Archive for April, 2017
The Perils of PDF
The Dodge Ram ProMaster cargo van we rented to haul our bikes to Glens Falls (and bring some furniture back) sat on their 2500 truck chassis, thus weaponizing an obvious phishing email waiting for me on our return:
Subject: About the Dodge ram 2500
Kindly review full details of your order.
Methner
The From and To addresses were identical, which is always a tipoff, as was the fact neither were any of my addresses. The email had an attached PDF, of course, although the context suggested handling it with the same nonchalance I’d use with any lump of high-level radioactive waste.
That brief text tripped my junk filters, but, somewhat to my surprise, all the scanners at VirusTotal passed Order 372.PDF without complaint (since then, one scanner woke up, smelled the scam, and tagged the file as “PDF/Phishing.A.Gen”).
Converting the PDF to plain text with pdftotext produced an empty file, so the PDF payload isn’t a script.
Passing the PDF through strings revealed a URL for a (probably compromised) server unrelated to the (obviously bogus) email address, wrapped with layout verbiage suggesting a clickable link:
<</Subtype/Link/Rect[ 205.25 467.11 369.91 499.51] /BS<</W 0>>/F 4/A<</Type/Action/S/URI/URI(http://bogus-domain-here.com/wp-settings/bloglist/hh/index.php) >>>>
Passing the PDF through pdftoppm produced this comforting image:
Bogus Order Form – Image
The “100% SECURE” padlock logo, with a green check for added confidence, is a nice touch.
At this point, if a product involves The Cloud, you can deal me out.
Gas Pump UI: FAIL
During our most recent trip, I stopped at a new-to-me gas station, managed to figure out the pump’s UI enough to swipe my card and fill the tank, then utterly failed at the Print Receipt? prompt:
Gas Pump Keypad Abrasion
A quick hike to the adjacent pump suggested pressing the illegible key above Enter, but the UI timed out before I got back and the promised “moment” never ended. The attendant generated a receipt showing I’d paid for the gas and told me to jiggle the pump nozzle, which didn’t improve the situation. We eventually agreed he’d handle it later and I drove away, never to return, hoping that the next customer didn’t get a free fill on my dime dollar C-note.
Surely I’d know what to do, were I a regular customer …
Badge Lanyard Reel Mount
Posted by Ed in Machine Shop, Software on 2017-04-28
A certain young engineer of my acquaintance now carries an ID badge and, so I hear, works in a PCB design & test venue. Seeing as how her favorite color is purple, this seemed appropriate:
Badge Lanyard Reel – front – overall
The guts came from Circuit Breaker Labs in the form of a recycled PCB trapped in acrylic resin atop a plastic housing with a spring-loaded reel inside.
It arrived with a plastic bullet at the end of the lanyard:
Badge Lanyard Reel – plastic bullet link
Which I immediately replaced with brass, because Steampunk:
Badge Lanyard Reel – bullet cross-drill
That made the plastic housing look weak, so, in a series of stepwise refinements, I conjured a much better case from the vasty digital deep:
Badge Lanyard Reel – iterations
All of the many, many critical dimensions lie inside the case, where they can’t be measured accurately, so each of those iterations could improve only one or two features. The absolutely wonderful thing about OpenSCAD is having it regenerate the whole model after loosening, say, the carabiner slot by two thread thicknesses; you can do that with a full-on relational CAD drawing, but CAD drawings always seems like a lot of unnecessary work if I must figure out the equations anyway.
The back sports my favorite Hilbert Curve infill with a nicely textured finish:
Badge Lanyard Reel – rear – oblique
It’d surely look better in solid brass with Hilbert curve etching.
Black PETG doesn’t photograph well, but at least you can see the M2 brass inserts:
Badge Lanyard Reel – lower interior
The first prototype showed the inserts needed far more traction than the usual reamed holes could provide, so I added internal epoxy grooves in each hole:
Badge Lanyard Reel Mount – show
Recessing the screw heads into the top plate made them more decorative and smoother to the touch. Button-head screws would be even smoother, but IMO didn’t look quite as bold.
After seeing how well the grooves worked, I must conjure a module tabulating all the inserts on hand and automagically generating the grooves.
The yellow star holds up the roof of the reel recess in the build layout:
Badge Lanyard Reel Mount – build layout – bottom
Slic3r produced the rest of the support material for the carabiner exit slot:
Badge Lanyard Reel Mount – bottom – Slic3r support
Those two support lumps on the right don’t actually support anything, but tweaking the support settings to disable them also killed the useful support on the left; come to find out Slic3r’s modifier meshes don’t let you disable support generation.
