AC air core coil of 0.2 T 100KHZ to 100MHZ

[rbobandcarly]

I am a student at Colorado State. Can someone in this forum help me with building an AC coil either an air core coil or C-frame iron or steel with so many turns of copper wire that has a maximum magnetic field of 0.2 T and its frequency range is 100KHZ to 100MHZ. I am searching on the internet.
Thanks,
Bob

 

[marcusl]

While reaching that field strength at DC is easy, doing it at high frequencies is very difficult. What is the volume of uniform field you are after? That will also determine the difficulty. Here are some general comments:

1) You can completely forget about iron due to the skin effect.
2) Look up Litz wire, this will probably be necessary even to get to 100 kHz. (Skin effect again.)
3) You probably won't reach 100 MHz. Interturn capacitance will eat you alive long before then.

Suggest you start Googling terms like RF inductor, Litz wire, interturn capacitance. Then come on back with more questions.

 

[rbobandcarly]

Thanks for responding. The bio sample is 1 1/2 inch diameter. I don't know how long the coil needs to be. I am assuming for high freq like 100 Khz to Mhz, it can't be that long, let's say 3inches long.
I'll check out litz wires.
Marcus, Would solid block copper metal work instead of iron?
Thanks,
Bob

 

[marcusl]

Read about skin depth--you'll see that the penetration of fields into a good conductor like copper is only 6 microns at 100 MHz. Any conductor that is thicker has zero field inside. That's why the coaxial cable used at these frequencies (and above) often has a steel center conductor that is plated with an almost infinitesimal layer of silver. Steel is cheap and poor conductor but at high frequencies all the current is carried through the surface layer of silver, which is the best room-temperature conductor.

Ferrites are the only practical core materials for high frequencies, and you won't have and easy time finding a yoke in the size you mention. You'll need to be an expert in RF to make this work, as well.

What is your application?

Read about skin depth, Litz wire and interturn capacitance before coming back. You'll also need to learn some RF engineering--maybe from the ARRL handbook, for instance.

 

[rbobandcarly]

Marcus, I have lots to read. Thanks for the info. My application at the University is to move iron nanoparticles attached to polymers. The goal of the project is to see if the nanoparticles are swept back and forth based on the magnetic field changing, for example 60 Hz the magnetic field would switch 120 times a second.
Thanks again, I'll be back when I gather more information.
Bob

 

[marcusl]

Good grief, why did you say you need to work to 100 MHz if you really only need 60 Hz? It's the difference between walking next door, and going to the moon and back.

 

[rbobandcarly]

You are funny, Marcus. 60 HZ is what I think would work, but the researchers still want from KHZ to MHZ range. I'm not sure if it is possible to move a big size magnetic field like 0.2 T 100 MHZ, I'll bet it can be done, but it won't be easy.

 

[marcusl]

Why don't you do a calculation of the expected motion of your particles in your medium. You are in a regime of low Reynolds number, so I expect you'll find that viscous forces exceed those of the applied fields at a rather low frequency.

EDIT: Better yet, have your "researchers" do their homework. They should calculate what is reasonable, or even possible, on the one hand, and utterly ridiculous on the other. You have my permission to share this thread with them )

 

[rbobandcarly]

Marcus,
I let my co-researchers know about our blog. I did find a commercial RF power supply that puts out 50 to 485KHZ at an RF power of 10KW, coil voltage of 1100 Vrms. I wonder what kind of magnetic field it would produce if I wind one to two turns of copper tubing say 1/8 inch diameter to 1/4 inch diameter.
Let me know if this is a realistic approach, to use the power supply and protective capacitors (across and in series with the coil to protect the power supply from reflected RF signal) and this coil arrangement.
Thanks,
Bob

 

[rbobandcarly]

Marcus,
Here's what I found out so far. To recap what's written below is use two copper tubing coils (2 turns each) in helmholtz configuration where their distance apart is equal to the radius of the tube coils. A high voltage pulsing circuit, pulses one coil and then the other microseconds later so that the magnetic field changes direction each time the control circuitry pulses the coil.
Take a look at the following.
Thanks,
Bob


A second very promising approach is also being pursued. Since work began on the high Q tank circuits,

IGBTs (very high current on-off solid state switches) have become available that can each switch 3000 A at

1500 V at 0.5 MHz for a several hundred microsecond pulse. Two power supplies are presently under

construction that will each use twelve IGBT’s in parallel. Figure 4 shows the circuit schematic.

