Mag coupling version 3.0 vid (updated new video 1/25/15)

[FJM]

I see below the listing prices as low as $8.50 from China! I might have to spring for a few. I'm OK with knock-offs as well. Unfortunately, I just ordered a set of motors. Might have to wait until I actually build and test something before upgrading...

Funny though, there are three separate vendors all of them have exactly 10 left.

Things look promising, so keep at it. If you run into troubles, post the failures as well. There's a lot of smart people out there that may have solutions that you can build on. I've been seeing quite a few people suggesting the use of magnetic coupling, but you're the only one that I've seen with a (somewhat) operational prototype; everyone else seems to be in the concept stage. It looks like there are some commercial versions, but nothing that looks affordable.

[a_shorething]

Thanks Gadgetguy, appreciate the encouragement.

I see below the listing prices as low as $8.50 from China! I might have to spring for a few. I'm OK with knock-offs as well. Unfortunately, I just ordered a set of motors. Might have to wait until I actually build and test something before upgrading...

Funny though, there are three separate vendors all of them have exactly 10 left.

Things look promising, so keep at it. If you run into troubles, post the failures as well. There's a lot of smart people out there that may have solutions that you can build on. I've been seeing quite a few people suggesting the use of magnetic coupling, but you're the only one that I've seen with a (somewhat) operational prototype; everyone else seems to be in the concept stage. It looks like there are some commercial versions, but nothing that looks affordable.

Thanks, I often say that I don't know enough to know what can't be done, so I'll try whatever I think of.

Just about every mag coupling I've seen has depended on a some kind of longitudinally attached coupling that puts the prop at the rear (which I guess seems conventional) but to me the idea of having the thruster turn around a pipe is less complicated and has more of a chance of working with physics instead of against it. I was thinking with this arrangement you could put a few of them along the same pipe and have multiple motors. You could angle the PVC with a 45 or a 90 or a pair of 45s in the middle and have four thrusters at the corners all using essentially the same section of plumbing, you could do two on the same pipe for vertical thrusters... Lots of options if it works.
And I will post my tests even if they fail, I think I'm close but I agree, if I can't work it out, someone else may be able to, based on my initial info/design.

[scubersteve]

That's why I wanted to try this because I think most of the hobby motors are at least $20 each and the ones that are this large typically draw 10 amps or more. This 1100gph motor is rated at 3 amps to run, 5amp max (requires 5 amp fuse..

While the 1100 may be rated as such, isn't that when spinning the stock bilge pump impeller? That spec is likely to be rather different under the load you are putting it under. I would be interested to see the amperage draw numbers for your loaded tests. For a strong, efficient, cheap brushless look at a dt700 from hobbyking. ($12) I have several and have been very impressed with them. I posted it's draw and thrust numbers on another thread.

I'm following this thread with interest.

[a_shorething]

Thanks Steve and you make a good point,

You may be right on that. And like I said, the design itself doesn't hinge on the bilge pump being the 'right' one, but I had to start somewhere. I actually got a cheap geared motor to test in the rig also. I figured I could go with a more aggressive prop profile at lower RPMs if it were geared. I'll give that a shot too.

I have to check with a friend of mine about testing the amperage draw, my Fluke meter doesn't do amps. Alternately I could take it to my local hackerspace, but that would be kind of embarrassing if I spent an hour setting up the rig only to turn on the power and have it go poof! Or worse, just sit there doing nothing while everyone looks on.

That's why I wanted to try this because I think most of the hobby motors are at least $20 each and the ones that are this large typically draw 10 amps or more. This 1100gph motor is rated at 3 amps to run, 5amp max (requires 5 amp fuse..

While the 1100 may be rated as such, isn't that when spinning the stock bilge pump impeller? That spec is likely to be rather different under the load you are putting it under. I would be interested to see the amperage draw numbers for your loaded tests. For a strong, efficient, cheap brushless look at a dt700 from hobbyking. ($12) I have several and have been very impressed with them. I posted it's draw and thrust numbers on another thread.

I'm following this thread with interest.

UPDATE: I started building the test rig yesterday. I put together something that should help me rapidly change out the thruster props and 'reload' quickly between tests. I might even be able to get it over to the river to test it, rather than using a test tank (which I don't have). I've got a digital scale and will be using something similar to the what the guys at OpenRov came up with. (This isn't mine, it's theirs)

[a_shorething]

OK, so I built a test tank (of sorts) with two goals in mind.
1. See if this thing will actually move some water and
2. See if I can get any kind of information out of it before my meager power setup maxes out*

I think both of them were successful tests which means this thing is beyond the point of wondering 'IF' it will work, now it's a question of 'HOW' do I make it work best.



My meager setup so far is this: Arduino motor shield passing through the power from a 12vdc 1000 ma transformer.

I'm doing it this way so far because I don't have a proper motor controller so I have limited my input voltage. I'm also throttling back the amount I'm actually asking of it using my software interface. It will accept a range of throttle inputs from 0-255 where 0 is no voltage and 255 is 12vdc. I started my tests at 50 and worked my way up to only 77 before I started overheating the Arduino and I blew the 1 amp inline fuse on the 1000ma transformer line I put in to limit my tests so I didn't cook anything.

Here are the results:

For 65 degree 3 blade prop:
At 50/255 input: 2.03 volts DC, 1.15 oz of thrust
at 75/255 input: 2.54 volts DC, 1.9 oz of thrust, about 53 grams

For 45 degree 3 blade prop:
at 50/255 input: 2.0 vdc, 1 oz thrust
at 70/255 input: 2.43 vdc, 1.5 oz thrust

So the results, while not astounding and certainly not worthy of much celebration, were positive.

The mag thruster works as a concept, now I need to get more equipment to be able to add power, and possibly try different motors to find a workable setup.

