Dear YE, I have just replicated your baseline design. It worked just fine. Since there is a small tilt or imbalance in the rotor I am unable to take it to the max RPM. I will come back again once I get that fixed. Thank you so much for the video. I request you to continue making more videos Thanks,
These videos are really interesting. I think the stacking idea is great, making the best use of relatively expensive magnetic flux. Doubling the magnet thickness adds more magnet mass further away from the stators, but stacking lets you put the additional magnet mass right next to the coils. Along that line of thought, I wondered if Disk 2 at 3:34 could be made just one magnet thick instead of two, without much loss. Anyway, really appreciate your videos. I'd be fascinated to see how you set up your simulations.
Dear tobuslieven, Thanks. Your thought looks intuitively valid. Another way of putting it is that deeper magnets allow more flux leakage between them. That is how the flux patterns look. Just one magnet in the centre would reduce power capacity by about 10%. The simulations are done by simple spreadsheet according to the steps listed in the appendix. Regards YE
I love your series! I've watched each video multiple times and learn something new each time. I was wondering what power handling capabilities you would expect if you were to replace the ferrous disk 2 in the 5 mm stack assembly with a third stator? This would keep the number of magnets the same but only increase the thickness of the assembly by about the thickness of the third stator.
Dear Joel Robertson Thanks. Net output voltage would remain much the same or less due to flux decrease, with output resistance x 3/2. FOM (V^2 /R) thus indicates power handling capacity decrease by more than a third. Best regards YE
@@YesEinstein Thanks for your reply. That's very helpful. So, if I wanted to add additional stators, I would want to add additional 10 mm rotors then, too, right?
Dear John Rupert. Thanks. Only sensible in the two outer rotors. Main e.g. 10x5 section magnets can be back to back in the internals with thin or no backing disks.Two up to 5x5 section magnets might be inserted at outer ends of the mains to some advantage as shown in Part4. Regards YE
Excellent work! A hidden treasure for sure. I'm curios, with increasing rpm so does the voltage. But what's the max rpm a generator will increase in power generation, not counting mechanical limitations? What's the mechanism that puts a cap on the power? Does the inductance cause the effective internal impedance to rise (due to frq increase) effectively stopping when the internal generator impedance matches the load (as per maximum power transfer law)?
Dear Luka S Power throughput capability of the coreless rare-earth machine is limited by the tolerable dissipation in its windings. The dissipation comprises separate load and eddy current effects. Both can be computed from the volume of copper comprising the windings and the rotation speed of the magnets. (Both losses are essentially independent of the wire diameter and number of turns filling the volume, and the consequent output voltage) Power capability is thus seen to increase rapidly with machine rpm until limited by the more rapidly increasing eddy loss. The transition can be delayed by adopting optimised Litz-type windings of finer wire. Efficiencies over 90% may be expected. Coreless inductance is very low and effects may be neglected. Maximum power transfer principle is inappropriate implying no more than 50% efficiency and unacceptable winding dissipation. Thanks for your pertinent questions. What seems needed now is comprehensive testing to isolate and confirm or otherwise the predictions. Regards YE
I would like to start by saying this is some very brilliant work you are doing . Seems similar to a halbach array if I am not mistaken? I would like to ask if the strong outside ring could be composed of ferromagnetic material without contributing to flux leakage? Also when you show isolated plates with deep inserts could they be connected to a central part of the disk but isolated for the length of the magnet and still achieve maximum flux density without creating leakage? Also do you plan to do a video on the construction or finished magnet rotors with deep inserts ? Would love to see them if you do.Any reply would be greatly appreciated! Thanks keep up the amazing work!
Dear Andy L . Thanks for your kind comment. Yes, similar to halbach. Permanently magnetised rare earth material has magnetic impedance similar to that of air in all axes. The idea is that the isolated iron pads have relatively very low impedance to much of the "bent" off- main-axes flux for greater efficiency. The pads must be totally isolated within the magnet footprints. Ferrous containing ring would not do. No immediate new alternator construction plans within the bike project. There could be something on the intelligent power controller.
Dear YE Thank you for the reply and the helpful information ! I have watched all of your videos on this project several times , but there are still a few things I am not sure about. If you have the time i would be most grateful for your help! i am currently building an axial flux machine of my own for the purpose of generating power from a waterwheel. It is a large generator at 20 inches in diameter. (i apologize for the units of measure). i have 24 N52 magnets on each rotor . the magnets are 1.5 long by .75 wide by .5 thick in inches . or 38.1 mm long by 19.05 mm wide by 12.7 mm thick . my question is how were you able to determine the optimum rotor gap for your project ? the other question is do you define the pole pitch by measuring from the center of one magnet to the next around the inside circumference? and lastly i am concerned that the gap between my magnets is too wide . the space between each of my magnets is almost exactly twice the width of the face of one magnet . will this large gap between magnets have a negative impact on the winding of my phases ? I was hoping to use the overlapping method you show in your project. if i make each phase one third of a pole pitch wide the magnets would only interact with two phases at any one time at most. your magnets appear to induce on all three phases on the inside of the radius . would it still produce a true thee phase output with 120 degree offset ? Thanks again for the inspiring videos on your project!