Here’s our power supply, suggest using two copper tube coils (2 turns) in helmoltz configuration and this circuit would pulse each of the copper tube coils. Helmoltz configuration means both coils are separated by the radius of the copper tube coil.







The IGBT output

will pass through a 20:1 air core transformer and into a parallel LC tank circuit (Q~100), where the inductor is

the RMF antenna. By presenting a high impedance only to its resonant frequency, the tank circuit transforms

the output waveform of the IGBT’s into a clean sinusoid. Additionally, the tank circuits high Q greatly amplifies

the real power delivered with circulating power, enabling the creation of a .01T RMF. With a low total

inductance of ~20nH in the primary side of the circuit, as the plasma load increases, the IGBT’s should be able

to provide the current necessary to sustain the secondary voltage, and hence maintain a constant RMF. For

Radius =

0.36 m Width =

0.6m

Separation = 0.36m

Quartz Tube

Radius = 0.2 m

Contours mark .001 T

increments of BRMF.

tens of microseconds, the maximum current output of 3000A can be exceeded by a factor of 5 to 10. The peak

output power may not be as high from the IGBTs as from the LC tank, but they have the advantage of both a

longer quasi-steady duration, and an easily variable frequency for experimentation.

FIGURE 4. IGBT Power Supply, Transformer, and RMF Parallel Resonant Circuit.

 

[rbobandcarly]

Marcus,
I forgot to mention in my last entry that the copper tubing and power supply will be cooled with running cold water through both copper tubing coils and power supply.
Bob

 

[marcusl]

Wow! Operating at 1500V through twelve 3000A devices in parallel. Very impressive. Sounds like you have an EE or two on the team, and, lacking the interest or ability to calculate field strength (or gradient), induced forces, frequency response, losses, expected SNR or any other aspect of your experiment, have gone with your strength—building a multi-Megawatt induction furnace of the sort used by heavy industry.

Please coordinate with your energy utility so you don’t brown out the city of Fort Collins when you turn it on...

 

[rbobandcarly]

Marcus, you are a very very funny person. Our goal as students is not to brown out Fort Collins, but to use house current and voltage.
Instead of copper tubing we are going to go with solid copper wire.
Again, I'm throwing out ideas, but let's include math from now on, if that's OK with you.
Let's say we go with 4/0 copper wire, it has a diameter of 0.522 inches and a resistance of0.047 ohms per 1000 feet.
Marcus, help me with this calculation.
Let's say we wind this wire 100 turns with diameter of 3 inches, the formula for Inductance L is r squared times N squared divided by ((9 times r) plus 10 times L), where r is inductance in nHenries, r is outer radius of coil in inches, L is length of coil in inches and N is number of turns.
Once we have L, I guess we can calculate B magnetic field if we have the current? Is that correct, I haven't calculated the inductance L, but I'll work on calculating both the inductance L and magnetic field B.
Talk with you soon, again please chime in if you have done the calculations.
Many thanks for your support,
Bob

 

[trevie1970]

Just out of interest is this Three phase or Single Phase. Do you know what the True power / apparent power will be? If this is watered Cooled, I am sure using water is not a good idea? How will you isolate the Water from the electricity?

 

[rbobandcarly]

Trevie,
As you read from Marcus' replies, we were going to go full speed ahead and use 3 phase, but we decided to apply the KIS principle, Keep it Simple.
So our plan is to use single phase no cooling at this time, although I have a funny feeling that because of the level of magnetic field of 0.2 T we may need to cool with water. Our requirement frequency has been lowered to 300 Hz, which may make this project a lot easier to deal with.
Now to answer your question, I don't know what the True power and apparent power is at this time, I'm at the stage of figuring out how many turns of wire I need and current and magnetic field, B.
We started out thinking about using copper tubing, but now we are heading towards using a solid conductor copper wire with insulation around wire and just winding it around air-core.
We are not at the water stage yet, but that's a good point you brought up, how are we going to keep the water from shorting things out. We are not there yet, very valid point you bring up. Feel free to add your ideas on constructing air-coil copper winding coil producing a B field of 0.2 T at around 300 HZ. Thanks, Bob

 

[rbobandcarly]