Also, FWIW, friction against the PVC tube was less than it was in air and didn't seem to be a factor at all.

[FJM]

Congrats on the successful tests!

You have a fair amount of range left in your testing. I'm interested in what the rest of the voltage range will bring. Maybe you should just wire it direct to a car battery and let 'er rip!

Have you tried a "regular" thruster with your test rig? Might be an idea just to make sure it's working as you'd expect and giving comparable numbers.

I'd be interested in the current draw as well as that will indicate battery life. Going slower but lasting longer might suite some applications. Sure seemed quiet as well!

Are there only three blades? I thought it had more, similar to a ducted fan.

Fred

[a_shorething]

Congrats on the successful tests!

You have a fair amount of range left in your testing. I'm interested in what the rest of the voltage range will bring. Maybe you should just wire it direct to a car battery and let 'er rip!

Have you tried a "regular" thruster with your test rig? Might be an idea just to make sure it's working as you'd expect and giving comparable numbers.

I'd be interested in the current draw as well as that will indicate battery life. Going slower but lasting longer might suite some applications. Sure seemed quiet as well!

Are there only three blades? I thought it had more, similar to a ducted fan.

Fred

Fred,

Yeah, 1 amp doesn't really show much, but it did prove out the things I was trying to accomplish and I didn't cook anything this time, so that was a bonus (So far the electronic death toll is at 2 Adafruit motor shields... may they rest in pieces. 1 was version 1.2 and one that was version 2.0).

For the comparison, that's a good idea. This is the only thruster I've seen up close and personal though, so nothing to compare it to. If anyone has a motor/controller/ESC they wouldn't mind loaning me, I'd love to have a go at it. But given my proclivity for letting the magic smoke out of circuit boards I doubt I'll get any takers. .

I had a couple of things kind of 'hacked on' to the test rig so I didn't show the whole thing in the video but once I get the real power supply and motor controller I'll show the whole thing.

I do want to monitor the current draw as well, my Fluke VOM doesn't have an ammeter though, may have to get to Harbor Freight and find a cheap one.

You're right, version 1.0 was more like a ducted fan on the 3" PVC. For version 2.0 it's on 1.5" PVC and only 3 blades for starters. The bilge pump motor is a higher RPM motor than many of the thruster motors I think (just guessing, but it typically spins a very small impeller very fast) and so it doesn't have much bottom end torque. For this reason I wanted to reduce the amount of water it would be pushing with each rotation, hoping to make up for it in higher RPMs with shallower pitch props, at least for the initial tests. If I had a high torque, low speed motor like a geared motor to test I might go with more blades and maybe even longer ones.

I'm just psyched that the design seems to work really well. From here I can set up pretty much any motor inside pretty much any size PVC and make it work. The trick will be fine tuning the relationship between the motor and the thruster prop design (which take about 2 seconds to change out by the way because you can just pull off one and pop on another by hand). The prop variables include pitch, number of blades, size of PVC and outer diameter. I have a template cross section that looks like it will work forward and backward and is basically a wing cross-section.

Also the fact that I was able to build two props with different pitch and outer ring (or attached nozzle) dimensions and they both seem to work pretty well right off the bat is encouraging. With version 1.0 it took a lot just to get it to turn on the dry 3" PVC without stalling. With the new design things just 'fit' together and work, dry, wet, whatever.

I can't wait to get some more stuff to test it with. Gotta wait for the hobby fund to be replenished, then I'm going to get some motor controllers and really see what it can do...


PS- I think I need to just do subtitles from now on. I sound stupid on video.

[a_shorething]

Update: Tried with a low speed, high torque geared brushed motor I found on some auction site.


Just hooked it up to a drill battery while I'm still waiting for the ESC.

14.4 volts DC DeWalt drill battery---

out of water: 2.4 amps draw
In water: 6.10 amps draw at full power.
1.7 oz of thrust with 65 degree prop (3 blade)
2.7 oz of thrust for 45 degree prop (3 blade).


Not great numbers I'll admit, but I'm not sure if the drill battery actually has any more than 6 amps to give and this one is a 200rpm geared motor so I think I still have some room to move if I use more power or a geared motor with higher rpms. It seemed to be moving quite a bit of water anyway.

I think the next try will be with the 45 degree prop but more blades to see if it can produce more thrust at low speed. A couple of times it spun the prop until it let go of the inner hub but then I got it to position properly. I think I may need to put a 'keeper' ring on there on either side of the thruster blades to keep it from walking forward or back. That would be cheaper than adding magnets to the thruster hub although I can do that too if I need to (just using 2 pairs right now but the spots are there for 4 pairs and I can add more spots if I need them.

Stay tuned, I just started printing in ABS and this stuff ROCKS!

The prints are coming our really clean and fitting perfectly. Things are moving along...

[olegodo]

Great news!

Good to hear you are getting good prints with ABS. The strength difference from PLA to ABS is fantastic.
I am still waiting for my parts, it has almost been 4 weeks in the mail now

[a_shorething]

Hey Ole!

Thanks. Yeah, I think the supplier you picked is a good buy, but not so strong on the quick delivery.

I have the same problem with the dome supplier I picked. They do great work but I think they only work in large batches so they wait until they have enough. Lead time on them (from cleardomes.com) is typically 4-5 weeks based on two orders from them.

I hope you get your stuff soon and get a chance to test.

My hobby fund is pretty much depleted until after the holiday so I doubt I'll be able to do any testing with a real ESC to seriously get a good idea of the amount of thrust my design can put out.

Right now I'm printing the first thruster prop design I'll be doing in ABS, not sure if I'll use the acetone trick to smooth the blades yet. It's going to be the same diameter but 6 blades at 45 degrees. Results to follow...