A Novel Desk Lamp Requires a Novel Design Process

The design of The Armadillo Desk Lamp started off a little bit differently. Typically when I get to the build phase on a project, my pencil is down and I’m mostly executing the plan. But I had spent months drawing in my sketchbook and the design just wasn’t coming to me.

I grabbed my ball-peen hammer and a piece of stainless and started shaping a lamp shade. I was pretty happy with where it ended up - the heat turned the stainless a nice bronze color and it felt good in my hand. I stuck a bulb in it, picturing it as a lamp of some sort and it seemed to fit the bill. But I still didn’t know what kind of lamp it wanted to be. It would make a nice sconce surely, but it would also make a nice bedside reading lamp on an arm extending off of the headboard. I wanted a more technical challenge though that neither of these options offered. Ultimately I decided that a desk lamp would require a bit more mechanical problem solving. And the lamp shade that I had just hammered out also felt good in the hands so it seemed like a shame to subject it to the solitary existence of a wall sconce.

So putting the cart before the horse and starting to build without a mature design had helped to create some momentum and bring some clarity to the design direction. I now knew I wanted to build a desk lamp that incorporated this shade I had built, I just needed to fill in the details between the wall outlet and the shade. I had designs for all sorts of crazy multi-hinged joints, counter-weights, and X/Y tracks to position the lamp. But the simplest design always wins. The arm had to be positionable up and down and left to right but the ability to rotate the shade about the axis of the arm was also important. A spherical joint is the simplest way to create so many degrees of freedom, although traditional spherical joints like rod ends are not necessarily the simplest type of joint to manufacture. I started to look at some rod ends and realized how they are made - a sphere is placed inside of the short cylindrical housing of the rod end, and then the cylindrical housing walls are very accurately deflected to capture the sphere. I knew from the time that I spent at Apple designing for high volume production that this process would require many sacrificial parts before the process was tuned just right.

Taking a step back, I realized that what made the rod end ball joint problem so challenging was the fact that the ball had to be constrained to handle loads in all directions. But my design only required load in one direction - the direction of gravity. So that meant that as long as the ball had enough engagement that the arm couldn’t easily be knocked off the base, it should work perfectly well to simply allow gravity to hold the spherical joint together. It also helps that there is a substantial amount of weight from the lead counter-weight holding the joint together. The other benefit to using gravity to hold the joint together is that gravity will hold the components of the joint tightly together even as the joint starts to wear over time.

The amount of friction that the spherical joint had to create is also important. Too much friction and the ball would likely gall, or not move smoothly, or perhaps bind. Too little friction and the desk lamp would not stay put where you want it. Step 1 was to determine how much friction would be right. I set my target friction by the amount of force that I wanted the user to exert. I simply pressed on a scale with the amount of force that I felt would be acceptable to break static friction and move the lamp into position. Around 3 lbs felt right to me. Since the arm is so long, the friction torque required at the ball to create 3 lbs is actually quite high. Luckily though the amount of friction torque can be finely tuned by playing with the geometry. The equation for F_user (force exerted on the lamp by the user) below helps understand how certain changes in geometry can be used to tune the force. Decreasing the angle theta has the largest effect on the friction torque, and is easy to do without impacting aesthetics. After that, we can see that increasing the radius of the ball or the weight of the arm will also increase friction. And of course decreasing the length of the arm will increase the amount of force that a user needs to exert. As the angle (theta) gets more and more shallow, the amount of force transferred from the surface of the ball to the surface of the cone gets larger. A 2° angle created the high forces necessary to hit the torque target but I also needed to check the contact stresses between the sphere and the cone. If these contact stresses are too high, the joint could bind and quickly become damaged.

Armadillo desk lamp user force and ball joint friction calculations

One of the other unique design challenges was figuring out how to counter-weight the arm. It’s important to be pretty precise with the weight balance. Calculating the weight balance in CAD will get close but parts come out with slight variations which can quickly throw things off. Because of this I purposefully oversized the weight so that I could drill balancing holes once everything was assembled. The amount of counter-balance required to keep the arm in balance was substantial and it quickly became apparent that I was going to have to use lead in order to hit the counter-balance target. A piece of stainless steel flatbar was rolled into an aesthetic shape first. I then welded some little tabs to the interior surfaces of the stainless steel flatbar so that when I cast lead into the shape, the tabs would be fully captured and there wouldn’t be the chance that the lead could wiggle itself free later. The parts were then cleaned up and the electrical components integrated to get the final weight. With everything assembled it was time to start drilling consecutively larger holes to get the lead counter-weight in balance. Once everything was in balance, I covered up the balancing holes with two little decorative pieces of brass. The lead makes the piece more intriguing - the large grain structure, the meniscus around the perimeter, the flow lines on the far left side of the casting.

Now just need some finishing touches. I cut out the brass base from a piece of 3/16” plate and machined it flat on my lathe so that the lamp wouldn’t wobble. Then I machined a custom knob for the lamp switch on my lathe and pressed it into place.

The result - a truly one of a kind hand-made desk lamp with a good design backstory. I found this project very satisfying to build because it has a few elements that make it novel - the cast lead counter-weight, the spherical joint, and the bronze colored forged lamp shade. The design process was also novel (to me at least) because very rarely do I start a project with an incomplete plan.

Thank you for reading, and I hope this post peaked your interest. As always if you have any feedback, questions, or comments please head to the contact us page and send me a note, I would love to hear from you.

-ODH

A Novel Desk Lamp Requires a Novel Design Process

Engineering the Customer Experience

The Product Lifecycle