Mag coupling version 3.0 vid (updated new video 1/25/15)
Re: Mag coupling version 3.0 vid (updated new video 1/25/15)
Looks good :--
A couple of comments if you don't mind (not that it needs any)
Controlling the speed of a DC motor by varying the voltage doesn't give you the best results in terms of heat or torque, It is generally better to do it with PWM. With PWM you can supply the motor with as higher voltage as it can reliable run at and thus gives you max torque even at very slow speeds. You should be able to run the thing slower with PWM than lower voltage and it not stall.
If you start the thing too fast is there a chance the reaction thrust of the impeller will throw it backwards and disengage with your magnets ? could you do with a couple of stuck on rings just to stop the impeller going too far if it try's too ?
Rather than making the center bore smooth make it with numerous semi circular (or any shape really) holes in it. That way you reduce the plastic to plastic contact area and force a small flow of water between the surfaces.
You could make the motor housing out of ally or stainless and then fill it with demin water that way you'll have a good cooling surface and thermal path to the outside.
A couple of comments if you don't mind (not that it needs any)
Controlling the speed of a DC motor by varying the voltage doesn't give you the best results in terms of heat or torque, It is generally better to do it with PWM. With PWM you can supply the motor with as higher voltage as it can reliable run at and thus gives you max torque even at very slow speeds. You should be able to run the thing slower with PWM than lower voltage and it not stall.
If you start the thing too fast is there a chance the reaction thrust of the impeller will throw it backwards and disengage with your magnets ? could you do with a couple of stuck on rings just to stop the impeller going too far if it try's too ?
Rather than making the center bore smooth make it with numerous semi circular (or any shape really) holes in it. That way you reduce the plastic to plastic contact area and force a small flow of water between the surfaces.
You could make the motor housing out of ally or stainless and then fill it with demin water that way you'll have a good cooling surface and thermal path to the outside.
-
- Posts: 289
- Joined: Sep 10th, 2013, 5:26 pm
- Location: New Jersey Shore
Re: Mag coupling version 3.0 vid (updated new video 1/25/15)
Thanks, I've numbered your questions so I can answer easily.perfo wrote:Looks good :--
A couple of comments if you don't mind (not that it needs any)
1.Controlling the speed of a DC motor by varying the voltage doesn't give you the best results in terms of heat or torque, It is generally better to do it with PWM. With PWM you can supply the motor with as higher voltage as it can reliable run at and thus gives you max torque even at very slow speeds. You should be able to run the thing slower with PWM than lower voltage and it not stall.
2.If you start the thing too fast is there a chance the reaction thrust of the impeller will throw it backwards and disengage with your magnets ? could you do with a couple of stuck on rings just to stop the impeller going too far if it try's too ?
3.Rather than making the center bore smooth make it with numerous semi circular (or any shape really) holes in it. That way you reduce the plastic to plastic contact area and force a small flow of water between the surfaces.
4.You could make the motor housing out of ally or stainless and then fill it with demin water that way you'll have a good cooling surface and thermal path to the outside.
1. I am using PWM. It's going through an IBT-2 relay.
2. I may go with the rings or guides to keep it still, but the acceleration will probably be controlled in software. This will address the issue of starting too fast for the outer hub (thrusters) but also the inner one, which is also printed plastic and is press-fit on the motor shaft. If it starts too fast it will start to work that connection loose. It's a delicate balance, but not something that can't be handled. Also the magnets are CRAZY strong so it's not that much of an issue at he prop pitch I think I'm going to need.
3. Very good idea. I have been planning to add small grooves inside the hub to induce a thin layer of water to flow in there (like brake ducting). It's still a work in progress. As it turns out, when the thruster is spinning it already has a very low amount of friction induced by actual contact. I think the majority of resistance by a great margin has been the force of the water itself, which I need to optimize with prop pitch and size, based on rotation speed of my motors.
4. I could, but my design is based on CHEAP and AVAILABLE which is why I'm working primarily in PVC and plastic. I plan to make a kit of plans and printable parts available when I'm done and so I want to work with things that can be readily obtained and assembled pretty much anywhere.
Re: Mag coupling version 3.0 vid (updated new video 1/25/15)
Sorry about the volt thing I thought I'd read in the thread you were altering the voltage to alter the speed...
Cheapppy yes I totally agree and it's one of my main restrictions (design constraints) however a pressure sensor from a washing machine is probably nothing in cost if you hunt about and even new is going to be less than £5. It will only give you a yes or no answer and will switch at a very low pressure but that may be enough to have confidence in your hull integrity. The car inner tube valve or equivalent is again dirt cheap (free possibly). I'm really thinking aloud about my own project and once I've got the camera and ESCs and raspberry pi and servo controller I need a quick way of seeing If I have a integral hull before flooding it. If you didn't want the feed back whilst underwater then maybe just visible LED inside. That way after opening the hull, seal up again and pressurise it slightly and wait five minutes, if the LED is still on then it'll give you confidence.
Hmm I just thought of a problem with my own idea.. washing machine switches and the like need one side to be at atmosphere and the other to be in the working fluid so you'd have to mount the switch on an air tight container filled at atmospheric pressure.. hmmmm more thinking needed.
Out of interest what 3d printer are you using ?
