct.



I'm in a similar spot with my grenades. I was noticing today as I was cutting tape for the tops that my hand-ground 3/32 grooving tool has shrunk to close to 1/16th. And there are a couple places where I may have been mislead by my earlier research, or that I find aesthetically compromised; the button should probably be shorter, the "crimp" further from the base, the grooves deeper. And I really should have worked out how to give the blue and green caps a different visual look.

Plus, there's multiple places I've learned better ways to machine, or necessary choices. Like adding a "tape slot" to the solid-body grenades, or like the changing angle on the "crimp" so it can be cut without scratching the metal of the base.

But I'm already shipping second orders to previous customers. So I have to keep as much of the look of the very first production run, in order that they continue to look like they are identical. And as much as possible, matching the look of the other major fan-made runs, so one cosplayer can get away with wearing a dozen cold-cast grenades from another builder on his webbing, and using one of mine as a "hero" prop.



I'm still tinkering with the process. The last two full-function ones took 2.5 hours each from start to finish. I'm leaving the compound rest at 15 degrees through the process and only moving the turret twice. As another trick, I clamp the bits into the tool holders at an angle (I can get away with stuff like this on aluminium!) That way, I can leave the tool holder aligned for grooving, and make several of the tool changes without losing zero on the DRO.

I'm still finding little tweaks. I just realized today I need to break the internal edge of the through-hole; what happens, is when I press the plug in the front of the grenade is very slightly deformed. And in some, the rough, sharp edge of the front of the hole now pinches the button and scratches it when it is pressed.

I've also realized that pushing a "dead" tool (that is, with the lathe stationary) into the metal to find zero is a bad idea; I'm leaving tiny pock-marks on the piece.

And so it goes. I'm eager to stop lathing these, and move on to something different.

Saturday, September 6, 2014

Falcon to Eagle to Pigeons

I may have to back-burner my Tomb Raider fanfic "Day of the Falcon." (Falcons so far; Horus, the falcon-headed Egyptian god, the Tribute of the Falcon of the Knights of Malta, and a rival tomb raider calling himself Juan Carlos "el" Halcon...)

I'm trying to throw together the 0.5 version of the Duck Light, despite funds being tight for ordering parts. Started laying out the PCB in Fritzing, which may actually work for my needs, but I've been meaning to learn the free version of EagleCAD. Oh, yeah -- and there's a nice-looking open-source freeware option called KiCad that is worth checking out.

In re learning software packages, I still haven't gotten facile with Blender, which means I'm starting work on the Jubal Early with Carrara. Sigh. I've been introduced to the slicing and tool path tools I need to turn it into g-code the CNC milling machine can use, but that's going to be a learning curve as well.

I'm very worried I won't have enough time to properly repair microphones before I have to start Feeding the Birds -- aka mixing Mary Poppins. And I'm renting out my own gear over the same time period, which at least will relieve some of the financial stress.

And the KP/-31 receiver is still sitting on my desk. Which isn't bird-related at all, sorry. I read through a thread at a gunsmithing forum and that just underlined how many operations there are to machine the thing from scratch. 3d printing would also be expensive, and the shapes are wrong for CNC milling (not all in one piece, anyhow). Very possibly making a master model and casting that would be the most efficient, despite my worries about the strength of the result.

So I'm looking again at welding. I'll take a grinder to the parts and see how that goes first...

Thursday, September 4, 2014

Minimizing

I'm in the cash flow hole right now. Contracts are started, but the first checks are weeks away. I hate being so low in cash. Makes me want to curl up at home with the lights low and hoard food instead of eating it.

The Jubal Early build seems plausible. I'd be surprised if more than a half-dozen people went for one, though. And the tool-up is a couple hundred bucks and 40-60 hours. Even right here at the start of it, it would help my 3d design if I had the Pachmayr grips to measure. But those are $36 a pop. 

The M40 grenades are a better story that way; I have enough metal to finish the current orders.

I'd still like to come up with better ways to do them. I did some research on the fourth axis of the Tomach CNC milling machine. And, actually, there's an even more clever trick; I've seen a guy chuck the stock in the head, and clamp the tools to the workbench, turning it into a sort of poor man's indexing CNC lathe.

But the set-up time for any of this is simply too large, in my opinion. The two things that would save the most time on these would be to have enough tool holders to properly use the quick-change tool post, and to have metal in lengths to where I could re-chuck and assembly-line it. 

Might actually make sense to cut billets to length and center drill them, then work between centers; at which point I could use a template or a saved set of tool offsets in the DRO.

In any case, there's good argument for moving on to props that cost more (even if just in terms of raw material); because even if the markup stays the same, 20% of $250 goes a lot closer towards paying for the next month of TechShop membership than 20% of $45.



Hrm. I just realized there's a couple of alternates to TechShop where I might be able to use a wire-feed welder. With some sacrificial blocks of aluminium or even wood, I might be able to grind out and re-weld the Suomi. Because as attractive as machining from scratch is, that's a heck of a project.