Dear @@andyl7935 I find your waterwheel setup very interesting and full of potential for axial flux. I will try to answer your points ASAP. A little preoccupied at the moment. Sincerely YE
Dear Andy L. Reference your queries. Optimum gap between rotors is explained in the Part 4A video; around 8mm for my set up, a wet finger 10mm for your machine. The windings should be sized to occupy the gap with minimum clearance. Large diameter is desirable; power capability, increasing with its cube, potentially avoids step up gearing. 48 of your magnets with overlap windings should be near optimum. Just 24 should work for less than half the power capability and a high ripple voltage. With 24, windings may still be overlapped with each phase up to one magnet wide. (inside circumference / (24*3)). Pole pitch is determined by inside radius and number of magnets. With the large spacing, each magnet will essentially interact with just one phase at a time. Yes, 3 phase 120 degree output but with distorted waveforms; you might say pulses.Happy New Year YE.
Dear Andy L. Just done sums on your 24 magnet setup. With 11 crossings of 5x2mm carefully stranded wire per magnet per phase; star rectified at 60 rpm could be 12 V with just 10% ripple. 250 useful Watt might be taken for100 W dissipation in the windings. Going to 17x1mm might reduce dissipation to around 60 W. 1kW might be taken at 200 rpm using 75x0.5 mm stranding. Errors excepted. Hope this helps your deliberations. Regards YE
Jaw dropping case studying, many thanks for sharing it. One observation, the deep inserts arrangement is more eficinte due to the increased number of magnets, therefore, rather than going through the hassle of making the halbach array configuration, why not, simply buy stronger magnets and stick with the baseline arrangement?
Dear CLEYTON CABRAL. Thanks. That is probably sensible. Bigger is better. The inserts show however effective enhanced flux in the space available. It is also shown that complex magnet arrangements for Halbach can simply be improved on. Regards YE
Wow, very interesting.. So just by adding smaller magnet in between the main 'sector' one can increased a lot the power output. Great. And it's all the same force n52 or n42 type ?
Dear JFAirplane, Reference my example plain sector plus inserts case. The table indicates 1.4/1.2 that is a 17% improvement in power handling for the same stator dissipation. Based on my Figure of Merit, all N52 rather than all N42 should reduce the baseline dissipation by around 26%. Deep inserts might nearly half it. Regards YE
@@YesEinstein Just to confirm.. if motor A is all N52 : it will have about 26% less heat to dissipate . And with deep insert, all n52, overall it will be 13% less than a regular one without insert.. ? if i do understand. thanks again
JFAirplane, Yes, by 0.73 ( -26%), ratio of squares of N40 and N52 maximum flux densities. It would apply to all options in the table. Deep inserts with N52 should give a net dissipation reduction from N40 plain sector magnets of 0.73*(1.2/1.6)^2 ie -59%. (Derivations are from the tabulated figures and the preceding FOM expression.
Hi. What would be a better option? 30x20x5mm magnets or 30x10x5mm magnets with 30x5x5mm inserts? Thank you. I tried to simulate it but with an unclear result for me.
Dear VclavZeleny. 30 x15 x 5 may be more appropriate to compare, still giving 24 around with the same material. Effect of the insert version is mostly to improve effective pole width at the outside diameter. Regards YE
3:00 these charts are a bit misleading, they should start at 0 in x & y axis. And do other rotors have more magnets than the baseline? If so, they are not comparable. It's like comparing a 1.0 L engine to a 1.6 L, and conclude that 1.6 L has more power...
Dear My Opinion. How modest! No data is hidden outside the plotted regions; unlike some popular indices. Comparison comment is fair, as far as it goes. Yours YE
aah wish could know this earlier before making a lot of prototypes lol. thank you for sharing.
Dear YE,
I have just replicated your baseline design. It worked just fine. Since there is a small tilt or imbalance in the rotor I am unable to take it to the max RPM.
I will come back again once I get that fixed.
Thank you so much for the video.
I request you to continue making more videos
Thanks,
Thank you, very useful!