The top plate required support all the way around the inside of the bezel:
Badge Lanyard Reel Mount – top – Slic3r support
I carved the original plastic housing in half, roughly along its midline, and discarded the bottom section with the belt clip (it’s on the far left of the scrap pile). The top section, with PCB firmly affixed, holds the lanyard reel and anchors the retracting spring in a central slotted peg. No pictures of that, as it’s either a loose assembly of parts or a spring-loaded bomb and I am not taking it apart again.
The lanyard passes through an eyelet that pays it out to the rotating reel. I’d definitely do that differently, were I building it from scratch, because mounting the eyelet in exactly the proper position to prevent the lanyard from stacking up on the reel and jamming against the inside of the housing turned out to be absolutely critical and nearly impossible.
The top plate presses the original housing against the carabiner, with the cut-off section inside the carabiner’s circular embrace, which just barely worked: the PCB bezel is a millimeter smaller than the shoulder of the housing.
All in all, I think it came out really well for a 3D printed object made by a guy who usually builds brackets:
Badge Lanyard Reel – front – oblique
I hope she likes it …
The OpenSCAD source code as a GitHub Gist:
Eneloop AAA Cells: First Charge
Posted by Ed in Electronics Workbench, Recumbent Bicycling, Science on 2017-04-27
With an AAA-to-AA adapter in hand, the Eneloop AAA cells looked like this:
Eneloop AAA – as received – Ah scale – 2017-04-20
The glitch comes from a not-quite-seated cell, showing that a poor connection matters.
The package touts “up to 800 mA·h, 750 mA·h min”, with asterisks and superscripts leading to “Based on IEC 61951-2(7.3.2)“, access to which requires coughing up 281 bucks. So it goes.
A full charge made them happier:
Eneloop AAA – first charge – Ah scale – 2017-04-22
The as-delivered 530 mA·h capacity represents 73% of the 725 mA·h after the first charge, so I suppose they’re more-or-less within the “Maintains up to 70% charge after 10 years of storage” claim. The 16-10 date code suggests they’re hot off the factory charger, so they must ship with somewhat less than a full charge.
Comparing the capacity in W·h makes more sense, because most devices (other than the Planet Bike blinky light these will go into, of course) use a boost converter to get a fixed voltage from the declining terminal voltage.
They arrived bearing just over 600 mW·h:
Eneloop AAA – as received – Wh scale – 2017-04-20
After charging, that went a bit over 850 mW·h :
Eneloop AAA – first charge – Wh scale – 2017-04-22
Call it 71% of full capacity on arrival. Close enough.
The Planet Bike blinky will be somewhat dimmer with two NiMH cells delivering 2.3-ish V, compared with the initial 3-ish V from a pair of alkaline cells. I generally burn the alkalines down to 1.1 V apiece, so perhaps they’ll be Good Enough.
Now, if I were gutsy, I’d install a rechargeable lithium AAA cell, with a dummy pass-through adapter in the other cell socket, and run the blinky at 3.7 V. At least for a few moments, anyhow …
Tour Easy Front Fender Clip
Posted by Ed in Machine Shop, Recumbent Bicycling on 2017-04-26
We rode the Feeder Canal trail during a recent bike vacation in exotic Glens Falls NY:
Feeder Canal Park Trail – Branches
The numerous downed branches along the trail and countless twigs on the trail came from a brush-clearing operation:
Feeder Canal Park Trail – Brush Clearing
As luck would have it, a twig snagged between my front tire and fender, snapping the clips holding the fender in place:
Tour Easy front fender mount breakage
Should it not be obvious, each ferrule formerly had two parallel jaws (on the left) gripping the fender, with the tiny screw digging into the fender. I affixed the fender to the broken clips with copious amounts of duct tape and we continued the mission.
It should be obvious why those ferrules are not suitable for 3D printing.
However, with the recent rear fender clip serving as inspiration, this didn’t take long:
Tour Easy – Front Fender Clip – Slic3r
The front fender fits a 20 inch wheel and is somewhat wider and flatter than the rear fender (I think they bent the same plastic strip around a smaller mandrel), so I did a quick copy-and-paste hack job on the OpenSCAD source code, rather than trying to parameterize the daylights out of the previous model.