Cheapppy yes I totally agree and it's one of my main restrictions (design constraints) however a pressure sensor from a washing machine is probably nothing in cost if you hunt about and even new is going to be less than £5. It will only give you a yes or no answer and will switch at a very low pressure but that may be enough to have confidence in your hull integrity. The car inner tube valve or equivalent is again dirt cheap (free possibly). I'm really thinking aloud about my own project and once I've got the camera and ESCs and raspberry pi and servo controller I need a quick way of seeing If I have a integral hull before flooding it. If you didn't want the feed back whilst underwater then maybe just visible LED inside. That way after opening the hull, seal up again and pressurise it slightly and wait five minutes, if the LED is still on then it'll give you confidence.
Hmm I just thought of a problem with my own idea.. washing machine switches and the like need one side to be at atmosphere and the other to be in the working fluid so you'd have to mount the switch on an air tight container filled at atmospheric pressure.. hmmmm more thinking needed.
Out of interest what 3d printer are you using ?
-
- Posts: 289
- Joined: Sep 10th, 2013, 5:26 pm
- Location: New Jersey Shore
Re: Mag coupling version 3.0 vid (updated new video 1/25/15)
I was monitoring voltage and amps and probably didn't make it clear so it probably looked like that, but it's using PWM for control now.perfo wrote:Sorry about the volt thing I thought I'd read in the thread you were altering the voltage to alter the speed...
Yeah, for the hull integrity check I do plan on coming up with a jig that will test it before dropping it all overboard (positive pressure as you mentioned) and then I plan to drop it to the bottom on a line in the ocean with no electronics in it to make sure it stays dry inside. Aside from that I'll probably use some kind of moisture detector and will be able to tilt my camera down far enough to see if I'm taking on water (maybe too late at that point, but whatever)perfo wrote: Cheapppy yes I totally agree and it's one of my main restrictions (design constraints) however a pressure sensor from a washing machine is probably nothing in cost if you hunt about and even new is going to be less than £5. It will only give you a yes or no answer and will switch at a very low pressure but that may be enough to have confidence in your hull integrity. The car inner tube valve or equivalent is again dirt cheap (free possibly). I'm really thinking aloud about my own project and once I've got the camera and ESCs and raspberry pi and servo controller I need a quick way of seeing If I have a integral hull before flooding it. If you didn't want the feed back whilst underwater then maybe just visible LED inside. That way after opening the hull, seal up again and pressurise it slightly and wait five minutes, if the LED is still on then it'll give you confidence.
I'm using a RepRap Prusa Mendel I built a few years ago. Still kind of moody, but it works OK.perfo wrote: Hmm I just thought of a problem with my own idea.. washing machine switches and the like need one side to be at atmosphere and the other to be in the working fluid so you'd have to mount the switch on an air tight container filled at atmospheric pressure.. hmmmm more thinking needed.
Out of interest what 3d printer are you using ?
Re: Mag coupling version 3.0 vid (updated new video 1/25/15)
Ok thanks..
I'm thinking of the makibox as the bits are on sale at the moment .... I think I need one though..
I'm thinking of the makibox as the bits are on sale at the moment .... I think I need one though..
-
- Posts: 289
- Joined: Sep 10th, 2013, 5:26 pm
- Location: New Jersey Shore
Re: Mag coupling version 3.0 vid (updated new video 1/25/15)
It makes prototyping and iterations SOOOO much easier. I wouldn't be trying the ROV without it.perfo wrote:Ok thanks..
I'm thinking of the makibox as the bits are on sale at the moment .... I think I need one though..
Re: Mag coupling version 3.0 vid (updated new video 1/25/15)
I've taken the plunge and ordered the bits for my first 3d printer.. I'll use it to print bits to make a bigger one.. Maybe
-
- Posts: 289
- Joined: Sep 10th, 2013, 5:26 pm
- Location: New Jersey Shore
Re: Mag coupling version 3.0 vid (updated new video 1/25/15)
Cool, good luck.perfo wrote:I've taken the plunge and ordered the bits for my first 3d printer.. I'll use it to print bits to make a bigger one.. Maybe
I wrote a blog when I was first setting mine up. It might help you get through some of the learning curve.
Here's the link in case you're interested...
http://itsa3dworld.blogspot.com/
Re: Mag coupling version 3.0 vid (updated new video 1/25/15)
Ok thanks.. I've had a look at your blog and some interesting stuff there.
I've got access to Autodesk inventor so hope to draw my solids up in that. I haven't quite worked out yet how I get from the Inventor part to 3d printer G codes yet. There is very little details on the net where I ordered my box of bits. It should end up a Makibox so I'm guessing the details will be in the box with it.. I should get it soon and start building...
I've got access to Autodesk inventor so hope to draw my solids up in that. I haven't quite worked out yet how I get from the Inventor part to 3d printer G codes yet. There is very little details on the net where I ordered my box of bits. It should end up a Makibox so I'm guessing the details will be in the box with it.. I should get it soon and start building...
-
- Posts: 289
- Joined: Sep 10th, 2013, 5:26 pm
- Location: New Jersey Shore
Re: Mag coupling version 3.0 vid (updated new video 1/25/15)
Yeah, you'll need to use a Slic3r program or maybe Skeinforge to translate the native drawings (AutoCad exports to .stl for stereolithography, the slicing programs may accept others now I don't know). The slicer program needs to 'slice' it up based on the parameters you enter and have set up for your printer. Lots of variables like speed, layer height, material, diameter, temperature etc.perfo wrote:Ok thanks.. I've had a look at your blog and some interesting stuff there.
I've got access to Autodesk inventor so hope to draw my solids up in that. I haven't quite worked out yet how I get from the Inventor part to 3d printer G codes yet. There is very little details on the net where I ordered my box of bits. It should end up a Makibox so I'm guessing the details will be in the box with it.. I should get it soon and start building...
It's a slippery slope...