These videos are really interesting. I think the stacking idea is great, making the best use of relatively expensive magnetic flux. Doubling the magnet thickness adds more magnet mass further away from the stators, but stacking lets you put the additional magnet mass right next to the coils. Along that line of thought, I wondered if Disk 2 at 3:34 could be made just one magnet thick instead of two, without much loss.
Anyway, really appreciate your videos. I'd be fascinated to see how you set up your simulations.
Dear tobuslieven, Thanks. Your thought looks intuitively valid. Another way of putting it is that deeper magnets allow more flux leakage between them. That is how the flux patterns look. Just one magnet in the centre would reduce power capacity by about 10%. The simulations are done by simple spreadsheet according to the steps listed in the appendix. Regards YE
Very nice
I love your series! I've watched each video multiple times and learn something new each time.
I was wondering what power handling capabilities you would expect if you were to replace the ferrous disk 2 in the 5 mm stack assembly with a third stator? This would keep the number of magnets the same but only increase the thickness of the assembly by about the thickness of the third stator.
Dear Joel Robertson Thanks. Net output voltage would remain much the same or less due to flux decrease, with output resistance x 3/2. FOM (V^2 /R) thus indicates power handling capacity decrease by more than a third. Best regards YE
@@YesEinstein Thanks for your reply. That's very helpful. So, if I wanted to add additional stators, I would want to add additional 10 mm rotors then, too, right?
@@joel.robertson . Yes, like Disk 2
@@YesEinstein Thank you!
A very good video! Thank you so much for this.
One question.
How can a stacked 5mm configuration have deep inserts?
Dear John Rupert. Thanks. Only sensible in the two outer rotors. Main e.g. 10x5 section magnets can be back to back in the internals with thin or no backing disks.Two up to 5x5 section magnets might be inserted at outer ends of the mains to some advantage as shown in Part4. Regards YE
instaBlaster
Excellent work! A hidden treasure for sure. I'm curios, with increasing rpm so does the voltage. But what's the max rpm a generator will increase in power generation, not counting mechanical limitations? What's the mechanism that puts a cap on the power? Does the inductance cause the effective internal impedance to rise (due to frq increase) effectively stopping when the internal generator impedance matches the load (as per maximum power transfer law)?
Dear Luka S
Power throughput capability of the coreless rare-earth machine is limited by the tolerable dissipation in its windings. The dissipation comprises separate load and eddy current effects. Both can be computed from the volume of copper comprising the windings and the rotation speed of the magnets. (Both losses are essentially independent of the wire diameter and number of turns filling the volume, and the consequent output voltage)
Power capability is thus seen to increase rapidly with machine rpm until limited by the more rapidly increasing eddy loss. The transition can be delayed by adopting optimised Litz-type windings of finer wire. Efficiencies over 90% may be expected.
Coreless inductance is very low and effects may be neglected. Maximum power transfer principle is inappropriate implying no more than 50% efficiency and unacceptable winding dissipation.
Thanks for your pertinent questions. What seems needed now is comprehensive testing to isolate and confirm or otherwise the predictions. Regards YE
I would like to start by saying this is some very brilliant work you are doing . Seems similar to a halbach array if I am not mistaken? I would like to ask if the strong outside ring could be composed of ferromagnetic material without contributing to flux leakage? Also when you show isolated plates with deep inserts could they be connected to a central part of the disk but isolated for the length of the magnet and still achieve maximum flux density without creating leakage? Also do you plan to do a video on the construction or finished magnet rotors with deep inserts ? Would love to see them if you do.Any reply would be greatly appreciated! Thanks keep up the amazing work!
Dear Andy L . Thanks for your kind comment. Yes, similar to halbach. Permanently magnetised rare earth material has magnetic impedance similar to that of air in all axes. The idea is that the isolated iron pads have relatively very low impedance to much of the "bent" off- main-axes flux for greater efficiency. The pads must be totally isolated within the magnet footprints. Ferrous containing ring would not do. No immediate new alternator construction plans within the bike project. There could be something on the intelligent power controller.
Dear YE Thank you for the reply and the helpful information ! I have watched all of your videos on this project several times , but there are still a few things I am not sure about. If you have the time i would be most grateful for your help! i am currently building an axial flux machine of my own for the purpose of generating power from a waterwheel. It is a large generator at 20 inches in diameter. (i apologize for the units of measure). i have 24 N52 magnets on each rotor . the magnets are 1.5 long by .75 wide by .5 thick in inches . or 38.1 mm long by 19.05 mm wide by 12.7 mm thick . my question is how were you able to determine the optimum rotor gap for your project ? the other question is do you define the pole pitch by measuring from the center of one magnet to the next around the inside circumference? and lastly i am concerned that the gap between my magnets is too wide . the space between each of my magnets is almost exactly twice the width of the face of one magnet . will this large gap between magnets have a negative impact on the winding of my phases ? I was hoping to use the overlapping method you show in your project. if i make each phase one third of a pole pitch wide the magnets would only interact with two phases at any one time at most. your magnets appear to induce on all three phases on the inside of the radius . would it still produce a true thee phase output with 120 degree offset ? Thanks again for the inspiring videos on your project!
