Tweet
Hello All,
I thought I would start the ball rolling on a design of Johns not commonly seen these days, the flux gate or G-Flux. For those of you who may not be familiar with the design it can be seen here.
IDEAS AND MOTORS
I have built a preliminary design and done some basic tests, here is a picture of my first model.
And another with considerably more turns of finer gauge wire and larger flywheel. This design (seen driven by an ac motor) was later driven by a 4 pole monopole. @ 850rpm the peak to peak voltage on my scope was over 400v. I set it to charge a 10000uF cap and discharged it via timed mechanical contactor approx every second into a 12v battery. At this pace the cap would discharge at around 20v
As you can see it is fairly close to the basic design. I have used transformer laminations as my "iron pole" pieces, however in light of other similar designs like the Kromrey I wonder if this is necessary, perhaps solid iron bar would yield better performance. The rotor is the same, an "x" made from laminations, rotated by a high speed DC brushed motor (first design).
John was dead right when he said low speed rotor though, this coil produces a lovely sine wave from a very slow speed, my current motor is far too powerful and fast, but it was what I had on hand to use.
John was also dead right about there being no BEMF force on the driving device, as far as I can tell the only thing the driving device must overcome is the cogging present, which is dependent on magnetic saturation (strength and proximity of permanent magnet) and the air gap between "x" rotor and flux paths.
I also noted that a Neo magnet like the ones pictured dont offer significantly more output, I was able to replace them with the standard "domino" sized ferrites we are familiar with from our SG endeavours and achieve similar results. And the permanent magnet side of the Gate must be built like John drew it, if one was to use two "c" cores, one with the coil wrapped around the middle the other with magnets strapped to its "back" there is no results. The magnet must bridge the gap between the iron bars, or extend from the back of the iron bar as can also be seen on Johns page where he says "this is how to gate the magnetic flux"
Can anyone comment on the use of multiple flux paths or phases? I have 6 "c" cores I want to experiment with and have started a simple 3 phase flux gate where an "x" rotor will be used but will only break one flux path at a time (I hope) every 30 degrees. Here is a picture of that assembly.
As you can see, there are no coils or magnets, or rotors yet, but I have my prime mover ready. If I stick to the original (and just add phases) there will only be three coils, and I will need to cut the middle sections out of the corresponding laminations so the magnets can fit. But I cant help wonder if there is a geometry that will allow for all 6 "c" cores to become coils. With a little guidance from a friend I've been considering changing the design to more closely resemble patent no. 7,109,671, but perhaps it would be best to replicate that patent as close as possible with one coil set first.
Any advice people?
Best Regards
Ren
I thought I would start the ball rolling on a design of Johns not commonly seen these days, the flux gate or G-Flux. For those of you who may not be familiar with the design it can be seen here.
IDEAS AND MOTORS
I have built a preliminary design and done some basic tests, here is a picture of my first model.
And another with considerably more turns of finer gauge wire and larger flywheel. This design (seen driven by an ac motor) was later driven by a 4 pole monopole. @ 850rpm the peak to peak voltage on my scope was over 400v. I set it to charge a 10000uF cap and discharged it via timed mechanical contactor approx every second into a 12v battery. At this pace the cap would discharge at around 20v
As you can see it is fairly close to the basic design. I have used transformer laminations as my "iron pole" pieces, however in light of other similar designs like the Kromrey I wonder if this is necessary, perhaps solid iron bar would yield better performance. The rotor is the same, an "x" made from laminations, rotated by a high speed DC brushed motor (first design).
John was dead right when he said low speed rotor though, this coil produces a lovely sine wave from a very slow speed, my current motor is far too powerful and fast, but it was what I had on hand to use.
John was also dead right about there being no BEMF force on the driving device, as far as I can tell the only thing the driving device must overcome is the cogging present, which is dependent on magnetic saturation (strength and proximity of permanent magnet) and the air gap between "x" rotor and flux paths.
I also noted that a Neo magnet like the ones pictured dont offer significantly more output, I was able to replace them with the standard "domino" sized ferrites we are familiar with from our SG endeavours and achieve similar results. And the permanent magnet side of the Gate must be built like John drew it, if one was to use two "c" cores, one with the coil wrapped around the middle the other with magnets strapped to its "back" there is no results. The magnet must bridge the gap between the iron bars, or extend from the back of the iron bar as can also be seen on Johns page where he says "this is how to gate the magnetic flux"
Can anyone comment on the use of multiple flux paths or phases? I have 6 "c" cores I want to experiment with and have started a simple 3 phase flux gate where an "x" rotor will be used but will only break one flux path at a time (I hope) every 30 degrees. Here is a picture of that assembly.
As you can see, there are no coils or magnets, or rotors yet, but I have my prime mover ready. If I stick to the original (and just add phases) there will only be three coils, and I will need to cut the middle sections out of the corresponding laminations so the magnets can fit. But I cant help wonder if there is a geometry that will allow for all 6 "c" cores to become coils. With a little guidance from a friend I've been considering changing the design to more closely resemble patent no. 7,109,671, but perhaps it would be best to replicate that patent as close as possible with one coil set first.
Any advice people?
Best Regards
Ren
Comment