Dehumidifier Scrapping

Dutchess County has another Household Hazmat / Electronics Disposal Day coming up, so I harvested some useful parts from the three dead dehumidifiers lurking under the bench.

The (perfectly good) blower motor in one unit lives inside a convenient plastic housing:

Scrap Dehumidifier Blower Motor - housing
Scrap Dehumidifier Blower Motor – housing

It’s sitting on three foam pads hot-melt glued to three wood blocks cut to fit inside three convenient molded features, making it nice & quiet & stable.

The motor uses a nice polypropylene run capacitor:

Scrap Dehumidifier Blower Motor - 6 uF cap
Scrap Dehumidifier Blower Motor – 6 uF cap

Which is also perfectly good:

Scrap Dehumidifier Blower Motor - 6 uF cap test
Scrap Dehumidifier Blower Motor – 6 uF cap test

The motor includes a wiring diagram:

Scrap Dehumidifier Blower Motor - wiring diagram
Scrap Dehumidifier Blower Motor – wiring diagram

I lashed it together with a chopped-off IEC cord, because the stock dehumidifier cords are just way too stiff. The motor and blower originally pulled air through the dust filter, the condenser, and the evaporator, before blowing it out the side, so it’s running pretty much unloaded. A quick test shows there’s not much difference between the high and low speeds:

  • High: 1050 RPM, 80 W, 12.5 m/s air flow
  • Low: 1000 RPM, 77 W, 11.7 m/s air flow

Low speed seems slightly less noisy, but the wiring now has insulated QD connectors just in case I ever want to run it at full speed.

For whatever it’s worth, the most recent dehumidifier failed one year into a two year warranty, but the company decided it was simpler to just refund the purchase price than to replace the unit. It seems the “sealed system” inside loses its refrigerant after a year and there’s no practical way to seal a small leak and recharge the system; unlike an automotive air conditioner, the tubes are soldered shut after the initial charge.

They all sport Energy Star badges, but throwing away the whole damned thing every year or two tells me we’re not measuring the right values. Obviously, somebody could make a worthwhile dehumidifier, but as of now Frigidare, GE Appliances (sold to Haier), and Danby are on my shit list. Next year, I expect to add HomeLabs to the list, because the dehumidifier is identical to the Danby unit (and, ah-ha comes with a 2.5 year warranty). They’re all made by Haier (or another Chinese factory) and nobody applies any long-term QC to their products.

Painting By Numbers, Redux

Five years later, the digits I painted with Rust-Oleum Rusty Metal Primer have weathered pretty well, while the original ink has fallen off the retroreflective sticker:

Mailbox numbers - original vs primer
Mailbox numbers – original vs primer

As before, I wiped off the crud with denatured alcohol and painted neatly inside the lines. The other digits on both sides still look as good as the day I painted them, with only a few bubbles and nicks.

Memo to self: Next time, buy a big sheet of 3M retroreflective film, make a stencil by vinyl cutting, paint the entire number in one shot, and be done with it.

Drill Press Vise Table Refresh

I built a small plywood work table for the drill press:

Drill press - scarred vise table
Drill press – scarred vise table

Obviously, that was a long time ago. It’s a plywood scrap with a small cleat screwed to its bottom, upon which one can position / clamp / hold / finagle smallish workpieces without worrying about drilling into the surface.

The most recent batch of aluminum backing plates prompted me to finally replace that relic:

Drill press - new vise table
Drill press – new vise table

The mill vise under the plywood grips the cleat and the whole affair rides on a Sears “Drill Press Milling Attachment Stock No 27585” which is basically a simple XY table with hand dials. It’s not rigid enough for actual milling (which you should never do on a drill press, anyway, because the end mill will pull itself out of the Jacobs chuck), but it’s good for tweaking the position before you drill something.

One should never hand-hold workpieces while drilling.

Don’t do as I do, do as I say. OK?

