03-28-2025, 09:14 AM
03-28-2025, 08:47 PM
(03-28-2025, 09:14 AM)Lasco Wrote: [ -> ]Very interesting concept,
Hope that it will be fruitful
It was looking good till I started to cut the contacts. I can't cut them precise enough, and the small metal slips are kinda flimsy. So it's going into the failure bin.
But I ain't done yet.. A while back I made this one below, and it was almost working right.. BUT I made the ring a little too tight for comfort (fitting the brush) And I was having difficulty sanding the inside smooth (also too tight).. But I think it would have worked if I got it right..
[attachment=1680]
So I plan on re-doing this one over with massive improvements..
- Make it tall enough to sand the inside smooth with a Three-Jaw Cylinder Sander
- Make it wide enough to easily fit a proper brush (or two)
- Secure the copper rods in with Epoxy Resin walls on top and bottom
I will get a working commutator soon I'm sure
03-29-2025, 04:56 PM
OK coming along.. 90 contacts!
Here's the progress
[attachment=1681]
[attachment=1682]
The contacts are 10 gauge wire. The brush will ride right where that middle support is on the inside. Only thing I wish I did different is I should have used more than 90 contacts, so I could close those gaps more. But it should still work (I predict)
I made it big so a 3 leg bore sander can fit inside to smooth and concave the contacts- like this
[attachment=1683]
The epoxy is curing now on 1 side, and I plan on pouring the other side tonight.. All 90 10ga contacts were straightened and glued / epoxied in place. The contacts will be mighty firm once everything sets.
The brush I choose is a standard vacuum brush. It's just wide enough so it barely touches 3 contacts, so it should meet Figuera's specs of always touching at least 2 contacts.
Here's the progress
[attachment=1681]
[attachment=1682]
The contacts are 10 gauge wire. The brush will ride right where that middle support is on the inside. Only thing I wish I did different is I should have used more than 90 contacts, so I could close those gaps more. But it should still work (I predict)
I made it big so a 3 leg bore sander can fit inside to smooth and concave the contacts- like this
[attachment=1683]
The epoxy is curing now on 1 side, and I plan on pouring the other side tonight.. All 90 10ga contacts were straightened and glued / epoxied in place. The contacts will be mighty firm once everything sets.
The brush I choose is a standard vacuum brush. It's just wide enough so it barely touches 3 contacts, so it should meet Figuera's specs of always touching at least 2 contacts.
03-31-2025, 01:25 AM
10 gauge, damn, you'll have current on tap - incidentally, can you even measure the resistance on those rods?
Great work, looking forward to this one.
Great work, looking forward to this one.
03-31-2025, 09:50 PM
I decided to rebuild with 120 contacts and do some corrections I wanted to.. If I am building this thing, I am going to do it right.. So I added 30 more contacts to reduce the spacing by 1mm between each contact. And I made a much beefier center support.
The rebuild is coming along nicely.. I got over 1/2 the contacts glued in place. (120 contacts will be installed in total)
The spacing between contacts looks excellent. The center support is much beefier, and I made the contact rods more uniform..
[attachment=1684]
[attachment=1685]
The top and bottom of the rod holders will be resin epoxy holding them solid. The rod contacts need to be smoothed and flattened from the inside when all is dried..
The downfall is the PLA filament.. If the contacts get too hot, they could deform the PLA. But it should withstand moderate use, the 10 AWG wire contacts should disperse heat pretty well and I will not be running it 12 hours a day at this point.
I have lets of ideas for this baby.
The rebuild is coming along nicely.. I got over 1/2 the contacts glued in place. (120 contacts will be installed in total)
The spacing between contacts looks excellent. The center support is much beefier, and I made the contact rods more uniform..
[attachment=1684]
[attachment=1685]
The top and bottom of the rod holders will be resin epoxy holding them solid. The rod contacts need to be smoothed and flattened from the inside when all is dried..
The downfall is the PLA filament.. If the contacts get too hot, they could deform the PLA. But it should withstand moderate use, the 10 AWG wire contacts should disperse heat pretty well and I will not be running it 12 hours a day at this point.
I have lets of ideas for this baby.
03-31-2025, 09:57 PM
(03-31-2025, 01:25 AM)unimmortal Wrote: [ -> ]10 gauge, damn, you'll have current on tap - incidentally, can you even measure the resistance on those rods?
Great work, looking forward to this one.
Virtually no resistance, the rods are only the commutator contacts. If I really want to work with rotating fields, I think it's imperative to have High-Resolution switching, meaning Lots and Lots of switches so the field can move in micro-steps instead of big jumps.
My electronic setup has 24 switches and is a nightmare to code and wire. Plus if I make 1 wrong move and fry some switches, it is a major PITA to find the blown components and repair the circuit.. But the raw commutator will solve all that.. With the advantage of no body diodes, and assurance raw copper can flow current in any and all directions.
Eitherway, I am NEVER going to know if rotating fields will work or not unless I give it my all.
04-01-2025, 03:06 AM
[attachment=1690]
Hi Jim,
I had my commutator frame made from PLA, and I noticed that the brushes do warm up the entire system. I measured temperatures exceeding 65°C, which is similar to chewing gum from PLA. To address this issue, I installed a small computer fan (120x120 mm) for cooling. With the fan running, the temperature stays at a maximum of 30°C in a room temperature of 24°C.
