My first project on this site, how exciting! Let’s get started..
The problem goes as follows; the current place I live at has ample room to set up a workbench, but the lighting here is utterly horrendous. The normal solution would be to just buy a desk lamp, but there’s 2 problems with that:
- Space on my workbench is in short order, and a standard light would take up precious real estate.
- It’s terribly, terribly boring..
That slanted roof really limits my options, and I’m pretty sure my landlord wouldn’t appreciate me drilling into the walls/roof, so mounting a surgical light to my ceiling is out of the question.
Okay! Now we know this isn’t impossible, let’s get planning!
Well, actually, no.
Let’s just get buying, and plan around the stuff we can find..
What we got is as follows
- 1 (one) 200 x 2 x 0.2 cm aluminium profile (very flexible!)
- 1 (one) 24v high CRI led strip; natural white
- 1 (one) 100w 24v power supply of dubious quality.
- 1 (one) 12v/24v led dimmer
- Some PLA filament
- An assortment of nuts and bolts.
Fantastic! That flexible aluminium profile is going to work perfectly, we just need to clamp it to both sides of the desk, and the profile will take an arch shape all by itself.
Now, for some reason I really wanted the arch to be a perfect circle.
Don’t ask me why, I’m no engineer. However, this should be perfectly doable, the only critical part here is the angle at which the profile is clamped to the desk.
Assuming we have a circle, we can say our desk is the chord of this circle, and our arch light is the arc.
Considering we know the desk is 150 cm long, and the aluminium profile is 200 cm long, we can figure the rest out with math.
Actually, I just plugged the numbers into an online calculator, and apparently the resulting diameter of the big circle is 156.6 cm. Neat!
To get the actual angle I just plugged the measurements into Fusion 360 et voila! Our bracket needs to clamp the profile at an angle of 71.4 °.
Trigonometry without doing a single calculation.
Now we’ve got Fusion 360 open, it’s time to start designing those brackets. Nothing very special about them. I just designed to press-fit on the side of the desk, hence the little bump on the lower prong. I also added a little hole in the bottom, so is could tie them together below the desk should the fit be too sloppy.
The final result.
The profile will just bolt to the outside of the bracket with two M5 screws.
Great! Now that those are printing, let’s have a look at that dimmer, we still need to find a proper way to attach it to the side of the desk.
Alright, so this needlessly big housing contains only a teeny, tiny circuit board and a single potentiometer. Kinda makes me wonder if it would be possible to integrate directly into the bracket. Back to Fusion!
This is what I came up with; widened the thing by a cm, added a recess to the bottom, and created a channel for the potwires to run through.
On the back a hole was added to screw the dimmer board to.
So, with the brackets done it was time to prepare the profile.
I simply marked the correct hole position, drilled a pilot hole and went afterwards at it with the 5 mm drill bit.
Some WD40 really helped my shitty drill get trough the aluminium without much problems.
Now, with that out of the way it was time to move back to electronics.
Now, let’s briefly touch on these lights.
These are what are called “CRI 90+” leds. Meaning that these leds have a relatively good “Colour Rendering Index”.
The CRI, ranging up to 100, defines how well colour is visible under a light source, compared to the sun.
My first plan was to go with some fancy RGB leds, so I could play with the warmth of the light later. However, as it turns out, RGB leds are, ironically, pretty bad at revealing the colour of the objects under them, and thus not very enjoyable to work under. So, high CRI leds were the way to go, even if they were just boring ol’ white.
To simplify it to terms I understand myself; light can appear white by just stimulating the three types of cone cells in our retinas. That’s just three wavelengths the lights needs to contain, the rest can be empty for all our eyes care. However, to interact with coloured objects, you need light with the specific wavelengths the pigments interact with. If that wavelength is not in the emitted spectrum, then the pigments cannot properly interact with the light and the object gets a weird, unnatural colour. Light sources with a low CRI (such as most LEDs and CFL tubes) can appear white to us, as our eyes are just happy with as long as our cone cells are stimulated, but objects with colours that lie between the emitted wavelengths cannot bounce the correct wavelength of light back (as it is absent) and thus appear to have no colour, or the wrong colour. Light sources with a high CRI (such as the sun, or incandescent light bulbs) emit a very broad spectrum of light, making the chances of correctly interacting with a pigment much higher.
The reality is much more complex, of course. But anyway..
While we were talking about lights I finished mounting the dimmer circuit to the bracket. The board just screws in place and sits flat against the back of the bracket. I might consider making a little cover for it, but for now it’s fine.
The brackets get attached to the aluminium profile.
The LED strip gets stuck to the aluminium profile, which acts both as the structural support and as a heatsink.
And the final solder connections are made.
Great, now it’s time to get this puppy mounted!
I’m positively surprised with how well the brackets clamp to the desk, I didn’t even need to shim them or anything! What I’m guessing is that the springiness of the aluminium profile in causing some sort of rotational force on the brackets, which locks them in place.
Boy! What a difference does this light make, I can finally see what i’m doing!
Well, that was pretty fun. Productive too..
I still need to mount the power supply somewhere out of sight, but that’s a problem for another time.
Right now I’m gonna bask in the light of my new desk light..