Ed Nisley's Blog: shop notes, electronics, firmware, machinery, 3D printing, and curiosities
Just to see if all the laser parts once again fly in formation, I cut a defunct cotton shirt into shop wipes:
Which worked as before:
Not that I need more wipes, but in this case the process is more important than the product.
Looks like I can start using the laser cutter again … whew!
A couple of test shots to verify the move hasn’t jostled the laser mirrors too far out of alignment:
The overlapping scorches on the left happened at the Mirror 3 position with the laser head at the far left and near right positions. Not quite as accurate as immediately after I overhauled the beamline, but close enough.
The pair of dot + disk scorches on the right show the beam position on the platform at the focus point and 20 mm below. The red-dot pointer definitely traces a wavering path as the platform goes down, suggesting the leadscrews may have taken a sideways jolt during the laser’s trip down the basement stairs and are now distinctly angled in their guides, but it’s good enough for my simple needs.
Looks like the laser survived the move pretty much intact!
The honeycomb platform in my OMTech laser cutter was secured by a pair of M4 screws passing through the surrounding frame into a pair of nuts requiring considerable contortion to install. As a result, I tended to use the screws as locating pins by just dropping them into the holes, which didn’t prevent me from jostling the honeycomb out of position on a few occasions.
With everything torn down as part of the move, I drilled out the holes in the frame and installed a pair of M4 rivnuts:
The scar around the hole in the honeycomb came from the factory; I have no idea what they were doing to cause that much wear.
Anyhow, installing the screws now requires zero contortionism and they locate the honeycomb much more securely.
I should conjure knobs for the top of the screws to eliminate the need for a hex key, although that’s definitely low on the task priority list.
The previous Basement Laboratory generally stayed above 60 °F = 15 °C, so I set the LightObject water chiller’s low-temperature alarm accordingly.
Having reached the point where I can set up the laser in its new home, I connected the chiller tubes, filled the reservoir with distilled water (and a squirt of algaecide), connected the alarm wiring, turned it on, and had the cool water trigger an alarm:
Which was relayed to the controller:
Silencing the chiller’s alarm clears the error indicator in the controller, so it’s possible to Fire The Laser with too-cold water if necessary.
As with the previous icemaker chiller, plotting the water temperature as a function of time shows the pump adds some energy as it moves the water around the loop:
The gap in the data shows where I had a few other things to do, but the exponential rise is obvious. The chiller compressor starts at just over 21 °C and stops at just under 20 °C, so the exponential curve had gone about as far as it could go.
The numbers in the upper right of the plot give the weight of:
The figures in the bottom mash the initial slope of that curve together with the weight of the water to find the 21 W required to heat the water at that rate, with a bank shot off British Thermal Units because why not.
A Kill-a-Watt meter shows the Q600 chiller draws 36 W with the pump running, which includes the controller and a column of blue LEDs behind the water level tube.
The pump (in the lower left) isn’t exactly water-cooled, but it’s not losing a lot of heat through that foam wrapper and maybe most of the heat really does come from the motor:
The basement temperature will rise as Spring becomes Summer, so the chiller will start working right away, and it’ll definitely get more exercise when the laser starts cutting again.
It turns out that if you drop a heavy sheet of laminated cardboard in exactly the right spot, you can shear the pot metal handle right off a two-step folding step stool:
I mean, it’s just a perfect target:
It was a clean break leaving gritty surfaces ideal for JB Weld epoxy and a clamp:
In truth, using epoxy in tension isn’t a good idea, but this is light duty and the repair ought to be good for a while.
Now, as to why I was standing on a two-step ladder fiddling with a heavy sheet of laminated cardboard, that story must wait for a while …
Every tub & shower looks like this after a decade or so:
Go look carefully at your bathroom if you don’t believe me.
Tubs have a raised lip around their perimeter, but our downstairs bathroom had a caulk crack that routed water running down the wall under the tile, over the lip, and onto the subfloor beside the tub. This had been going on unnoticed for years, but we apparently take showers differently enough to put a puddle of water on the basement floor.
Some exploratory surgery revealed a patch of rotted subfloor (which is why we know it was an ongoing problem), but no structural damage. A few hours of tedious razor knife and hook work extracted the old caulk, after which squirting new caulk took almost no time at all:
I screwed a small fan across the subfloor opening to pull air across the wet area:
A few days dried things out nicely, so I can proceed with a project involving the adjacent shower stall, about which more later.
They don’t install drain pipes like that any more! Judging from the many scorch marks on the joists, the plumber had considerable difficulty keeping enough heat on the fittings for good solder joints.
The Kevlar cord on Mary’s bike survived the crash without breaking, but it was badly scuffed and holding on by only a few strands. Unlike in years gone by, Kevlar cord is now cheap & readily available, so I decided to restring the thing:
The cord path isn’t at all obvious, even given the smudges on the seat struts:
Pictures of the original cord as installed at the Easy Racers “factory” served as guidance:
The knots joining the cord at the top, taken juuust before I pulled the right knot apart:
Those are in addition to my Tour Easy a few feet away, but you can never have enough pictures.
A 3.5 meter cord will be plenty long enough and marking the midpoint simplifies equalizing the two sides. The cord crosses the seat frame at the bottom from the lower guides, although I’m reasonably sure it wouldn’t matter if you ran separate lengths up the two sides with a knot in the lower guide.
The new cord claims to be 1000 pound test (200 pound working), but the vital dimension is its 2.6 mm diameter to match the OEM cord. It does not claim to be UV stabilized, which may turn out to be a problem over the course of a few years.
Tightening the cord proceeded as before and a test ride indicated the installation was all good.
The Smell of Molten Projects in the Morning 