Hi Jim,
I had my commutator frame made from PLA, and I noticed that the brushes do warm up the entire system. I measured temperatures exceeding 65°C, which is similar to chewing gum from PLA. To address this issue, I installed a small computer fan (120x120 mm) for cooling. With the fan running, the temperature stays at a maximum of 30°C in a room temperature of 24°C.
04-01-2025, 04:52 AM
Hi Jim,
Sounds like you are making good progress in the right direction! A mechanical commutator addresses a lot of the shortcomings of semiconductors.
Would it not be worthwhile printing it in PETG or ABS instead of PLA at this point? If you're putting a lot of effort into building something with a degree of precision then warm temps are going to deform PLA and possibly throw the device out of alignment.
Another option would be to anneal the PLA print you have, to give it better heat resistance properties. This post describes one way to do that: https://www.reddit.com/r/fosscad/comment...some_mags/
Just a thought.
Best regards,
Lee
Sounds like you are making good progress in the right direction! A mechanical commutator addresses a lot of the shortcomings of semiconductors.
Would it not be worthwhile printing it in PETG or ABS instead of PLA at this point? If you're putting a lot of effort into building something with a degree of precision then warm temps are going to deform PLA and possibly throw the device out of alignment.
Another option would be to anneal the PLA print you have, to give it better heat resistance properties. This post describes one way to do that: https://www.reddit.com/r/fosscad/comment...some_mags/
Just a thought.
Best regards,
Lee
04-01-2025, 07:42 AM
Thanks Guys. Some good idea's to think about. I used to anneal all my prints in grain alcohol until I ran out. It does help to some degree. And my printer can print anything, so ABS remains a possibility.
I am not too worried about the top and bottom supports, as the contacts will be encased in epoxy resin. And the contacts (when finished) will be around 100mm long so the ends should be manageable with regards to heat.
The crucial part is the middle ring where the brushes will ride. It is 30mm high, beefier than the rest, and I used 70% infill. The computer fan idea doable also.
This is mainly a proof of concept to see if this method is plausible. If the center support can't take the heat, I may be able to wrap some thermal rubber on top and under the center frame and hose clamp it in place to restore the rigidity.
If the build works out well and I end up rebuilding, I will have to address this issue.
One of the methods I want to try is use this with 2 brushes like a motor commutator. Then rewire a large universal motor armature with 120 taps, something like every 5 loops per tap. If the armature has 12 slots, I can do something like 10 taps per coil slot. This should produce High-Resolution rotating magnetic field instead of 12 jumps it will have 120 steps per revolution.
Anyway- more soon
I am not too worried about the top and bottom supports, as the contacts will be encased in epoxy resin. And the contacts (when finished) will be around 100mm long so the ends should be manageable with regards to heat.
The crucial part is the middle ring where the brushes will ride. It is 30mm high, beefier than the rest, and I used 70% infill. The computer fan idea doable also.
This is mainly a proof of concept to see if this method is plausible. If the center support can't take the heat, I may be able to wrap some thermal rubber on top and under the center frame and hose clamp it in place to restore the rigidity.
If the build works out well and I end up rebuilding, I will have to address this issue.
One of the methods I want to try is use this with 2 brushes like a motor commutator. Then rewire a large universal motor armature with 120 taps, something like every 5 loops per tap. If the armature has 12 slots, I can do something like 10 taps per coil slot. This should produce High-Resolution rotating magnetic field instead of 12 jumps it will have 120 steps per revolution.
Anyway- more soon
04-03-2025, 06:39 AM
Epoxy is set- ready to continue.
I want the option of running both negative and positive, so I need to be able to accommodate 4 brushes. 2 to ride on the slip rings, and 2 for the contacts.
I can not do machine-shop precision, so I need to use a motor with a long enough shaft where I can fit it all. So this is my choosing
[attachment=1693]
This motor is a brushless induction type with a nice long shaft. It's an over-kill But it should leave me enough room to attach a fan blade also to keep it cool.
I was thinking of using a dual shaft motor where the shaft comes out both ends, but there was no way to connect the slip ring wires to the brushes. I was also considering making flat slip-ring plates on top of the brush assembly, but the 120 contacts come out the top so there was no logical way to design the slip-ring brush holders looping over the contacts. So the guts will all have to be inside the cylinder.
No problem tho, I think I can design a brush / slip-ring assembly all in 1 piece, and the slip-ring brush holders can easily and securely mount on to the existing frame bolts.
I want the option of running both negative and positive, so I need to be able to accommodate 4 brushes. 2 to ride on the slip rings, and 2 for the contacts.
I can not do machine-shop precision, so I need to use a motor with a long enough shaft where I can fit it all. So this is my choosing
[attachment=1693]
This motor is a brushless induction type with a nice long shaft. It's an over-kill But it should leave me enough room to attach a fan blade also to keep it cool.
I was thinking of using a dual shaft motor where the shaft comes out both ends, but there was no way to connect the slip ring wires to the brushes. I was also considering making flat slip-ring plates on top of the brush assembly, but the 120 contacts come out the top so there was no logical way to design the slip-ring brush holders looping over the contacts. So the guts will all have to be inside the cylinder.
No problem tho, I think I can design a brush / slip-ring assembly all in 1 piece, and the slip-ring brush holders can easily and securely mount on to the existing frame bolts.