Here's what I calculated from the inductance formula for air core coil. Diameter of coil is 3 inches, wire diameter is .52 inches, coil length is 13.3 inches, and number of turns is 25.5, the inductance is 10 microHenries.
Now using the formula N times I = (B times L) / u0
Substituting for the B we want of 0.2 T in the above formula, L is the length of the coil or 0.338 meters, N is number of turns or 25.5 round off to 26, uo is 4 times pi E-7
the result is 2113 amps, a continuous ac wave won't work, but a short interval of time pulse may work, we may also have to cool the coil by running tap water through it.
Any suggestions from the forum is appreciated? I'm not sure if I'm on track with the equations used.
Thanks,
Bob

 

[rbobandcarly]

Now this power supply can put out 2000 amps at millisecond interval:

http://www.microwelding.co.uk/3-1-1-2-1.asp
This power supply accepts single phase input.
Millisecond time intervals work out based on the L/R time constant.
With L equal to 10 micro Henries and R for the 13 inch long coil is around 0.001 ohms so the L over R time constant is 10 milliseconds, which works out, the only thing that we have to do is cool the coil to keep it from burning up, I believe water cooling it will prevent burn up.

 

[marcusl]

Hi Bob,

Talk with you soon, again please chime in if you have done the calculations.

I think we have some mutual misunderstandings here.

1. This is your research project, not mine. This means that you get to do the calculations for your project. I won’t do them for you.
2. It also means that you need to research and master the background topics needed. That means taking books on, for instance, electromagnetics out of the library, understanding the principles, and identifying the appropriate equations to use. Physics, chemistry and engineering faculty at your school are also good sources of information and advice.
3. You miss my point when you ask me to validate design features of your induction furnace, arc welder or whatever you are calling it today. I again encourage you to plan your experiment from the bottom up, not from the top down.
   
   a. What is your concept for detecting the signal? How do you plan to see the signal in the presence of the driving field? How much isolation between them do you need, and how much are you likely to achieve? (You don’t need to tell me the answers to these or any other questions, but you need to think them through if you want to get good results.)
   b. What is the background noise level, and the minimum detectable SNR, of your sensor/receiver/detector? Is it good enough to see your signal?
   c. Part of the answer to 3b) involves the force on the particles, which in turn depends on the magnetization induced in them. Have you calculated these?
   d. Had you tried to calculate the force, BTW, you would have realized that your Helmholtz coil field generator will not work.
   e. Another part of the answer to 3b) depends on the particle motion in response to the applied force. Nanoparticles move in an unfamiliar regime of high viscous forces (low Reynolds numbers), and motions of polymers in so-called melts are even more complex, so the answer might contain surprises.​
   
   In other words, the machine you are designing might not be the machine you want for your experiment to succeed. Constructing what you know how to do, while disregarding theory and other aspects of the experiment, is an uncertain path to success.
4. I’m hoping that someone else on your team, like your principal investigator or mentor/advisor, has already looked at these things.

That’s a long explanation of why I’m not planning to comment further on the details of your megawatt machines.

 

[chitrageetam]

Hi ..
I have posted another question on AC coil before seeing this post.
My question was...
What happens to the magnetic field in a solenoid coil
when the frequency is increased, and by keeping the current constant?
As per my idea, The field must remain constant as the current is constant.
But I am not very sure..
Can you answer this question?
Consider a coil of length 2 inch, inner diameter 0.2 inch.
Frequency range - 100 Hz to 10 kHz.
Power consumption increases because of eddy and skin effects but how about H?
Regards
Chitra

 

[rbobandcarly]

Marcus,
I appreciate your input. Thanks for sharing. Good points you made, however, We (our research team) need to press and put this together as soon as possible, after the Thanksgiving holiday. Your comments are well taken and feel free anytime to write about this project. Believe me, I need all the help I can get since this is new territory for me, however, I won't delay it either.

chitrageetam,
Interesting comment. We are limited by sample size of 1.5 inch for the nanoparticle wafer, I assumed 3 inches so we have room to wiggle.
With a 3 inch diameter, what do you recommend the length of the air core coil should be?
You are right, the higher frequency, the more significant that we have to take into consideration skin effect, since electrons like to travel towards the outside of the conductor.
But in our case, a frequency of 300 Hz, skin effect may not be overly significant.
Thanks,
Bob

 

[chitrageetam]

