PCB coil to replace hand wound coil.

[rickford]

I have an application in which I am transferring power from a stationary device to a device that is rotating. The device that is stationary outputs an alternating magnetic field at 26kHz (sine wave) using a U shaped ferrite. The device that is rotating has a large, 7" diameter, coil wound in air with 24 turns of 26 gage wire. The device that is being powered by this coil has capacitors across the input such that the coil and capacitors form a resonant circuit. This aids in the power coupling across the gap. Power transfer works quite well. I have since designed a PCB with 4 layers of 6 turns each, one ounce copper, with traces .01" wide with .01" spacing between coils on the same layer. The board house tells me that the inter layer spacing is .04". The coil windings on the PCB are all in series, so it is also 24 turns. In the same way, capacitance is added to the device under power to bring the system to resonance. Problem is, this PCB coil does not pick up power nearly as efficiently as the hand wound coil does. It also doesn't seem to resonate nearly as well. I need to make changes to the PCB to improve this but changes are very expensive, so I need to understand what's going wrong. First off, the DC resistance of the hand wound coil is about 3 ohms, where as the DC resistance of the PCB coil is about 35 ohms. I tested the hand wound coil with a 30 ohm resistor in series with the coil, but still, ample performance was found. The resistor created loss of power transfer, but not nearly enough to cause failure. There are 2 other issues that I'm thinking of, but I can find very little about either on. First is skin effect. Second is the air gaps between the coil windings in the PCB are greater than those of the hand wound coil. I'm wondering if there is anyone that has a better understanding in this area that could shed light on what I should do to improve the coil design. Thanks.

 

[waross]

Any thing that you can do to reduce air gaps will help proportionately. When calculating amp turns per inch, one inch of air is equal to over 10,000 inches of good transformer grade iron. Amp turns per inch are important and when the air gap is relatively large, calculations may be based on the length of the air gap alone, neglecting the path through iron, without serious error.
The spacing between the coils may be reducing the self inductance of the windings. This will affect your inductance and your tuning. Try different values of capacitors.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[rickford]

Doesn't pass thru the copper - around it.

I have a hard time getting my head around the minute details of this. If the field doesn't pass through the copper, then how does the copper know it's there? Magnetics was not my strongest class back in school. It's as if the wire has an invisible lever arm sticking out and the magnetic flux passing by pushes on this lever. I wonder if I think about it this way, if it would help. The wire is like a screw with a nut on it. The nut has several paddles on it, and as this "magic" field passes by, it strikes the paddles, rotating the nut, and propelling the screw in one direction or another. Of course, the direction of the screw movement would be the current. If the screws are too far apart, the spinning nuts don't work in concert with one another and much of the field cuts through the middle, producing an opposite spin on some of the nuts, or, at least subtracting from the force of rotation.

As for the layout, the wraps are all in the same direction. The only issue, which I have now discounted, is that the board house could have reversed a layer. After looking at the layout again, it wouldn't have any continuity at all if that happened.

 

[itsmoked]

I have a hard time getting my head around the minute details of this. If the field doesn't pass through the copper, then how does the copper know it's there?
[red]It passes around it and verrrrrrry close to it, thereby affecting the free electrons in it.[/red]

Magnetics was not my strongest class back in school. It's as if the wire has an invisible lever arm sticking out and the magnetic flux passing by pushes on this lever.
[red]Naw. I find it way easier to think about the actual effects on individual electrons. An electron in free space would react the same way (vector right-hand rule). In a wire it's constrained like it's in a pipe.[/red]

I wonder if I think about it this way, if it would help. The wire is like a screw with a nut on it. The nut has several paddles on it, and as this "magic" field passes by, it strikes the paddles, rotating the nut, and propelling the screw in one direction or another. Of course, the direction of the screw movement would be the current. If the screws are too far apart, the spinning nuts don't work in concert with one another and much of the field cuts through the middle, producing an opposite spin on some of the nuts, or, at least subtracting from the force of rotation.
[red]Well I guess you could think about it that way. I prefer to think of it more like if the wires are bundled tightly the magnetic field that has to divert around each wire would instead divert around the entire bundle and therefore affect the eletrons equally in all of the wires. Whereas gaps will cut out various wires from being included in the mass effect.[/red]

As for the layout, the wraps are all in the same direction. The only issue, which I have now discounted, is that the board house could have reversed a layer. After looking at the layout again, it wouldn't have any continuity at all if that happened.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ijl]

"the wraps are all in the same direction"
Does this mean that the current circulates in the same direction in each layer?
It should, but the wording makes me wonder.