Dear @@andyl7935 I find your waterwheel setup very interesting and full of potential for axial flux. I will try to answer your points ASAP. A little preoccupied at the moment. Sincerely YE
Dear Andy L. Reference your queries. Optimum gap between rotors is explained in the Part 4A video; around 8mm for my set up, a wet finger 10mm for your machine.
The windings should be sized to occupy the gap with minimum clearance. Large diameter is desirable; power capability, increasing with its cube, potentially avoids step up gearing. 48 of your magnets with overlap windings should be near optimum. Just 24 should work for less than half the power capability and a high ripple voltage. With 24, windings may still be overlapped with each phase up to one magnet wide. (inside circumference / (24*3)). Pole pitch is determined by inside radius and number of magnets. With
the large spacing, each magnet will essentially interact with just one phase at a time. Yes, 3 phase 120 degree output but with distorted waveforms; you might say pulses.Happy New Year YE.
Dear Andy L. Just done sums on your 24 magnet setup. With 11 crossings of 5x2mm carefully stranded wire per magnet per phase; star rectified at 60 rpm could be 12 V with just 10% ripple. 250 useful Watt might be taken for100 W dissipation in the windings. Going to 17x1mm might reduce dissipation to around 60 W.
1kW might be taken at 200 rpm using 75x0.5 mm stranding. Errors excepted. Hope this helps your
deliberations. Regards YE
Jaw dropping case studying, many thanks for sharing it.
One observation, the deep inserts arrangement is more eficinte due to the increased number of magnets, therefore, rather than going through the hassle of making the halbach array configuration, why not, simply buy stronger magnets and stick with the baseline arrangement?
Dear CLEYTON CABRAL. Thanks. That is probably sensible. Bigger is better. The inserts show however effective enhanced flux in the space available. It is also shown that complex magnet arrangements for Halbach can simply be improved on. Regards YE
Wow, very interesting.. So just by adding smaller magnet in between the main 'sector' one can increased a lot the power output. Great. And it's all the same force n52 or n42 type ?
Dear JFAirplane, Reference my example plain sector plus inserts case. The table indicates 1.4/1.2 that is a 17% improvement in power handling for the same stator dissipation. Based on my Figure of Merit, all N52 rather than all N42 should reduce the baseline dissipation by around 26%. Deep inserts might nearly half it. Regards YE
@@YesEinstein Just to confirm.. if motor A is all N52 : it will have about 26% less heat to dissipate . And with deep insert, all n52, overall it will be 13% less than a regular one without insert.. ? if i do understand. thanks again
JFAirplane, Yes, by 0.73 ( -26%), ratio of squares of N40 and N52 maximum flux densities. It would apply to all options in the table.
Deep inserts with N52 should give a net dissipation reduction from N40 plain sector magnets of 0.73*(1.2/1.6)^2 ie -59%.
(Derivations are from the tabulated figures and the preceding FOM expression.
Would a ring of horseshoe magnets 🧲, with thestator in-between the north and south poles, work in this setup? 🤔
Yes; same principle
but not so practical or potentially efficient.
Hi. What would be a better option? 30x20x5mm magnets or 30x10x5mm magnets with 30x5x5mm inserts? Thank you. I tried to simulate it but with an unclear result for me.
Dear VclavZeleny. 30 x15 x 5 may be more appropriate to compare, still giving 24 around with the same material. Effect of the insert version is mostly to improve effective pole width at the outside diameter. Regards YE
@@YesEinstein Do you think 30x20x5?
@@vaclavzeleny5717 No. I think 30x15x5 which has the same magnet mass and number of poles around the circumference as the 5mm insert variant.
@@YesEinstein but I have 30x10x5 or 30x20x5.
3:00 these charts are a bit misleading, they should start at 0 in x & y axis.
And do other rotors have more magnets than the baseline? If so, they are not comparable. It's like comparing a 1.0 L engine to a 1.6 L, and conclude that 1.6 L has more power...
Dear My Opinion. How modest! No data is hidden outside the plotted regions; unlike some popular indices. Comparison comment is fair, as far as it goes. Yours YE
@YesEinstein your video is inspirational..
i want to design my own for 20hp water pump..