Update 17.01.14


Update 17.01.14



Since last update I have made progress on a couple of things. Just before christmas, I had a go at machining the struts for the shroud, and I skimmed some metal off the motor in order to get a true mating surface when I install it into the tailcone. During the christmas holidays, I redesigned the throttle mechanism so that I have an individual knob for speed adjustment and a lever for on/off operation. The last couple of days I have finished the part of the magnetic coupling that is connected to the motor.





This is the CNC mill I used to machine one of the struts for the shroud. It was built as a summer project by a friend of mine who is also a student at NTNU. He was kind enough to program the mill as well as aquire the stock. My job was spraying lubricant and compressed air to keep the cutter cool and lubricated. Unfortunately this was not enough, as some of the cutters became gummed up by aluminium. We finally managed to produce one good part, but had some problems with tool chatter. I have ordered an endmill that is better suited for milling aluminium, and we will try again once it arrives.

On the top picture you can see the strut that holds the handle inserted into the shroud. The bottom picture shows tool chatter. This part will be machined again to hopefully achieve a better finish.

Here is the 500w BLDC motor from goldenmotor inserted in the chuck and indicated so that it will turn true. The purpose of this was to skim the surface of the motor housing so that I will have a true mating surface once I install the motor into the tailcone of the DPV.

Once I returned from christmas holiday, I continued on machining the flange that holds the magnets inside the DPV. Some of the machining on this part was done before christmas, as well as mistakingly drilling the center hole 0.35mm oversize. This is what connects the flange to the motor, and such a large clearance resulted in a very sloppy fit. I fixed this by boring the hole to Ø30 and press fitted a bushing into the hole. The bushing was then drilled and reamed to correct dimensions. Finally, a 5mm keyway was broached.

The flange will have a thrust bearing to take the attraction forces from the magnets. This force is calculated to be around 26 kg, and such a large shear force on the motor bearings would probably cause premature failure.

After the center hole in the flange was repaired, I mounted the part on a rotary table to mill the slots for the magnets. Here I am indicating the part with a test indicator to line up the axes of the flange and the rotary table.

The freshly machined slots. I had to mill the slots in two passes because I only had a 7.5 mm endmill available. DOC was 2.5 mm and the spindle was running at around 1200 rpm.

The flange mounted to the motor. Also note how the outside of the motor has been turned to achieve a true mating surface.

This is a sketch I did of the throttle mechanism. On the top is a knob that can be rotated. The knob houses two magnets whose magnetic field is picked up by a linear hall-effect sensor. As the knob is rotated, the direction and intensity of the magnetic field changes, and a corresponding voltage is transferred from the hall-effect sensor to the controller. On the bottom is a reed contact. This is activated when the when the lever is released. The reed switch activates the braking circuit on the controller, that is, the cutoff that is initiated when the brake lever on the bike is pressed.


The on/off switch is operated with the little and ring finger. Since I mostly dive in cold, norwegian waters, I have to use thick gloves. This design eliminates any need for dexterity, as well as keep the fingers together in order to minimize heat loss.

A 3D model of the DPV. Here you can see the four 12V SLA batteries that will power the DPV. The improved handle is also included.