Electromagnetic Coil Design

[travin69]

Hello All.

Looking for some overall suggestions from the forum on designing an electromagnetic coil. Here are the parameters I have so far:

DC powered at 12V. 50% duty cycle. Using 0.500" 1018 CRS as core. 26AWG magnetic wire wound on a bobbin that is 1" long, 0.577" center ID, 2" OD, and 1.750" long. Wire is wound bifiliar and wires are parallel to each other. Each wire is wound about 1000 turns on the coil. Coils are wired electrically parallel with a total resistance of 4.5 ohms. Each coil has an individual resistance of about 9 ohms. For the sake of argument, lets just assume that the coils draw 1 amp each.

Here are my particular questions:

1) Since I am powering the coils in the same direction, do the ampere-turns equal 2000? If not, can someone explain how bifilar coils wound parallel act when wired and powered in the same direction (polarity).

2) Would the coil be stronger if I wound 2 seperate coils of 1000 turns each on the same coil but one coil was wound behind the other?

3) Is there any thumbrule that says that windings that are further away from the coil are less effective? Bascially, are short, fat coils less strong than long, skinny coils assuming windings and power are the same?

4) Do pulsed dc electromagnetic coils have hystoresis / eddy current losses like AC coils do?

Thanks for any help you all can give me.

Daniel

 

[Comcokid]

1) Since the two coils consist of parallel wires for the purposes of any calculation they are treated as if they were 1 wire. Your 1000 turns of bifilar are still just 1000 turns, and 1 amp into each wire is treated as 2 amps on a single coil. It is the same as if you wound with a single wire with twice the cross-section area. Bifilar, multifilar, or litz wire where the wires are connected in parallel give advantages with skin-effect, and skin-effect is of concern with higher frequency switching.

2) 3) For your purposes with DC excitition the two coils can be treated as equal. Now, the inner coil will have less length, and therefore slightly lower resistance, and slightly higher current. As you pulse the coils with increasing frequency, leakage flux issues arise which means the inner coil will magnetically couple stronger to your core, and the outer coil may setup a magnetic field that has a portion that does not include the core, but again these are effects that for your application you can ignore. The high permeability of the steel core u of maybe 30,000 or more) will make sure the magnetic field will want to travel through it. If you were using a low permeability core material (say, koolmu with u of < 300) leakage flux issues would be greater.

4) Since you are pulsing your current, there will be eddy currents. Eddy currents result in core heating, Since you seem to be making an electromagnet, your heating will probably be dominated by I^2 R resistance losses of your coil. Hysterisis is a charcteristic of magnetic cores.

 

[travin69]

Comcokid,

Thank you for the explanation. With respect to questions 1 and 2, would it be true to say no matter how I wound (2), 1000 turn coils, as long as they are connected electrically in parallel, I would always only have 1000 Ampere-Turns?

What I was trying to do was wind multiple coils on the same core and get the maximum power for the least amount of power input. So my thought was to wind bifilar windings or "stack" multiple coils (like you would stack magnets NSNSNS) to increase the power of the electromagnet via more windings. I know as coils are connected in parallel, the resistance goes down. With a lower resistance, I would be able to lower my voltage and get the same current draw, thus lowering my power input, but keeping my Ampere-Turns the same.

Since my application will not allow me to have both poles on the same side, I am kinda stuck with an electromagnet with a pole on each side. Are there any other options to wind a coil with the greatest strength and lowest power consumption?

Thanks for the help,

Daniel

 

[Comcokid]

What it comes down to is current and number of turns. Also at some point you reach the top of the hysteresis curve and additional current & turns does not result in a stronger magnetic field as the field will begin to spread-out from the core (core is saturating) and more flux-lines are traveling through the air. To counteract core saturation you either make the core area larger, or select a material that will handle a higher flux density before saturating. For your application which is essentially DC excitation of the core, I don't know what core material will be optimum.

 

[RyreInc]

Paralleling a second winding is equivalent to doubling the size of the wire. With that in mind, if you inject 1A into a single-wire 1000-turn winding you'll have 1000 A*t; if you inject 1A into 2 parallel 1000-turn windings, the current splits evenly and you have 0.5A in each winding, 0.5A * 2000 turns = 1000A*t. The only difference if the windings were in series would be the impedance, which would be 4 times as high.

With a lower resistance, I would be able to lower my voltage and get the same current draw, thus lowering my power input, but keeping my Ampere-Turns the same.

Equivalently with two windings in series you could halve your current draw for a given voltage. Note that whether in series or parallel the power consumption is the same. It's the increased amount of copper that is saving you power here, not additional windings.