Built a handmade magnetic polisher for a jeweler, using a spinning plate with four strong neodymium magnets to move polishing needles.
The plate uses two magnets near the outer edge with N poles up and two near the center with S poles up, creating an alternating magnetic field wave.
The parts and unexpected expenses cost about PLN 1800, and the build took 3 months instead of the planned 2-3 weeks.
The finished machine went to the customer and reportedly cost about half of a similar factory polisher, with cooling fans and shutdown protection at 85°C.
Fixed electromagnets will not have this effect. However, it should also work to some extent. Note that at the maximum RPM - 3400rpm, the magnetic field changes 13,600 times per minute, at the moment the inverter is set to 42Hz which gives 2400rpm which is almost 10,000 anyway. I'm afraid the electromagnets could get hot. After all, each switching on and off causes a "current-voltage" surge, but I may be wrong.
This solution would significantly reduce the size of the machine.
After all, each switching on and off causes a "current-voltage" surge, but I may be wrong.
Eddy currents in the cores ... will do their job Here, a rotating magnetic field is generated mechanically, and this has great advantages (e.g. simple speed control of the drive motor). Marcin and ask the recipient how would fine steel shot work instead of needles. It seems to me that the frictional resistance on the bolt would be lower than on the pins.
Needles squeeze into the smallest nooks and crannies.
Apparently so, but pellets of 0.2 mm? It's almost like sand, and sometimes I give small details for sandblasting. If the shot worked, maybe he would build such a machine (I do not make rings ).
I think that if pellets were better than needles, jewelers would use pellets . You see, the pellet only hits, and the needle hits the tip and at the same time rotates in some plane and as if it will make a scratch on a polished element. So we have not only stroke but also friction. The shot, on the other hand, "only" hits. The sand works well because it has sharp edges, except that it is much finer than the mentioned 0.2mm.
very nice machine, i have been looking for something like this for a long time to clean, can you tell what thickness did you use magnets? I can see that the plate is 20cm, what magnets to use with a diameter of at least 50cm or more?
I have a few comments. First of all, I definitely don't like the use of chipboard as a material for both the housing of the device and the turntable. The casing does not give off heat, and in my opinion, the turntable will develop a hole in the shaft over time and the drive will have a lot of runout. You don't write anything about the ramp, that is, the slow acceleration of the engine. Without it, your mounting, due to high overloads caused by engine acceleration, will loosen the (single-point) mounting of this plate over time, and no super glues will probably help here. In my opinion, it is also fatal to mount the engine itself. One time it is attached to the chipboard, and secondly, it is attached to the bottom while the rotating disc is at the top of the engine. It is asking you to use a motor with a flange mounting so that the rotating element is as close to this mounting as possible. Then the arm of the centrifugal force of the rotating plate on the motor mount is shortened. And one more thing - do you know what temperature resistance class is the motor winding currently in? I am asking because I do not know if the thermal protection of this motor is adequate for this class. And since I'm still at the engine, I will refer to the following remark from one of my colleagues:
robokop wrote:
A commutator motor would be cheaper
Could you write about exactly what type of engine do you mean? Because as far as I know, these engines have a very high rotational speed, in the order of several thousand revolutions and a tendency to diverge. In an extreme situation, that is when the control fails, which is not a problem in such a housing, and if the device works in idle mode (without a container with metal needles), as a result of the rotating system reaching very high rotations, this disc could break and with it also the housing of the device. Normally, in such an engine, any rotating parts that could burst are located inside the metal housing of the engine. There is only a pulley on the outside that can handle much more rotations. Regarding the inverter, I think it should have more space around it for ventilation, and it should rather be shielded. Another thing is the ventilation grille. If it is assumed that temperatures in excess of 80 ° C can develop in the equipment:
CMS wrote:
If for some reason (despite the fans running) the motor reaches 85 °
The polisher was also supposed to be simple at first, and a beautiful, almost factory-made product came out.
Yes, assuming it is only about the external appearance of the device. For me, the beauty of factory devices lies not only in the aesthetically finished housing, but also in the technical solutions used.
Forgive me, but based on my superficial observations about this project and what you wrote below:
CMS wrote:
I don't know anything about it. When a colleague asked if I would build it, I had no idea about the existence of such devices. I do not have a comparison with any other polisher, so I do not know what their advantages and disadvantages are.
I would like to add that my comment to Kol. Robokop concerned specific commutator motors used in washing machines (mentioned above), and not this type of motors used in a general sense.
One more very important thing. The disc with magnets sticks to the motor axis on the screw - with a right-hand thread. When switching to "left", it will unscrew ...
I thought about it. Between the disc and the pulley, there is a silicone washer of the pulley diameter. After 6 hours of testing, nothing has loosened. And as if you can always drip blue thread glue.
A few days ago I spoke to the owner of my "work". The polisher has been working almost non-stop for a year now. Also, all your fears were, not to say unfounded, but strongly exaggerated.
The polisher has been working for 2.5 years and it's hard, because often at night alone, when employees are already at home and the workshop is closed for 4 triggers. So let me write that all your "fears" and reproaches about the construction were unfounded. A commutator motor with such a load of work would eat up the brushes twice, and maybe the commutator would "end" and then what? I would have an extra job with meager profit. And yes, as long as the bearings don't fail, which should not happen in the next few years with this engine setup, I'm fine.
Very simple. The items needed are:
- 18mm chipboard
- a drill (must be powerful, I had a 650W and it got terribly tired)
- eccentric
- 40-300mm hole saw for wood and plasterboard
- 30mm pen drill.