(One downside to welding is that crosses the Federal firearms line again. Unless I very carefully destroy some of the receiver in the weld process to make sure it can not seat the full-auto bolt.)



Just to be silly, I investigated what Tomb Raider 2013 plays like if you completely ignore the entire "salvage" mechanism, and make no effort to maximize skill progression either. That did make the battles a bit tougher -- the boss battles are quite annoying when you don't have anything but the base Dodge/scramble, for instance. But I got through to the radio tower in one hour, and through to the helicopter in about two and a half hours. Skipping through the cutscenes also saves, I dunno, maybe twenty minutes there as well.

Oh, yeah. And got through the major battles with the base, unmodified pistol (no salvage, remember?) Which also proved annoying, as it is significantly less damage and even head shots don't take down many mid-game mooks. I got killed a time or two, but mostly made it through the big set-pieces on the first try. Which is still better than I'd doing on those dratted QTE's.



In the last of this thematic topic, I'm still working up the circuit for my Duck Light, and pretty much determined it is going to be largely (if not entirely) SMD even for the prototype. But I'm thinking it may be worth doing a quick through-hole version right now so I can offer it to the design team on the next production. Except, of course, for the whole not-enough-cash-to-purchase-pcb's-this-week problem.


Tuesday, September 2, 2014

Where's Henry Ford?

Actually, I'm not sure I'm even up to Winchester. No luck in assembly-lining the grenades -- each one is still a manual process and each one is a little different.

Went in today. First task was to find a way to hold one from the other day that I'd forgotten a cut on. And then I cut too deep. Drat.

Ran off a second solid-body and was starting to get into the groove. Set up for the finishing pass, threw on the half nut for a powered feed -- and engaged the cross-feed instead of the longitudinal. Drat. Fortunately, this time I knew I could rescue the piece as long as I didn't re-chuck it. So turned it into another full-function body. Of course, that meant bringing out a bunch of other tools, and pretty much killed the assembly line.

Then ran short of metal. Was just enough on the last bar to finish the order if I was clever. I wasn't quite clever enough. Did two passes in the wrong order. Set up for a left-handed cut, got the power feed set correctly for that. And going "backwards" must have thrown me, because when it came to the moment where you throw off the half nut and take up the motion smoothly with the traverse wheel...I grabbed the wrong wheel. Drat.

Oh, and finished the day destroying another chunk of brass stock trying to part it. At least this time I figured out a better method; just cut a v-groove where the part should happen, and hack-saw it off right in the lathe.


Next set, in progress; the solid-bodies need to be painted, the two full-function need the spring-loaded buttons installed. Unfortunately, all the primers I ran off yesterday are a trifle large; only two of those in the bag were useable. The two grenades on the right are mistakes; my original (at the old dimensions), and the one I tried to rescue and failed.

The one thing I can say is I'm starting to like that 7075. It is very springy; I have to adjust to the tools springing back on me, and dealing with chips on the rough passes is still a problem. Also, it is so tough I can't really dig the knurling in as deep as I'd like on the grenades. But on the upside, with the right feeds it makes a very shiny, hard surface. It feels tougher, and even weighs noticeably more.



I've been looking at the Suomi some more. The plasma cuts they made through the receiver left some pretty large gaps. Just welding isn't enough; I'd need to do filler welds to bridge those gaps. And it also made a total mess of the threads where the end-cap goes.

But machining is no simple task, either. There's a lot of complex slotting for the magazine well, and some big chunks of metal that need to be constructed for that, too. And it seems impossible to do without welds -- which means getting checked out on MIG, and then getting enough practice in to do them clean.

Monday, September 1, 2014

The Men Who Stare at Gats

I was given a project some months back. I said I'd know more about its feasibility when I'd had a chance to lay out the parts I have and stare at them for a while.

I stared at them today.


This is a de-militarized KP/-31 "Suomi," a Finnish-made submachine gun introduced in 1925 and used in World War II. The "de-militarization" consists of -- as you can see from the photograph -- using a torch to cut the receiver into little bits.

My task is to make a new receiver. Not, I hasten to add, a functional one. In fact, there's some nice little laws I'd run afoul of if I made more than 80% of a functional receiver and sold it or was otherwise recompensed for the work. Figuring out where that line lies, and making sure what I build can't accidentally be fired, is going to be part of the problem.

So how to it?

There are two basic strategies (both of which are also applied by gunsmiths who do intend to end up with a functional weapon at the end). The first is re-weld. This might be the fastest, especially as I have no need of the special jigs and resulting accuracy of the gunsmiths. I just need the parts to more-or-less fit.

The other is a new receiver. And where I part company with the gunsmiths is in materials. My first impulse was to sculpt the entire thing and cast it in resin. I might even be able to take some impressions from the existing scraps (or negatives from the parts that mate with them) in order to save on hand-sculpting.

The best argument for sculpting is that it is the most iterative process. At any point, I can add more material back in; there's no risk of cutting too deep and ruining the piece. The biggest downside is lack of strength; the necessary shapes include some very thin portions, and the weight of the barrel and the pressure of the trigger assembly are going to put a lot of stress on them.

