Actuators, PBSI

While working at PBSI I redesigned our two most-used actuators to make them easier and cheaper to manufacture.

LED Module

We had been using an off-the-shelf LED module for several years, but it was/is stupidly expensive (10+ USD per piece) for what it is (an old LED on a PCB with high thermal conductivity). It is also a generic part that needed extra supporting parts to work into our mounting system.

To replace it, I designed my own LED module on an aluminum substrate with a big copper plane on the top layer, connected to the thermal pad on the LED. The PCB probably isn’t as thermally conductive as the off-the-shelf one because they have their own unique stackup, but I also used an LED that needs less power for a higher luminous flux so we could run it dimmer if I found that it got too hot while testing it.

This board is used in a 5V system so I added two surface mount resistors. I went with 2 instead of 1 so I could avoid using a resistor in a huge package. These are both 1206 and have a sufficient power rating (500mW) for this application (~300mW).

This simplified design with a new smaller jack (the same one used on the new microcontroller board) cut the size and part count of this actuator by more than half. And the new LED module is 80% cheaper than the old one.

Prototype assembly of the new actuator, MUCH smaller than the one shown in the fist pic in this section.

The new actuator assembled and mounted in a sculpture prototype.

When I received the prototypes, I did some stress testing o make sure the LED didn’t get to hot, and that the LED+heatsink didn’t get too hot for the 3D printed (PLA) enclosure. The temperature of the LED itself never exceeded 78 degrees C (measured with a thermocouple because we didn’t have a good thermal imaging camera), and the assembly with the heatsink only reached 55 degrees C, which is below the temperature that PLA begins to soften (60C).

Vibration Motor Actuator, “Moth”

Our other most commonly used actuator is this vibrating frond that we call a “moth”. They are a terrific output device because they create both movement and a pleasant rustling sound when coupled with the right materials, and especially when they are in large groups. And with just a small LED to illuminate themselves, they are one of the most effective actuators we have for creating interesting field behaviour.

The original moth design used a pair of dirt-cheap brushed DC ERM motors from China, but they only had a lifespan of 6-12 months, and the studio’s method of assembling them involved a lot of splicing wires together and wrapping it in heatshrink.

Once we realized how bad the original motors were (when they started failing in an older sculpture in Indiana), I found some better motors from a reputable British company. They were bigger than the old ones so we modified the design of the actuator a bit, and I made some changes to improve the manufacturability and reliability of the motor assembly and wire harness. I made a simple PCB with through hole pads for soldering the motor leads, the LED, and harness leads, which eliminated all the wire splicing that was done previously. This worked well to make this new design a drop-in replacement that I installed in the Indiana sculpture (Amatria).

A look inside the heatshrink of the motor assembly.

When we made the jump to the new hardware system, I also took the opportunity to simplify the design. One of these new motors was easily as powerful as two of the cheaper ones, so I reduced the actuator to a single vibrating element and integrated all of the components into a single PCB. These changes reduced assembly time by almost 80%.

Assembly before encapsulation with heat shrink.

← Return to Project List