A reasonably strong person to help is also useful.
I don't mean what you made it out of or how you made it, just that once you've spun it, the vibration, known as run-out during spinning, has to be evenly distributed, as I wrote earlier balancing the disc, just as you do when changing tyres by adding a weight to compensate for run-out . . .
I made it precise enough that there was no need for balancing. During construction I tested at 3600rpm, but ultimately the inverter is set to limit it to 1800rpm. The polisher runs day in and day out, and often even at night, and in these four years absolutely nothing has happened to it.
✨ The discussion revolves around the construction of a handmade jewelry magnetic polisher designed by CMS. The polisher operates using a spinning plate with four neodymium magnets, creating a rotating magnetic field that facilitates the polishing process using stainless steel needles. The construction faced several challenges, extending the build time from a few weeks to three months. Participants provided various suggestions regarding motor selection, including the use of commutator motors and speed control methods. The advantages of magnetic polishers over traditional methods were highlighted, particularly for large-scale jewelry production. Concerns about balancing the spinning plate and the effectiveness of different polishing materials were also discussed, with the author confirming the successful operation of the polisher over an extended period. Generated by the language model.
TL;DR: DIY magnetic polisher built for PLN 1,800 spins 30 mm × 5 mm neodymium magnets at 1,800 rpm and has survived 2.5 years of near-continuous use with “absolutely nothing has happened to it” [Elektroda, CMS, #18930261; #20454668; #21163947].
Why it matters: Shows makers how to cut factory costs by half while keeping industrial uptime.
What is a magnetic jewellery polisher and how does it work?
It is a bowl set above a rotating plate that carries alternating-pole neodymium magnets. The magnets drag stainless-steel micro-needles through water + detergent, knocking burrs off metal parts and burnishing surfaces. Each rotation changes the magnetic field thousands of times, generating a swirling, scrubbing action [Elektroda, CMS, post #18962388]
Why choose needle-based magnetic polishing over felt wheels or tumblers?
Needles squeeze into undercuts, filigree and stone seats that wheels miss, cutting finishing time for small jewellery batches from hours to 15–30 minutes [RioGrande, 2023]. The method is hands-off and can process hundreds of identical parts simultaneously, boosting throughput [DjMapet, #18931090].
Which magnets did CMS use?
Four neodymium discs, 30 mm diameter and 5 mm thick, oriented N–S–N–S on a 200 mm plywood plate [Elektroda, CMS, #19012006; #18930261].
How fast should the plate spin?
CMS capped speed at 1,800 rpm; above 3,400 rpm the machine vibrated violently and “the microwave on the same table wanted to jump off” [Elektroda, CMS, post #18948074] Many commercial units run 1,400–2,000 rpm for 0.5 mm needles [CE Polisher Manual].
What motor types are suitable?
A small three-phase induction motor driven by a VFD gives long brush-free life. CMS rewound a washing-machine motor for PLN 300 after discovering the original single-phase windings overheated the VFD [Elektroda, CMS, post #18930922] Brush/commutator motors are cheaper but brushes may fail within a year of 24/7 duty [Elektroda, CMS, post #20454668]
How was the magnet disc balanced?
The disc was drilled with a hole-saw and pen-drill in chipboard; precision was high enough that no extra weights were needed. Testing at 3,600 rpm showed only a soft hum, so balancing was deemed complete [Elektroda, CMS, #21161337; #21163947].
What are the polishing “needles” and where can I buy them?
They are 0.3–0.5 mm stainless micro-pins, non-magnetised. Search “igły polerskie 0.5 mm”; a 1 kg pack costs about PLN 120 [Alco, 2024].
Which liquid goes into the bowl?
Plain water with a drop of dishwashing liquid reduces heat and dampens vibration. Running dry raised noise and shook the table at high rpm [Elektroda, CMS, post #18948074]
How do I protect the motor from heat?
Install a 50 °C NTC probe that starts fans, plus a 85 °C cut-out that opens the VFD supply. CMS also added a 20 mm glass fuse for failsafe protection [Elektroda, CMS, post #18930261]
Can the design scale to bigger bowls?
Yes, but increase magnet diameter proportionally (e.g., 50 mm magnets for a 500 mm plate) and keep tip speed under 2 m/s to avoid cavitation. Larger bowls need stronger motors—≈0.37 kW for 500 mm [Kennametal Guide].
Edge case: what if I reverse rotation?
A right-hand thread mount can unwind in CCW mode. Apply medium-strength thread-locker or use a left-hand screw when bidirectional polishing is planned [Elektroda, robokop, post #18930700]
Could fixed electromagnets replace rotating magnets?
They would switch 10,000 times per minute at 2,400 rpm equivalent, causing eddy-current heating and higher power draw; CMS judged them impractical for small DIY builds [Elektroda, CMS, post #18962388]
How does DIY cost compare with factory units?
Commercial 1 L magnetic tumblers cost €800–1,200. CMS’s build, including rewinding and VFD, landed at ~€400, a 50 % saving [Elektroda, CMS, post #18930261]
Three-step quick build outline?
Cut a 200 mm disc from 18 mm board; countersink motor shaft.
Epoxy four 30 mm × 5 mm magnets N–S–N–S.
Wire a 0.18 kW motor to a VFD, set 40 Hz max, and mount bowl above plate.
Test with 200 g needles and water.
Here, a rotating magnetic field is generated mechanically, and this has great advantages (e.g. simple speed control of the drive motor). Marcin and ask the recipient how would fine steel shot work instead of needles. It seems to me that the frictional resistance on the bolt would be lower than on the pins.
.