Which leaves metalwork. The shape is too complex and the bounding box too large to make subtractive methods optimal, as much as it has a surface attractiveness to simply CAD the entire thing and CNC it out of a single block of aluminium.

And additive methods bring back some degree of iteration; you can work on a sub-assembly without risking the entire piece.

In the real weapon -- and also in the "80% kits" that used to be commercially available (they seem to have dried up), this is how it is done. The "tube" and "sidewalls" are machined separately, then welded together.



So that's my working assumption; take a tube that's as close as possible to the required ID, lathe it down, cut the rear threads, mill out the lugs, mill ejection port and a long slot along the bottom. Then cut into a slab (or several pieces) of flat stock to create the sidewalls, the part that hooks into the stock, and the magazine well.

Press-fit is not going to work here, though, so aluminium welding, and then clean-up of that. That's a lot of work. And a lot of the cuts and holes are necessary because so much of the receiver is exposed; is the look of the gun. And they have to be right to fit the existing parts that show externally, like that trigger group. There's also a question of how to treat aluminium so as to fit in with the look of the other metal.



There may be a third route. It might be possible to simulate the look of the receiver and support the other parts of the gun with rods and careful bits of water-soluble glue. I'll have to stare at it longer, but at the moment that isn't looking very plausible.

In any case, it doesn't seem the best "first project" for the mill. It might actually be a little easier to tackle this thing:


This is one of an excellent run in pewter-cast of the pistol used by the character Jubal Early on Firefly. Not my favorite design, but apparently there are enough people interested in one to justify looking at trying to build a few. It looks like a natural for CNC, with a little traditional machining to open some of the mating slots and tapped holes.

Well, I still have seven grenades to go to fill the current orders. That's at least two more days of machining, plus painting time. Pity I'm not making enough money at this yet to pay for next year's TechShop membership...

Keep them spinning

I feel like I'm backsliding on machining. Does not do much for my confidence in finishing that submachine gun...or doing a limited run of Jubal Early's for sale at the RPF.

I've got several orders of solid-body grenades to finish. Those do go faster, but I'm still at almost two hours (with set-up and clean up) each. So I tried a couple of tricks to speed things up, with mixed results.

The lathe has a quick-change tool post and the ability to remember tool offsets. But that doesn't help if you move the post or rotate the cross-slide. So the first trick was to clamp my cutter at a slight angle, meaning I can leave the tool post aligned for parting off and not have to change that.

The next trick was to leave the cross-slide at 15 degrees instead of 45. That meant I didn't have to rotate between cutting the chamfer (and doing the knurl) and the rest of the cuts. It does make it slightly harder to align some of the other cuts, though, and it means I'm bringing the knurling tool in at right angles to the work instead of along the normal of the resulting knurl.

Unfortunately, that put the cutter to where the work was rubbing along the relief. I figured out what was going on soon enough to rescue the piece, though. But I didn't get so lucky with the next one. Cut the first groove at .174 instead of .274, realized the mistake, parted off. Then realized I could chop the front off and re-use it for the other style of grenade. Except I'd parted already. I cut down the rim, and was just able (with a lot of gauge work) chuck it in over the rim and make the cuts. But I was cutting backwards now to make the final cuts. I cleaned up a battered left-handed cutter from the bin, but it still made a poor finish. And I'd slipped up in a couple other places trying to take the measurements backwards from my usual references.

So two hours of clever improvisations and I had a ruined piece of metal anyhow. Well, at least I learned a little. Didn't help me, by the way, that all the way through this another TechShop member was plowing through some really pretty lightsabre hilts. With the cleanest knurling I've seen done in that shop, too.

Well, it helped in one place. Seeing him hog on 4" billets made me rethink my first cuts. The book value for Aluminium with HSS bits is 250-350 surface feet per minute. I'm starting with a 1" billet and coming down to .790, and that means the machine RPM should be in the 9600 range. However; carbide tools can be run at twice the speed of high-speed steel.

Well, I'd been running at 9100. I bumped up to 1250 RPM and it felt pretty good. And I reduced it down to two roughing passes instead of four, taking over a tenth of an inch of metal off with each pass. This does stress the metal a little, but I'm leaving .004" consistently (and adding it to the other cuts), for a single finishing pass. Since these don't have to be fitted into other machinery, only the nominal finished dimension, and having a smooth finish, actually matter.

Unfortunately I've still got several tool changes, and because I can't chuck around the rim, I can't remove the main body until I've made that finishing pass. And the solid-body ones require a flawless cap, so I can't drill those out and hold between centers. But, still, there's probably some way I can make this a little more efficient. Oh, yes, and the tool holders seem to be going astray; there's not even enough to leave all my tools in holders as I work (which kind of negates the usefulness of the quick-change turret).



I got into a conversation with a volunteer at one of my theaters (the one where I'm trying to fix up the sound system), and realized I don't have any of the music I've created for plays on my current laptop. So I spent an afternoon or two fixing that. And listening to a lot of old work. And might post some of that here some day. Yeesh. I've got that new keyboard, and the new library, and I really should hook that up some day...