The posts around the wire stays are 6 diameters deep and reamed to fit; the stays won’t be flopping around even without fiddly mechanical hardware retaining them. The holes extend about halfway into those posts to mimic the dimensions of the original ferrules.
All of us can predict where the next break will occur, right? That’s OK: I want this to break, instead of wrecking the fender, so the only question is how much abuse those simple joints can withstand. The printing orientation wraps the perimeter threads from the posts around the clip, making it about a strong as it can be.
The ferrules should splay outward by a few degrees to match the angle from the fender to the fork eyelets, but that’s in the nature of fine tuning.
The arch accommodates a strip of double-sided foam tape holding the clip in place along the fender curve, with those cute little hooks capturing the fender to keep the tape in compression:
Tour Easy Front Fender Clip – installed
I really must get some black foam tape …
The picture shows the fender sitting well away from the tire, due to the upper fender mount bending in response to the splash flap snagging on curbs and random debris; the wire stays didn’t seat completely into the posts.
The extender I made during the cracked fork episode remained perfectly straight, though:
Tour Easy – new fork – fender extender
So I re-bent the upper fender mount (not the extender!) to its original angle, thereby moving the bottom of the fender much closer to the tire. Now the stays seat fully, the clip holds the fender firmly in place with no rattles, and it’s all good.
The OpenSCAD source code as a GitHub Gist:
Microscope 60 LED Ring Light Adapter
Posted by Ed in Machine Shop, Photography & Images, Software on 2017-04-25
The Barbie-themed microscope light I built from an angel eye LED ring worked fine for the last six years (!), but a much brighter ring with 60 aimed 5 mm LEDs for $17 delivered from a US seller caught my eye:
Microscope 60 LED ring light – in use
Although this ring looks much more professional, it didn’t quite fit the microscope, being designed for a round snout rather than a squarish one. This snout has a 47-ish mm threaded ring intended for filters & suchlike, so I built an adapter between that and the 60 mm ID of the LED ring:
Microscope 60 LED Ring Light Adapter – top – Slic3r
The ring came with three long knurled screws which I replaced with much tidier M3 socket-head screws going into those holes:
Microscope 60 LED ring light – assembled – top
The part going into the snout threads is deliberately (honest!) a bit small, so I could wrap it with soft tape for a good friction fit. The Barbie Ring didn’t weigh anything and I wound up using squares of double-sticky foam tape; it could come to that for this ring, too.
The adapter features a taper on the bottom for no particularly good reason, as the field-of-view tapers inward, not outward:
Microscope 60 LED Ring Light Adapter – bottom – Slicer
Seen from the bug’s POV, it’s a rather impressive spectacle:
Microscope 60 LED ring light – assembled – bottom
The control box sports a power switch and a brightness knob. Come to find out the ring is actually too bright at full throttle; a nice problem to have.
That was easy!
The OpenSCAD source code as a GitHub Gist:
Cylindrical Cell Adapters
Posted by Ed in Electronics Workbench, Machine Shop, Software on 2017-04-24
An octet of Eneloop AAA cells arrived, I wanted to measure their as-delivered charge (the package says “Factory Charged With SOLAR ENERGY”, so you know it’s good), and discovered I’d given away my AAA cell holders. You can actually get inter-series adapters on eBay, but what’s the fun in that? Plus, I didn’t want to delay gratification for a month; you know how it is.
Soooo:
AAA to AA Adapter – top – Slic3r
It’s basically an AA-size sleeve that fits over the AAA cell, with a lathe-turned brass post conducting juice from the + terminal of the inner cell outward:
AAA to AA Adapter – parts
Not much to look at when it’s assembled:
AAA to AA Adapter – assembled
The AAA cell fits deliberately loose, because this goes into a metal clip holding everything firmly in place for the battery tester:
AAA to AA Adapter – in use
The source code tabulates the sizes of several cylindrical cells, exactly zero other pairs of which have been tested; I expect most won’t work correctly. In particular, the table entries should include the contact button OD and thickness for each cell, so that I can turn out the proper terminal for each pair of cells. If I ever need a different adapter, I’ll beat some cooperation out of that, too.
Discovered I needed an adapter after breakfast, started testing cells after lunch. Life is good!
The OpenSCAD source code as a GitHub Gist:
The original doodle:
AAA to AA Adapter – sketch
-
You are currently browsing the archives for April, 2017
Blowback