Soaker Hose Clamps

Having figured out the geometry for two- and three-channel soaker hoses, I cranked out more clamps:

Soaker Hose Clamps - production
Soaker Hose Clamps – production

Actually, those are the remainder of two production runs devoted to reducing the amount of water sprinkling the garden paths. A 50 foot hose runs along both sides of one 14 foot bed, crosses the path, then continues along the adjacent bed. The hoses have (deliberate!) sprinkler holes along their porous rubber body and sometimes the layout puts a hole where it waters the path.

The blue silicone rubber strips provide a bit of sealing to prevent the absurdly high pressure water from streaming through the orange PETG clamps. It’s OK if the clamp leaks, but less flow is better!

I’m getting really good at making those aluminum backing plates and, in fact, I think it’s faster to run the blanks past the disk sander, then drill the holes, than to CNC-machine them. Could be wrong, but Quality Shop Time is not to be sniffed at.

Solid Modeling: Support Puzzle

I’ve been putting this type of support structure inside screw holes & suchlike for years:

Browning Hi-Power Magazine Block - solid model - Generic 1 - support detail
Browning Hi-Power Magazine Block – solid model – Generic 1 – support detail

It’s basically a group of small rectangles rotated around the hole’s axis and about one thread thickness shorter than the overhanging interior.

I’ve found that incorporating exactly the right support structure eliminates Slic3r’s weird growths, eases removal, and generally works better all around.

So doing this for the baseplate of the Glass Tile frame came naturally:

Glass Tile Frame - octagonal support
Glass Tile Frame – octagonal support

This OpenSCAD snippet plunks one of those asterisks in each of four screw holes:

  if (Support)
      for (i=[-1,1], j=[-1,1])
          for (a=[0:45:135])
                translate([0,0,(Screw[LENGTH] - ThreadThick)/2])
                  cube([Screw[OD] - 2*ThreadWidth,2*ThreadWidth,Screw[LENGTH] - ThreadThick],center=true);

The “cubes” overlap in the middle, with no completely coincident faces or common edges, so it’s 2-manifold. Slic3r, however, produces a weird time estimate whenever the model includes those structures:

Slic3r - NaN time estimate
Slic3r – NaN time estimate

NaN stands for Not A Number and means something horrible has happened in the G-Code generation. Fortunately, the G-Code worked perfectly and produced the desired result, but I’m always uneasy when Something Seems Wrong.

Messing around with the code produced a slightly different support structure:

Glass Tile Frame - quad support
Glass Tile Frame – quad support

The one thread thick square on the bottom helps glue the structure to the platform and four ribs work just as well as eight in the octagonal hole:

  Fin = [Screw[OD]/2 - 1.5*ThreadWidth,2*ThreadWidth,ScrewRecess - ThreadThick];
  if (Inserts && SupportInserts)
      for (i=[-1,1], j=[-1,1])
        translate([i*InsertOC.x/2,j*InsertOC.y/2,0]) {
          for (a=[0:90:360])
                translate([Fin.x/2 + ThreadWidth/2,0,(ScrewRecess - ThreadThick)/2])

Which changed the NaN time estimates into actual numbers.

One key difference may be the small hole in the middle. The four ribs (not two!) now overlap by one thread width around the hole, so they’re not quite coincident and Slic3r produces a tidy model:

Glass Tile Frame - quad support - Slic3r
Glass Tile Frame – quad support – Slic3r

The hole eliminates a smear of infill from the center, which may have something to do with the improvement.

In any event, I have an improved copypasta recipe for the next screw holes in need of support, even if I don’t understand why it’s better.

Glass Tiles: Matrix for SK6812 PCBs

Tweaking the glass tile frame for press-fit SK6812 PCBs in the bottom of the array cells:

Glass Tile Frame - cell array - openscad
Glass Tile Frame – cell array – openscad

Which looks like this with the LEDs and brass inserts installed:

Glass Tile - 2x2 array - interior
Glass Tile – 2×2 array – interior

The base holds an Arduino Nano with room for wiring under the cell array:

Glass Tile Frame - base - openscad
Glass Tile Frame – base – openscad

Which looks like this after it’s all wired up:

Glass Tile - 2x2 array - wiring
Glass Tile – 2×2 array – wiring

The weird colors showing through the inserts are from the LEDs. The red thing in the upper left is a silicone insulation snippet. Yes, that’s hot-melt glue holding the Arduino Nano in place and preventing the PCBs from getting frisky.

