Saturday, 21 September 2013

Thruster Oil Temperature (and Volume) Increase

After a small tweak to the pressure tolerant (hopefully, yet to be tested) brushless DC motor drive PCB I'm getting very reliable use out of the thruster now.

So in view of this I thought I'd now put the PCB back into the thruster and fill it with oil.

I was testing the PCB outside the thruster with the motor control lines, three phases, power to the hall effect sensors and feedback from the hall effect sensors running to the motor in the thruster through the flexible tube and placing the PCB in a small oil filled container outside of it.

Now with the PCB in there, there's just power (24VDC) and communication (RS485) from my laptop travelling down the flexible tube.

So with my load cell (fish scale :)) and temperature sensor in situ it was time to do some testing!

The other thing I noticed is, as I was using a PVC pipe originally for the thruster housing from the local plumbers merchants, but this didn't seal very well and it wasn't quite the right size, so I now have an Acetal tube which I made on my lathe (which is a lot nicer than PVC to machine, as PVC is a bit "stringy"), is that I can see any air bubbles trapped in the thruster through the housing.

I'm not sure if this is good or bad thing as I haven't pressure tested the housing yet (probably the next thing i'll do), but we'll see.

Anyway I could purge the thruster of air by jiggling it around, see how many bubbles were in there and letting the air bubbles run up the flexible tube :)

So here's a picture of the setup I used today.

I let it run for three hours at full load and the oil temperature in the flexible hose/tube rose

The load cell is measuring about 1.82kg of force and the temperature sensor (the white thing with a probe sticking into the flexible tube) initally started off about 17 degrees C.

After three hours the temperature rose to about 19 degrees C and the thruster housing was definitely a lot warmer to the touch, but not hot. The Acetal is transferring some of the heat from the brushless DC motor control PCB and then the oil into the water.

The thing that was very noticeable though, which I didn't expect and thought was initially a leak was the oil moving up the tube as the temperature rose! I had to empty some out as the test went on initially. The reason I know it wasn't a leak is when the oil cooled down the level dropped back down in the flexible tube and the oil remains clear not cloudy.

I intend doing a few more detailed measurements on the oil movement in the flexible tube when I test the thruster next time, but the main thing is I now know that the ESC PCB is happy to carry on turning the motor and propeller, under full load (about 3 amps) with oil in the housing (as it will be in the final ROV) for over three hours, as compared to the my last test with the BLDC control PCB being external to the thruster and in the small oil filled container (for initial testing purposes), so it's a good result all in all :>

Saturday, 17 August 2013

Two brushless DC ROV thruster motors controlled independently over an RS485 link

Having modified the LabVIEW GUI (Graphical User interface) code and the BLDC (Brushless Direct Current) ESC (Electronic Speed Controller) firmware I can now control multiple motors independently over the same RS485 link using addressing.

The first part of the meassage on the serial link specifies the address, with a value from 0 to 15 and only the ESC that has that address, based upon four jumpers on the ESC PCB, will then use the following value in the message to control the speed and direction of the brushless DC motor it is attached to. If it isn't the right address it then simply ignores the message, until its message comes along.

A message is transmitted from the GUI software to one ESC, then the other, then back to the first one, continuously, at about 50ms (milliseconds) per message.

Here's a short video of the GUI and the motors in action.

Sunday, 11 August 2013

Two thruster motors controlled by RS485

Not exactly a giant leap forward but it's all about slow deliberate steps for me, with as much testing as possible in between.

So here is a video of two thruster motors, two BLDC motor controllers made by my own fair hand and a LabVIEW programme basically telling them how fast to go and in what direction over an RS485 link.

The oscilloscope at the end is just showing the communication on the RS485 channel.

Next step is to provide independent control to each motor so they're not quite as synchronised as they are currently :)

Thursday, 18 July 2013

Temperature of FETs

I thought I'd check out the temperature of the FETs on my ESC whilst submersed in oil (i.e. the way they will be running inside the thruster), driving the thruster at full load (about 3 amps), just to see if the oil dissipates most of the thermal energy, and it looks quite encouraging.

The temperature probe resting directly on the FET measures just over 63 degrees C (the ambient is about 23 degrees C today) ..... :)

EDIT: Actually up to 73 degrees now after about 2 hours continuous running but still not bad.

Saturday, 23 February 2013

Too Much Thrust (if that's possible)

Well, I never really realised when I referred to churning up some water in my previous post what level of churning would eventually ensue!

Now I have digital, rather than analogue control over the thruster, the amount of thrust I'm seeing has dramatically increased, check out the video .....

I definitely need a bigger test tank!

Saturday, 16 February 2013

Thruster Under RS485 Control

I've recently received and populated a new thruster control PCB (ESC), which seems to be working well.

Here's a picture of a stack of nine of the blank PCBs that I need to populate (solder all the components on to) next.

Here's a video of the motor being controlled via the populated PCB, a Windows GUI and a USB to RS485 converter. The GUI is written in National Instruments LabVIEW, which I find to be a very good programming language to use.

I've also put four off two pin headers on the PCB so I can address the boards on the multidrop RS485 bus and "talk" to and control them individually.

Next step is to get the new PCB loaded up electrically in the test tank, churn some water up so to speak :)

Saturday, 12 January 2013

Thruster Kort Nozzle

I've just tried out a proper Kort nozzle on my thruster and it seems to work a treat, definitely more consistent thrust above 1kg, I did see over 2.3kg on the fish scale, equating to about 1.15kg from the thruster, but here's a shot of 2.22kg, or just over 1.1kg at the propeller.

Here are a couple of pictures of the new Kort nozzle mounted to the thruster.

It's a lot closer fit around the propeller than the shroud I made earlier, which is probably increasing the thrust.
Next step is to manufacture a mount for the thruster itself, so I can attach it to the frame of the ROV..... it's still only in "Martin's 3D World" at the moment, and unfortunately so is the frame of the ROV for that matter :)