Bot Cousteau- My ROV project

[Zaibach]

Nice interface
Makes me wish I was better at GUIs, as in anything graphics scares the living bajeebers out of me lol

[a_shorething]

Thanks. I was trying to set it up so that the control meters (progress bars) went from right to left on the 'left stick input' but it turns out you need a different version of VS to do that easily so I just left it.

Basically I wanted an easy way to verify every signal was being sent through and to see what the codes were.

[duncan]

Thanks. I was trying to set it up so that the control meters (progress bars) went from right to left on the 'left stick input'

Haha this is exactly how I decided to set it up when I connected the xbox controllers in my vb.net app. Thats weird

I'll be interested in your project since it seems we're both attacking it from the same angle (i.e with a programmer mindset I guess). Of course you're a fair bit further along so I'll be stealing all your ideas soon

[a_shorething]

Thanks. I was trying to set it up so that the control meters (progress bars) went from right to left on the 'left stick input'

Haha this is exactly how I decided to set it up when I connected the xbox controllers in my vb.net app. Thats weird

I'll be interested in your project since it seems we're both attacking it from the same angle (i.e with a programmer mindset I guess). Of course you're a fair bit further along so I'll be stealing all your ideas soon

Nice. No problem. If you can use it, help yourself.

I haven't coded it yet, but I came up with a pretty cool idea for doing vectored thrust controls. I'll post the code when I get it tested out.

I have a couple of bugs to work out in the application in 'serial' mode then I plan to upload the Ethernet libraries and see what's involved in working with them. With any luck the serial version will be done tonight. I found a hackerspace near me and I'm going there for 'Microcontroller Monday' tonight. I should get some time to work out the kinks and hopefully get some cool ideas for some other stuff.

[a_shorething]

Well, after a bit of a hiatus during which time I was not idle, but nearly so, I have resumed work on my ROV in earnest. (I can't believe my last update in this thread was over 12 months ago!)

I posted a short update in the programming section under the miscellaneous category because the platform/language I've finally decided to go with isn't in the list!

As I mentioned in that post I had considered various combinations of Raspberry Pi, Arduino and possibly even the Beaglebone Black (used by OpenROV) even though every time I looked at specs I saw that the Netduino had just about all I needed in a tight little package. Well a few weeks ago I decided to follow my research instead of my 'feelings' about what I should use.

I am a programmer by trade and C# is my language of choice. As such it's probably easier for me than for most to 'get' the programming and debugging interface of Visual Studio, but I REALLY like the Visual Micro environment for Netduino. This is definitely going to work for me.

I have worked out the UDP protocol that I will use over my Cat-5 ethernet tether and I should be able to add that capability to the XBOX interface shown at the beginning of this thread pretty easily. It's fast and extremely simple, sending things at the byte level so bandwidth shouldn't be an issue at all. Best of all, the protocol is built into the Netduino so it's almost plug and play. As mentioned in the previous post from 11/2013, ethernet comm was a bit of a hurdle for me because the learning curve from 'hello world' stuff in the Arduino to the point where it's talking over ethernet cable is a hardware as well as software challenge. I was using an Uno which would have required an ethernet shield and then you are limited or need to get creative with other shields based on space and pin pass-thru. I know there were probably options that would make it easier, but I don't think any provided the simplicity and price point of the Netduino Plus 2 with all of the built-in options it offers (micro SD, Ethernet, processor speed, onboard memory, all for under $70, and it accepts Arduino shields as well). The Netduino code-debug-repeat process is also very simple and intuitive to me so I think it's going to work really well. I highly recommend it for anyone else who's still on the fence about a controller.

In addition, I've got a few IBT-2 relays so I can finally resume my thruster tests, this time with full power (previously I was limited to some 2 amp relay 'shield' boards I was using with the Arduino, and cooked one of those at a relatively low power setting). The IBT-2 can handle over 30A so I should be good to go on that front.

My 3D printed mag-coupled thrusters will be the featured design element of this first ROV and I think I'm going to go simple in version 1.0 (two side thrusters and one vertical instead of full-vectored thrust in all axes) so I can test it out as a proof of concept. On the bench the IBT-2 relay and motor combo show no signs of heating up at 100% power from my 8 amp battery with no load. Hopefully I can run at 75% or better without cooking anything. (one goal of this project is to provide identical or superior performance to the current state of OpenROV at an equal or lower price point, even for a one-off).

From where I am, I'm hoping to have thruster results done and posted by the end of this month and with any luck, I'll be ready to launch by spring!!!

Standby for updates, hopefully more often and soon.

[olegodo]

Good to hear you are back! And with interesting developments. Keep it up!