Soak a handful of glass tiles overnight in paint stripper:

Glass Tiles - paint stripper soak
Glass Tiles – paint stripper soak

Whereupon the adhesive slides right off with the gentle application of a razor scraper. Rinse carefully, dry thoroughly, and snap into place.

Tighten the four M3 SHCS and it’s all good:

Glass Tile - 2x2 array - operating
Glass Tile – 2×2 array – operating

So far, I’ve had two people tell me they don’t know what it is, but they want one:

Glass Tile - various versions
Glass Tile – various versions

The OpenSCAD Customizer lets you set the array size:

Glass Tile Frame - 3x3 - press-fit SK6812 LEDs
Glass Tile Frame – 3×3 – press-fit SK6812 LEDs

However, just because you can do something doesn’t mean you should:

Glass Tile Frame - 6x6 cell array - openscad
Glass Tile Frame – 6×6 cell array – openscad

Something like this might be interesting:

Glass Tile Frame - 2x6 cell array - openscad
Glass Tile Frame – 2×6 cell array – openscad

In round numbers, printing the frame takes about an hour per cell, so a 2×2 array takes three hours and 3×3 array runs around seven hours. A 6×6 frame is just not happening.

The OpenSCAD source code as a GitHub Gist:

Screw Thread Measurement

While I was cutting threads for the Floor Lamp poles, I tried measuring my progress over wires:

Floor Lamp - tube fitting - thread measurement
Floor Lamp – tube fitting – thread measurement

Those are three lengths of music wire, slightly bent from their storage roll, held in place with a precision clamp metric micrometer. Given the crudity of the setup, the uncalibrated wire diameter, and my lack of thread-fu, the results came out both close and unconvincing.

A set of real thread measuring wires being cheap & readily available, I’m prepared for the next time around this block:

Thread Measuring Wires - eBay set
Thread Measuring Wires – eBay set

The 185 mil “wires” (they’re all allegedly ground rod) will let me cut threads matching things like a Jesus nut; they’re suited for 3 TPI / 8 mm pitch screws. Mostly, wires from the front row will be all I ever need.

Which look like this in action:

Thread Measuring Wires - eBay setThread Measuring Wires - detail
Thread Measuring Wires – eBay setThread Measuring Wires – detail

The black doodad (the set includes half a dozen for all the wire sizes) fits over the micrometer anvil and holds two wires betwixt anvil and screw, leaving me to manipulate the screw, the third wire, and the micrometer with my remaining hands. Hence the vise holding the micrometer, which is known to be Very Bad Practice.

From the side:

Thread Measuring Wires - overview
Thread Measuring Wires – overview

All of the smaller wires measure 0.5 mil too thin, which is likely due to my lack of calibrated measurement equipment:

Thread Measuring Wires - scant 24 mil
Thread Measuring Wires – scant 24 mil

The few thread pitch diameters I measured also came out slightly too small, again likely due to calibration and screw tolerances.

The LittleMachineShop description of measuring threads over wires seems entirely adequate.

To forestall link rot, a slightly rearranged version of their tables of wire constants:

Thread Wire Measurement Constants
Thread Wire Measurement Constants

The lower table has metric thread pitches with the wire sizes in inches.

You measure the distance over the recommended wire (in inches or millimeters, as appropriate), subtract the constant, and get the pitch diameter in the same units. Conversely, add the constant to the desired pitch diameter to get the target over-wire distance, carefully cut the thread until it measures a bit less than that, back up sixty seconds, and cut it spot on.

Verily, it is written: there is no UnDo key (⎌) in machine shop work.