It’s been a while.

My CNC mill has seen some action recently, doing work for my co-op job and some side jobs.

I’ll start by explaining the project since there is nothing confidential about it, it’s a low frequency desktop shaker table being made on the cheap. How cheap you ask? Scrap bin parts cheap. Low frequency shakers are generally used for simulating earthquakes with high accelerations. They are also massive in terms of size and weight becasue the models they shake are thousands of kilograms. So I present to you, the economy desktop low frequency shaker table:

The shaker itself is made of a linear actuator, servo motor, and scrap aluminum plate. The linear actuator/motor combination comes from the Animatics Smartmotor line. To transform the pile of scrap metal into a functional instrument, there is control software and electronics, in the form of a LabVIEW interface and data acquisition system packed in a nice enclosure as seen below.

Basically all the machining was done on my CNC mill for the reason that there was a lot of patterned holes and whatnot. Here we trim down a plate to be square for the shaker platform:

After trimming a series of thru holes are drilled for mounting test samples to the shaker, held down on a precision ground piece of MDF. CNC drilling beats laying out and center punching any day.

The electrical enclosure front panel was the next most difficult part to machine. I needed a series of holes for mini-DIN connectors, strain relief nuts, and an emergency stop button. I needed a number of different hole sizes so it made sense to cut them out on the CNC mill. To avoid scratching the paint off the enclosure, I used a folded up sheet of paper.

In order for the shaker system to be useful, we also needed to log measured accelerations. The shaker is designed for relatively low accelerations, in the order of 1g or so, which made finding an appropriate accelerometer difficult. For analog accelerometers, sensitivities nearing 1v/g are available for ridiculous costs relative to the budget of the shaker project, ($1000+) usually due to the fact that they are precision instrumentation. To keep costs down, I designed a breakout board for use with some MEMSIC MXR7900 accelerometers, good for +/- 1g in 2 axis, sensitivity of 0.9v/g, and simple 0-5v scale centered at 2.5v. the board is about 1″ square with 4 mounting holes for #6 bolts and uses all surface mount components including a locking connector.

The boards were made by Seeedstudio, and they arrived in 2 weeks. Since their board size was up to 5cm square, I patterned 4 boards per board and cut them down myself with a sheet metal shear. The design itself was done in Diptrace, and it was the first time I designed a PCB.

As I said before, this is a single sided board with only surface mount parts. This would allow the bottom of the board to be mounted to metal surfaces without shorting anything out. It also gave a nice flat surface to mount with. They’re working great so far and I have no complaints given that 5 accelerometers were made for 1/8th the price of an industrial one.

In addition to my work-work, I was also asked to help make a non-magnetic, non-EMI generating beaker stirrer for a start up in Waterloo. They were doing very sensitive bio-sensor testing and a normal magnetic stir plate was causing too much noise. They also needed this fast to do testing with. After some thought and iteration, I came up with rotating the entire beaker back and fourth with a hobby servo I had on my desk. This was the original plan:

The whole concept was simple so it could be manufactured in a few hours. To further simplify the design, the final product ended up being a circular platform fully supported by the servo and an Arduino soldered on to some perf board to control the oscillating motion.

The images below show a circle being cut from some leftover aluminum and a pocket being made for mounting to a servo horn. With my very precise 2×4 work holding jig, I also managed to break a 1/4″ endmill when chips welded to the flutes.

Good lighting ensures fewer screw ups and more shiny machining marks.

And the final product moving a can of WD-40:

That’s it for now.