[a_shorething]

Update:
I put the new power setup (Netduino over ethernet through IBT-2 H-bridge instead of Arduino with motor shield using USB serial) to the test the other day and the results were not promising. I had tried with some parts I'd already been using and it just didn't really generate that much force (25 grams or so max) at full power. It was only drawing about 1.5amps at that point from my battery too though. I'm not sure what that means but I'm thinking it means that either a) the bilge pump motor I'm using isn't going to cut it or b) the thruster design I was trying just wasn't aggressive enough (not enough surface area/number of blades/angle maybe?). What I thought I'd do is design a set of props with more surface area and see how that affects the numbers before scrapping the current motor plans (which have the dual benefits of being cheap, and being here ). So I set about modding my prop design.

That took about 2 hours and then it was on to fabrication.

I tried to crank up the 3D printer and it looked like all was going well until about the 4th or 5th layer when it started to get a little funky. Checked a few things and determined that the ABS I'd had sitting on the spool probably had absorbed some moisture over the months its been sitting there. (this is a known issue, just not something I'd run into).

After doing some research I went and got a food dehydrator and worked out a way to fit a spool of ABS in it. Now I'm on hold until I can toast my ABS for about 10-12 hours to get all the moisture out and try again.

Anyone else got any theories about the lack of power/draw? Is it possible that I'm just not going to be able to generate enough power with these motors and these are the symptoms? I was thinking this way:
If it was trying too hard and not generating much power then it would clearly show I needed to move to a bigger/more powerful motor. Something like 6 amps draw and very low power numbers.
Since it's only drawing like 1.5 amps and ALSO not generating much force (and doesn't seem to be bogging down or working very hard) I'm going to try to make it work harder and see how that changes things.

Will update once I get the new thruster set up.

PS- Now that I've got plenty of power available and I'm not sweating the fact that I could blow up my whole setup by pushing too hard, I'm really liking the mag-coupler design. It's nice and smooth and looks like it will work just fine. Cost: less than $100US per thruster including the H bridges required for each one. And the motor NEVER GETS WET. It will cost more if I end up going with specifically hobby motors, but will probably still not go over $100 each, I'm not sure.

[olegodo]

Have you checked that the motor and prop isn't slipping in any way? My first thought at least.

As for the propeller you were testing, what is the diameter, blade count and angle of the blades?

When I first started out I planned to use bilge pumps as well (you might remember). I printed a 3 blade propeller, aprox 70mm dia, but don't remember the angle of the blades. It generated a pretty decent amount of force, but I was testing it coupled directly to a 12v 30A power supply and the motor gave out after five minutes.
So the bilge pumps them self should be good for quite a bit of thrust.
Note that because of the generally high rpm nature of these motors, the prop size and angle needs to be kept relatively low. A larger prop with larger surface area require a higher torque/lower rpm motor to be efficient.

Just my thoughts! Looking forward to hear about your next revision.

[a_shorething]

Hi Ole,

Thanks for the response. No slippage. There is a definite linear relationship between speed and power generated, it just doesn't get high enough.

When I opened it up afterward it was still nice and tight.

The diameter of the props tested was about 4.5-5" but the angle was 45 deg. and 60deg. (tested both) with three blade props and the props themselves are less than an inch long. I started out with a conservative profile so I could ramp up to it. I think I'll stick with 45 degrees but go with 4 slightly longer blades and see how that goes.

The motor hasn't been getting hot in the current configuration and hasn't been drawing more than 1.5 amps either, so I feel like there is still something on the table.

Hopefully sometime tomorrow I'll have something to report. I have a few more ideas for what to do, including a bizarre twin-engine setup that might be cool (two motors turning one thruster). Stay tuned...

[a_shorething]

OK, brief update. Probably not enough info on it but it is encouraging.

I'm working on a unique (I think) mag coupler design that uses magnets in a radial arrangement that fit inside the thruster prop rather than having it on the back. This means you can actually put more than one motor or thruster on a single PVC pipe and power them together or separate if you wanted to.

Anyway, I'm still working on the 'small' one in a 1.5 inch PVC pipe and I just got over 100g of thrust for the first time from it. It's extremely light and simple to build (print actually, using a 3D printer) and being sealed from the water means that it will last longer with less wear and tear than one exposed to the water. Ironically enough I am using a bilge pump motor, but in a dry environment.

I had a LONG delay because I couldn't get my 3D printer calibrated for the ABS after a short hiatus. It took me about a month and probably over 100 prints to figure out the right combination but it's back up and running better than ever.

The current test was with 4 20mm blades at 45 deg angle. I think I might try 8 in the same setup and see how that works.

I wasn't able to test the amp draw because it was a rushed test before company gets here today, but based on my previous tests I think I'm still under 2-3 amps of draw so there is still quite a bit left in the 'tank' if I can figure out how to get to it.

I know there is probably a way to optimize a propelor for the motor being used, but I don't know how to do that, so I'll iterate my way to something that works.

I'm pretty sure I already have something that will work for my small ROV, I just want to see if there is a 'sweet spot' for this motor/hub design that I can find without too much difficulty before moving on to another set for testing. As of right now, the entire mag thruster setup would be under $50. If I can squeeze enough power out of it, I think there might even be a market for it.

Standby for (hopefully) more frequent updates).