Mr.Bob..
Thanks for reply..
I mentioned the coil size not as a suggestion to your task but only to
understand the problem ..."weather the field in AC coil is same as DC coil?"
Did you ever notice any drop in field with the increase in frequency, with same current?
Thanks

 

[rbobandcarly]

chitrageetam,
After Thanksgiving if the team agrees on this design and we begin ordering the parts and putting 0.5 inch diameter solid copper coil #4/0 and wrapping around 30 turns will give us a foot and a few inches in length for the air core coil length and diameter of coil is 3 inches with a current of 2000 amps ON for milliseconds should provide around a magnetic field of 0.2 T. The goal is to use a DC power supply that is used for welding that puts out 2000 amps for short period of time, use two 2000 amp thyristors or triodes to switch the field from one side of the coil to the other side of coil (it is important that we switch the field direction to cause the nanoparticles to move back and forth) and water cool the copper coil to keep it from burning up, and isolate the water with PVC pipe or something of similar material, and then measure the magnetic field with a hall sensor and see if the magnetic field is around 0.2Tesla.
I'll let you know what the magnetic field is after we begin ordering the parts and assemble the project, we'll need to bolt the PVC pipe or the whole thing will fling out so fast with such intense force.
Thanks,
Bob

 

[chitrageetam]

hi Bob..
Did you say that you want to switch the direction in milliseconds,..meaning kHz?
Okay, if you use DC switching, it will be like a sqaure wave rather than a sine wave..isnt it?
It will not ramp up or ramp down.
Well, even before conducting the experiment, do you have theoretical numbers?
Thanks

 

[rbobandcarly]

chitrageetam,
Yes, the plan is to use the triode or thyristor that can withstand 2000 amps and switch it ON and OFF through a low voltage control circuit that puts out say 2 volts at the control input line of the triode. We will also have to cool both triodes, besides both helmholtz coils.
Instead of a ramp, we are looking to provide a square wave DC voltage, low voltage around 6 volts and 2000 amps on one coil and turn OFF the first triode and turn ON the second triode like a NOT gate so that the same 6 volts and 2000 amps is directed at a similar second coil (helmholtz coil arrangement) so that the 0.2 T field is moved in the opposite direction, this sudden change in 0.2 T field will cause the nanoparticles to sweep directions and this is the whole gist of this project, the see motion of the nanoparticles sweep back and forth, which can be done with luminescence and other techniques.
Thanks,
Bob

 

[rbobandcarly]

Hey there Physics Forum:
We met for our design of the 0.2 Tesla coil with iron bar added in the middle instead of air core and the team made the decision, I'm not the decision making body of the team, they made the decision, I'm the advisor.
Here's what they came up with to generate 0.2 Tesla with iron core in the middle of around 12 inches long with #25 AWG copper wire wound with 40 turns, they bought a 3000 dollar power supply so that the power supply would generate 20 amps at 25 volts. To me that means that this group wants to generate 0.2 Tesla with a power input of 20 amps times 25 volts or 500 watts. I repeatedely mentioned to the group that they would need kilowatt power input to generate the 0.2 Tesla field they are looking for. The group has not revealed why they want 0.2 Tesla, I asked why can't you get away with milli Tesla, no response.
Can people in the Physics forum read my statement and see if I am way off or am I on track that 500 watts of power is not enough power due to hysteresis and iron core losses and copper losses, they already ordered the power supply, but at least I have comments from you that back my viewpoint or your comments back state that 500 watts is enough power. I'd like to know.
Thanks,
Bob

 

[trevie1970]

I thought I would post a link, to give some understanding into the powers of Tesla and may give some idea what type of power supply you will need to achieve your 0.2T. From what can see, it probably does not need the most expensive PSU other than something you can probably find around the home.

http://en.wikipedia.org/wiki/Orders_of_magnitude_(magnetic_field )

 

[rbobandcarly]

trevie1970,

According to what you are saying buying a 3000 dollar power supply was not a wise decision, where you can put some components for 50 dollars together and get the same results, that does make sense, what about the 40 turns of 25 AWG wire around an iron bar, can you achieve the magnetic field intensity of 0.2 tesla with a power supply that you can build at home.

Bob

 

[rbobandcarly]

The website you sent doesn't show how to build a power supply and a coiled iron core to generate 0.2 Tesla. Do you have an example?