Energy harvesting using a brushless generator

Thread Starter

cmartinez

Joined Jan 17, 2007
7,633
I'm not so sure that the illustration is correctly showing how the magnetic force would be in the way it's expected. With the magnets arranged like this . If they were arranged with all of the poles the same yes, but with them alternating I don't think so. But could be wrong, and would like proof. I'm basing my thinking on how a magnetic work holding chuck is made, how the lines of force are in them.
I didn't understand it either, until I visualized that not only the magnet's poles are alternated but the coil's directions too. For instance, the wire in coil #1 is wound in a counterclockwise direction, while the wire in coil #2 is wound clockwise.

Generator.png

Let's assume that the magnets are rotating clockwise. When magnet A crosses the pair of coils #1 and# 2, it induces a current flowing away from the center of rotation, while magnet B induces a current that flows towards the center of rotation (because it has the opposite polarity of magnet A) as it crosses coils #2 and# 3. This creates a current flowing counterclockwise in coil #1, and clockwise in coil #2. And the sum of both flows in the same direction inside the wire, because of the way that coils are connected. If the coils directions and the magnet's polarities were not alternated the current in each coil would "crash" against each other, and it wouldn't flow out of the device. The only result would be that heat would be generated.

All you have to do now is visualize this same situation in the rest of the coils. And as you can see, only alternating current can be produced this way.
 
Last edited:

Danko

Joined Nov 22, 2017
1,273
I'm not so sure that the illustration is correctly showing how the magnetic force would be in the way it's expected. With the magnets arranged like this . If they were arranged with all of the poles the same yes, but with them alternating I don't think so. But could be wrong, and would like proof. I'm basing my thinking on how a magnetic work holding chuck is made, how the lines of force are in them.
I think cmartinez gave us comprehensive explanation.
@cmartinez :
Tell me please dimensions of your magnets and number of magnets you will use in generator.
 

Thread Starter

cmartinez

Joined Jan 17, 2007
7,633
I think cmartinez gave us comprehensive explanation.
@cmartinez :
Tell me please dimensions of your magnets and number of magnets you will use in generator.
Magnets are 10mm diam x 5mm thick, and I'll be using 8 of them. I've already made a drawing using the proportions you suggested and it fits perfectly. My last question to you would be the wire thickness and number of turns per coil that you'd suggest. Of course maybe the only way to know whatever configuration works for what I want is to actually build this thing and test it.

I was thinking about using ga #42 wire (the one normally used for guitar pickups, for instance) but maybe its diameter is too small and the resulting resistance will be too high, diminishing the total output power. So I was thinking about building three coil sets using wire of #42, #38 and #34, see which one performs best.
 

Danko

Joined Nov 22, 2017
1,273
Only way to lower coil resistance is lower gap between magnet and soft iron disk.
Stronger magnetic field - lower turns for same voltage.
So, if gap will 2mm and coil thickness 1.5mm, then gaps coil-magnet and coil-iron disk will 0.25mm each.
 

Thread Starter

cmartinez

Joined Jan 17, 2007
7,633
That last (coil) drawing wasn't included before.
Oh yes it was ... Danko posted it more than a month ago. But the image is embedded in his post as a thumbnail. Maybe that's why you missed the coil drawing detail.

I thought you wanted DC not AC?
yup ... I want DC, so I'm going to somehow have to rectify its output, depending on the minimum voltage that I get out of it at the minimum rpm's that I'm planning on running it at. Or maybe I'll use a special harvesting ic that I found out there (but that is a bit too expensive for my taste, about $7.00 dlls) that is capable of handling an ac input and outputting regulated dc.

As for the EHT game ... nah.. I'm neither that mysterious, or fancy spoken :p... so if you want further clarification, ask away and I'll see how much I can disclose.
 
Last edited:

Thread Starter

cmartinez

Joined Jan 17, 2007
7,633
Only way to lower coil resistance is lower gap between magnet and soft iron disk.
Stronger magnetic field - lower turns for same voltage.
So, if gap will 2mm and coil thickness 1.5mm, then gaps coil-magnet and coil-iron disk will 0.25mm each.
I see ... but my intuition tells me that there's a "sweet spot" on the ratio of turns vs the wire diameter and the coil thickness (which can be interpreted as the gap between the magnet and its opposing iron plate) in which power generation can be optimized... as far as rpms are concerned ... never mind the torque for now.
 

Danko

Joined Nov 22, 2017
1,273
I see ... but my intuition tells me that there's a "sweet spot" on the ratio of turns vs the wire diameter and the coil thickness (which can be interpreted as the gap between the magnet and its opposing iron plate) in which power generation can be optimized... as far as rpms are concerned ... never mind the torque for now.
OK. You can check your intuition by easy way:
1. Build generator without coils for now.
2. Rotate generator by DC motor with controlled, stable, measured RPM.
3. Made Probe Turn:
Probe-Turn.png
4. Place probe turn accurate in the middle of the gap.
5. Measure V at desired I_load through R_load (I_load is calculated using V and known, step by step changed R load).
6. Write V and gap values to chart for constant RPM and constant I_load.
7. Slightly change gap and repeat points 4...6.
8. Write calculator program, using data from chart and specific different gauge wire resistance.
Program will calculate generator voltage at desired RPM and I_load, for different gaps and wire gauge.
EDIT: Of course program will calculate cross section area and number of wire turns in coil.
 
Last edited:

Thread Starter

cmartinez

Joined Jan 17, 2007
7,633
I've already finished assembling, programming and testing the coil winder controller (along with its software) and it's performing beautifully. Also, I'll be receiving all of the mechanical parts and commercial components this week. So, God willing, the machine will be operational this coming weekend, hopefully.

We have already established that the metal plate in front of the magnets is extremely important:

Generator.png
My question now is: How important is it for that plate to be spinning along with the magnets? I'm asking because the device would be much simpler to build if that plate were fixed, instead of rotating. I know that parasitic eddy currents will be generated on the plate due to the magnets spinning relative to it, and some power will be wasted. But how much would efficiency be affected if that plate were fixed? How much would efficiency be affected?

As a reminder, the magnets are 10mm in diam x 5mm thick, and they'll be spinning at between 100 to 300 rpm.
 

Danko

Joined Nov 22, 2017
1,273
I think it will dissipate much more then 50% of energy.
1. Eddy currents will create big braking torque.
2. Magnets and plate will attracted to one another with enormous force, therefore problem with bearings, generator will simple stopped.
 

Thread Starter

cmartinez

Joined Jan 17, 2007
7,633
Again, it seems that I've underestimated the importance of that plate ... but now my curiosity has been piqued ... It shouldn't be that hard for me to test the difference once the device's setup is complete. Many thanks to both, Alec and Danko, your observations are always thoroughly appreciated.
 

MisterBill2

Joined Jan 23, 2018
11,624
I am suddenly seeing this thread again after not seeing it for a long time. While I am pleased to see it, since it is interesting indeed, I am very puzzled as to why I did not see it for so many weeks. Now it is at 114 comments, it was still on the first page when I saw it last. And I do have some questions about automotive alternators but that is probably better in a separate new thread.
Glad to be back,
MisterBill2
 
Top