Linear Actuators Redux

[Dumpsterholic]

Hi all:

My diligence in scoping out a nearby Dumpster has paid off in the form of a complete controller circuit for a "Megamat MBZ 50899" linear actuator. Along with the LA itself, I also netted an "Ultra Control Box" and a "Pride" thumb switch. These circuits are used in power-lift recliner chairs for handicapped people, and every now and then a store close to where I work will throw away entire units: chair, control box and power switch. The LA's themselves and the controllers cost a bundle of money new, so when I finally got my unit home and tested it for functionality (Yay! It works just fine!), I was curious to find out what kind of fancy components might be inside.

The answer: They comprise nothing more exotic than a transformer, some chunky diodes, a capacitor and a few resistors, plus a pair of relays. For some reason, there is also a pair of 9V batteries, even though the control box runs off of AC mains power. I suspect that planned obsolescence may be at work here, but so what? I'm not the one ripping off handicapped people (or the insurance companies that pay for their care); I just want to learn how to use LA's in microcontroller-based projects.

So, here's my question: Given that a rectifier circuit and a couple of relays (one for forward, one for reverse?) are obviously sufficient to power a LA that can exert a push/pull force of 4,000 Newtons, why is there a need for more elaborate motor-driver circuitry in other applications? I suspect the answer is twofold. First, the units I scavenged have a very limited duty cycle of two minutes "on" and eighteen minutes "off", so in an application that required faster response times, the relay-based approach would not work because the relays themselves would quickly burn out.

Second, these brute-force controllers make no provision for velocity or position control: The switch either makes them go forwards or backwards at one set speed, with no way to stop at a preset location. If an application required speed adjustment, then additional circuitry would be required to supply either a variable DC voltage or a PWM signal, and in order to tell the LA to stop at some desired position in the middle of its travel, other circuitry would be necessary to provide feedback to the MCU controlling the power supply to tell it when to turn on and off.

Am I on the right track? It seems to me that with just the circuitry I have, all I would need to do to build a "Star Trek" style IR-activated sliding door is a microcontroller and a Darlington pair to fire the relays, plus whatever IR sensors would be required to tell the MCU when to activate the LA to do its thing. (Not to mention the mechanical parts: door, frame, etc.) Needless to say, there are many other possible uses for this kind of circuitry....