Back stepping a bit, the golf ball is traveling 4 ft. / s or 1.22 m / s which seems a reasonable final velocity
for the stick. The stick is about 18 X heavier than the ball so we only need about 18 X more energy. That is for one coil, so for maybe 3 coils we need only 6 X more energy or with 8300 uF @ 78 V ??

 

Magnetic field strength is directly proportional to the amps times the number of wire turns in a coil, so in theory you could use very thick wire to compensate for the low voltage. But thicker wire also means a larger diameter coil, and that could also affect performance

Well, this is where I started: Fat wire, pulling 50ish AMPs. The metal JUMPED, but I wasn't able to control the switching of that much current; I blew the FETs over and over. This inability to switch the current resulted in me going to the 4ohm coils of thin wire where the cores are lethargic, but controllable. 50 Amps would be great, just don't know how to control.

It is my opinion that said iron cores should stick at least 2/3 of the way into each of your device's coils before activating them. Or at least make sure that your design allows for a wide adjustment of each of its hall sensors to make sure that optimal switching sync is achieved.

I really appreciate this solenoid dissection and share. Thank you.
the setup I have now enables the coils to fire (in theory) when both the following are true:
1. Steel is only on one side of the coil
2. The steel slug is at least half way in the coil

So, sometimes slugs will be 2/3rds in, 4/5ths in... sometimes less, but never less than half way in.
Note: The "In theory" part assumes the hall sensors switch on/off when the steel edge passes in front. Realistically, I'm guessing the new hall sensors will have hysteresis: switching ON when the steel is a bit beyond the detection point and switch OFF when the steel is a bit beyond being absent.

 

A solenoid can only pull the armature into the coil.

Agreed. Fundamental concept behind the design.

I also don't see the reluctance to using a higher voltage on this. Since you talked in other posts about aesthetics, you don't plan on having exposed wires do you? Most things in even a home type environment are using 120V.

I would like to not have a heart attack if I see a curious observer sticking fingers past covers. I also know enough about myself to know I'm going to get shocked at some point while tinkering. It isn't a deal breaker to go higher voltage, just would like to avoid if possible.

And you do know the original pool playing robot used two different actuators, one for most shots and another just for breaks.

Yup, one for hard shots, one focusing on the soft/softer shots. This is a likely outcome.

But that is not how a variable or switched reluctance motor works, in real life. The next coil turns on when the next rotor pole is "near" not under the coil pole piece. It's kind of like the advancing of the ignition in a car engine.

Ok, this is officially not a variable or switched reluctance motor. Please review the video explanation posted in #421. The sequence of coil firing and associated steel positioning is detailed there.

 

So, sometimes slugs will be 2/3rds in, 4/5ths in... sometimes less, but never less than half way in.

What I'd consider is having the first slug 2/3 into the first coil, then the next one 1/2 (or so), the next one 1/4, and so on. The last one can be completely outside of the last coil. What I'd do is fire the first coil, then switch it off and fire the next one when it's 2/3 in, and then switch it off and fire the third one when it's 2/3 in, and so on ... that way you'd end up firing only one coil at a time while imparting almost constant acceleration to the whole assembly.

 

Back stepping a bit, the golf ball is traveling 4 ft. / s or 1.22 m / s which seems a reasonable final velocity
for the stick. The stick is about 18 X heavier than the ball so we only need about 18 X more energy. That is for one coil, so for maybe 3 coils we need only 6 X more energy or with 8300 uF @ 78 V ??

My hope is that the coils can add force to one another... sequentially... as if the golf ball got a second kick mid-air, then another, and another...
The forces then accelerate the stick multiple times increasing the final velocity. Again, this is the hope. I believe I can get 9 kicks in one stroke, but need to look at it again.

Side note: I received the new hall sensors today. Should be able to test tomorrow to see if I can advance or have to back step to find another means of detecting steel positioning.

