Back EMF vs. Collapsing Magnetic Field Spike

[nilrehob]

You quoted me but I didn't say that

My mistake, I quoted Matt, sorry

[nilrehob]

I would like to discuss how to minimize the energy-loss in a coil by using the established formulas in the textbooks (to start with).

According to textbooks:
* the time-constant for a coil is T=L/R
* the current in a coil is I=U/R (1-e^-t/T)
* the energy in a coil (possibly coming out of it) is E(o)=1/2 LI^2
* the energy fed into a coil is E(i)=UIt

If I remember my calculus class correct, then E(i)=U^2 /R (t+Te^-t/T).

In the attachment I have made a graph, x-axis is t going from 0 to 5T (which is considered when a coil is saturated).

The lines are:
* blue E(i) energy going in the coil
* red E(o) energy coming out from it (when current is shut off)
* yellow E(i-o) energy loss
* green E(o/(i-o)) energy out:loss ratio
* brown E(o/i) energy out:in ratio

My conclusions, based on this, is that the time the current is going through a coil is best to be kept at a minimum.

My guess is that this explains the rather complex commutator in Newmans motor, where the coil is fed by many short pulses for each turn?

Any comments or corrections anyone?

[nilrehob]

Is there really no one who dares to comment on this?
I'm not saying that there are no such thing as radiant energy, and I never will (simply because I don't know), thats not the debate anymore. Really!

So come on; has anyone had any experience concerning shorter or longer pulses into a coil in regard to efficiency?

[lighty]

For the optimum input/recovery ratio of some particular coil the duration of the energising impulse should be such to provide maximum core saturation and should not last a moment longer. When you stop energizing current at the point of maximum core saturation (depending on the permeability of core material and on the hysteresis curve characteristic) you will get maximum recovery for that particular coil.

The point is, you can only saturate core material up to a maximum point and after that you can prolong duration of energising impulse to infinity, waste as much energy as you want and you will still get same recovery. In short, one should adjust duty cycle so that it ensures maximum core saturation and should not prolong its duration after that point.

[nilrehob]

Thanks for the reply, lighty!

So Your recommendation is to energize it as long as 5T (where T=L/R) ?
I'm thinking perhaps of a much shorter time.

Have You looked at the graphs in my post above?

[lighty]

The simplest way for you to optimize duration of impulse would be to do it with oscilloscope by measuring current curve and look for the saturation plateau. I described the whole procedure somewhere in Lindemann attraction motor thread. I mean you can do calculations to get some general values but the moment something changes (different batch of material, different geometry, different semiconductor etc.) you will have to recalculate it all over again. I personaly calculate some general values but then I adjust everything manually.

[Electrotek]

Is there really no one who dares to comment on this?
I'm not saying that there are no such thing as radiant energy, and I never will (simply because I don't know), thats not the debate anymore. Really!

So come on; has anyone had any experience concerning shorter or longer pulses into a coil in regard to efficiency?

I did a test on just this very concept. Using two four pound repulsion coils, the first discharge was at a normal speed, but the second discharge was faster due to magnetic quenching. With the same energy in both pulses, the faster pulse produced more kick.

[nilrehob]

It all depends on what the goal is: do You want as much energy from the coil in the spike as possible, or do You want as high energy out/in ratio as possible.

I'm aiming for the second one: to get as high energy out/in ratio as possible, to make it as efficient as possible and not waste any energy.

[nilrehob]

I did a test on just this very concept. Using two four pound repulsion coils, the first discharge was at a normal speed, but the second discharge was faster due to magnetic quenching. With the same energy in both pulses, the faster pulse produced more kick.

Cool! I wonder how short the input can be without making the output energy too small for any good use?

[lighty]

@nilrehob
You already got your answer. Core can be saturated to some maximum value and no matter what you do and how much additional energy you give you cannot saturate it further.
So the ratio is the same in both cases- if you want maximum energy from the coil you drive it to saturation and then immediately discharge it. If you want best input/recovery ration you drive it to saturation and then immediately discharge it. The only thing you have to observe is to drive coil to the point of saturation (on the time scale) and then immediately discharge it.

The easiest way to set the optimum ratio (which is at the same time maximum recovery) is to watch current graph on your scope. You can do that either by current transformer or by measuring voltage drop over shunt. Start with the shortest duty cycle and then slowly prolong it. You will see, as your duty cycle gets longer that current graph on your scope also gets longer but it rises in a rather linear way. At a certain point your graph will reach a kind of plateau where it suddenly is not linear any more. That means that you either reached point of complete saturation or that you're at least so close to it that you would have to add unproportionally large amount of energy to saturate it even a little more. So, now that you've determined best duty cycle for that coil and that given frequency you can check out the recovery spike and you will notice that no matter what you do it won't get any higher. Also, now that you know the current needed for complete saturation you can easily calculate number of ampere/turns so you can adjust number of turns of your coil (when using that core) to suit your particular needs (this is when calculation is very handy).

It's all extremely simple and I don't see a need to make it more complex for all the practical reasons. Yes we can discuss some other topics like Q factor, effect of temperature on the maximum core saturation, the effect of capacitance of windings, geometry of core and coil, the ways to further increase maximum saturation of core etc. However, I doubt most of the people will be able to follow the discussion and there are more suitable forums on the net dealing with EE on which one can discuss such things. Also I think that info on setting the proper duty cycle for some given coil and frequency is what most people are seeking after.


@Electrotek
Input/recovery ratio is almost linear up to a point of core saturation. It's very obvious when looking at the current curve on the scope and it's easier to measure it. Nice experiment though.

[nilrehob]

@lighty Your missing the point.

[lighty]

@nilrehob

How so?

[Matthew Jones]

@Lightly
What would be the best way to calculate core saturation based on the materials with in a core. Or where is a good place to find some of that data.

Thanks
Matt

[nilrehob]

@lighty

By using the the formulas that I'm aware of I have found a (well-known) non-linearity in the way coils behave.
The formulas are probably not correct in the way that we are using the coils (with our abrupt disconnection), just as Newtons formulas are not correct when speed is very high, but these are the formulas that I currently know and understand.

If my calculations are correct and they have some connection to reality, it seems like we might increase efficiency by not waiting until the coil is fully saturated (which theoretically never happen but by practical reason is being considered at 5*T where T=L/R) and then cut off the supply, but to only connect it for a fraction of T before disconnecting it.

If I'm right one should energize the coil as short as possible with as high voltage as possible. Only in this way will the reactant part of the coil be fully used.

It's all in the non-linearity of the functions in the graphs:

[nilrehob]

@matt

Here are lots of info on coils, maybe You can find something useful:
Producing wound components