 

What I'd consider is having the first slug 2/3 into the first coil, then the next one 1/2 (or so), the next one 1/4, and so on. The last one can be completely outside of the last coil. What I'd do is fire the first coil, then switch it off and fire the next one when it's 2/3 in, and then switch it off and fire the third one when it's 2/3 in, and so on ... that way you'd end up firing only one coil at a time while imparting almost constant acceleration to the whole assembly.

AND, as Sensacell suggested, I have a second set of 3 coils 1/4 step out of phase. This means that 2 coils are always firing, one with 3/4 steel in and one with 1/2 steel in. (or more)

Yes, almost constant acceleration is the hope, but I do expect some torque ripple, just not sure how much yet. (I guess it isn't "torque" since there isn't twisting... but the equivalent in linear terms. Perhaps "thrust ripple")
A LOT depends on how these new high sensitivity hall sensors perform.

 

The capacitor bank is there to help the system with the sudden current demand when you switch on the mosfets, this is because the batteries (or power supply) that you use always have an internal resistance that limits current flow and affects the rate at which the magnetic field rises in the solenoids.

So, if the capacitor bank is charged by the batteries (let's say 2x12V = 24V), then the capacitor bank can only charge up to 24V, does that help alleviate the internal resistance issue, increasing the rate at which the magnetic field rises?
This is what I'm considering... https://www.amazon.com/gp/product/B001DXAVTS/ref=oh_aui_detailpage_o08_s00?ie=UTF8&psc=1

 

Perhaps "thrust ripple"

yeap ... that'd be the correct term ...

AND, as Sensacell suggested, I have a second set of 3 coils 1/4 step out of phase. This means that 2 coils are always firing, one with 3/4 steel in and one with 1/2 steel in. (or more)

Then I suggest you start with baby steps. First, try to smoothly sequence two coils before adding a third to the circuit, then a fourth, and so on.

A LOT depends on how these new high sensitivity hall sensors perform.

I'd say that a lot depends on the repeatability of said sensors. And how perfectly distanced you install them from the slugs.

 

does that help alleviate the internal resistance issue, increasing the rate at which the magnetic field rises?

It depends on the coil's resistance itself. And, regarding solenoids, internal resistance and total inductance are two things that must be balanced together, especially when space is restricted ... which it always is. That's because the more layers of wire you add to a coil, the new layer will be at a longer distance away from the slug than the previous one, and hence the magnetic field it induces will be less. There comes a point in which adding more layers becomes detrimental to performance.

 

Ok, this is officially not a variable or switched reluctance motor. Please review the video explanation posted in #421. The sequence of coil firing and associated steel positioning is detailed there.

Again you can call it what ever you like. But in configuration your going in at the moment, it is a SRM, switched reluctance motor, that is the name for it in any literature you will ever find. Other than this thread.

 

yeap ... that'd be the correct term ...

... now that I think of it ... the correct term in motion control for an unsmooth acceleration profile is the word "jerk" ...

 

The "In theory" part assumes the hall sensors switch on/off when the steel edge passes in front. Realistically, I'm guessing the new hall sensors will have hysteresis: switching ON when the steel is a bit beyond the detection point and switch OFF when the steel is a bit beyond being absent.

The Hall sensor could be affected by multiple things: (1) the mere presence of the steel slug, (2) magnetic pulses produced by energising the nearby coil and (3) magnetic pulses induced in one coil by energisation of the neighbouring coil (since the two coils would be magnetically coupled via the steel slug). It would be interesting to see a scope shot of the sensor output.

 

My take is that the rod at the bottom of the solenoid is there to help concentrate the magnetic field's lines created by the coil when it is activated,

That secondary, fixed rod within the solenoid coil has had me thinking.
I wonder if a thin metal tube between the coils and the steel slug in my system would concentrate the magnetic field lines. If that is the case, manufacturing the coils would certainly be easier... just wrap the coils around a thin metal tube. The thin metal tube would also be far more durable than having the cue slide inside the epoxy/coil inner surface of the coil. Definitely a next version idea, but fabrication would be easy and if performance is better...
Any experience related to metal tubes wrapped in wire vs no metal tube, just coils?

 

the two coils would be magnetically coupled via the steel slug).

Interesting. I hadn't considered that the coils would still be magnetically coupled. The gap between coils is equal to one steel slug width (with the exception of the coils between the 2 sets of 3 coils, that gap is 75% the thickness of a slug). So, the same piece of steel shouldn't ever be within 2 coils at once, with the exception of that one 75% gap width.

It would be interesting to see a scope shot of the sensor output

Yes, yes it would. I'll see if I can score a scope. What would be even better would be to have precise position data to accompany the scope data.

 

I wonder if a thin metal tube between the coils and the steel slug in my system would concentrate the magnetic field lines.

If you mean one long steel tube with many coils, won't the whole tube be magnetized as each coil is turned on?
Where this would work and is done in all motors with more than one coil is if the steel tube was on the outside of the coils. In a motor it is known as "back iron". A better way of concentrating the magnetism, and one I mentioned in the beginning of this thread, is to make individual steel 'bobbins' to wind the coils on. Look at a sewing machine thread bobbin to see what i mean.

 

he gap between coils is equal to one steel slug width shouldn't .

Again this is a commonly known thing, if researched. The gaps between coil poles and "slugs" is a ratio that has been proven out over the years. And most people making them follow those ratios. I may still have the PDF but won't take the time to look for it unless you really want it. You may not because it comes from a motor called a SRM not what you're making.

 

My hope is that the coils can add force to one another... sequentially... as if the golf ball got a second kick mid-air, then another, and another...
The forces then accelerate the stick multiple times increasing the final velocity. Again, this is the hope. I believe I can get 9 kicks in one stroke, but need to look at it again.

Again not how I look at this or how it should be done. Or how a motor rotary or linear normally works. Wouldn't it make more sense that more that one coil and more than one 'slug' gets accelerated at a time? It is how something is moved smoothly, like when your arm makes a pool shot. One group of muscles in your arm doesn't contract then another group takes over after the first is done. They all contract at one time, that gives a smooth and strong movement.

I don't mean all of the coils turning on at once just the ones next to the slug. I'll have to take the time to draw this out and post it.

 

I think @shortbus has it right " if the steel tube was on the outside of the coils." If on inside it could shield the slug from magnetic forces and act as a low resistance shorted turn.

 

I only had a brief moment to test the new hall sensors prior to getting pulled away from the shop. Results weren't encouraging. I'm hoping the problem was the test jig. I'll solidify the test setup and try again Monday.

If you mean one long steel tube with many coils, won't the whole tube be magnetized as each coil is turned on?

Clarification: Yes, my question was about each coil getting its own tube, not one long tube... for the reason you describe here.

if the steel tube was on the outside of the coils. In a motor it is known as "back iron". A better way of concentrating the magnetism, and one I mentioned in the beginning of this thread, is to make individual steel 'bobbins' to wind the coils on

After doing some reading about Back Irons (thanks for the vocab heads up) I'm seeing what you're saying about having metal on the outside of the coils. This enables the magnetic field to "flow" (correct term?) through ferrous material which is better than through air. If metal on the outside helps, I'm guessing a thin tube on the inside would do the same... but I'm seeing Bernard say:

If on inside it could shield the slug from magnetic forces and act as a low resistance shorted turn.

The bobbin idea is definitely metal on the inside... Testing will help clarify. But not a top priority now.
I do have hesitations about the metal on the end of the tube in the bobbin example... feels like that would interfere directly with the field.

Or how a motor rotary or linear normally works. Wouldn't it make more sense that more that one coil and more than one 'slug' gets accelerated at a time?

This project is certainly not a normal application, so I'm fully on board with being out of line with what normally works.
It does make sense to activate multiple coils. The video in post #421 and the table in post #423 show that 2 coils will always be on.

 

The bobbin idea is definitely metal on the inside.

Which is how a solenoid is made, especially a solenoid for a solenoid valve, or the one on a car starter motor. Sorry that I keep trying to get/give you help instead of cheering you on down a path that goes against all of the available facts on doing this. But won't stop until forced too. Too old and stubborn, and spent too many hours reading and buying books on motor theory.