Have you measured the inductance of the two coils?

Was the 30 Ohm resistor really part of the resonance circuit in your test? That is, one end of the capacitor was connected to the inductor and the other end to the resistor?

 

[rickford]

"the wraps are all in the same direction"
Does this mean that the current circulates in the same direction in each layer?
It should, but the wording makes me wonder.

Yes, the current should be running in the same direction, as the loops all go around the circle in the same direction.

Have you measured the inductance of the two coils?

I don't have an inductance meter, but I did put each into a tank circuit and using a function generator and a scope, I measured the resonance frequency and solved for the inductance. They are not the same. The PCB circuit inductance is much lower.

Was the 30 Ohm resistor really part of the resonance circuit in your test? That is, one end of the capacitor was connected to the inductor and the other end to the resistor?

Yes, the resistor absolutely was in the circuit. Note that it also changed the results.

itsmoked, I like your explanation of the magnetics. It helps.

It sounds like my windings may have too much gap between them. Sounds like I need to tighten up the spacing, not only of the gaps between the traces on any given layer, but I need to specify a narrower gap between the layers.

 

[waross]

The current in each wire develops a magnetic field which affects the surrounding wires. This is mutual induction.
I hope that the outside of each layer is connected to the inside of the next layer. If identical layers are connected inside to inside and outside to outside then the coils will tend to cancel each other.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[VEBill]

Why not reproduce the PCB coil geometry using the wire and see if the problem can be reproduced? This would confirm or refute the geometry theories in about an hour.

 

[ijl]

If the inductance is much lower, then the Q-value of the circuit is much lower, and you get a (much?) lower voltage. Simply increase the number of turns, so that the inductance is as large as or larger than the inductance of the other coil. If possible, use thicker copper and wider traces in order to get the resistance down. And to not worry too much about the gaps!

Please forgive me, but I still do not understand that "the loops go in the same direction". As I understand, there is a planar spiral in each layer, and the spirals are connected vertically. Correct? If you connect them in the shortest way, the spirals should go "inwards" and "outwards" in aternate layers.

 

[rickford]

I hope that the outside of each layer is connected to the inside of the next layer. If identical layers are connected inside to inside and outside to outside then the coils will tend to cancel each other.

If this is true, then you may have hit on the problem... though I have trouble seeing why it would be true. The layout is this... looking down on the board, a imagine a connection point at 12:00 position. The connection point is 2 terminals. From the right connection, a trace travels upward in the 12:00 direction about 3/4" to meet the inside of the coil. It then makes a right turn and goes clockwise around the coil, with each wrap getting slightly larger than the last. After 6 turns, the trace enters a via where it drops to the next layer. In the next layer, it continues to travel clockwise starting at the outside, with each wrap getting slightly smaller than the last. After 6 turns, the trace enters another via to go to the next layer, and the coil winds back out again. On the next layer, it winds back in again until finally meeting where the coil originated, then turning down and into the left terminal of the connection.

How does returning to the inside of the coil at the start of each layer improve the issue? Isn't the current always traveling in the same direction around all the loops? When I wind a coil by hand, the windings go across instead of in and out radially, but still, they go across one way, then back the other, then continue back and forth as the coil grows. What's the difference?

 

[IRstuff]

A simple screen shot of your layout would clear up this question immediately.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[rickford]

This is about the best I can do with the program I have. See attached pdf file.

 

[rickford]

Well, I guess that didn't work. Looking for another way...

 

[rickford]

Try this...

 

[rickford]

Note on the pdf that the 4 inner layers are used for the coil. There is no copper on the top and bottom layers except for the via's and the connection pads. This is done this way to protect the coil in case it momentarily rubs while spinning during setup.

 

[MikeHalloran]

It looks like you have room to make the traces _much_ wider.
That would reduce the amount of copper you have to remove from the board, reducing your waste stream volume, reduce the coil resistance, and possibly make it less sensitive to alignment errors or runout.



Mike Halloran
Pembroke Pines, FL, USA

 

[rickford]

I have done some more experimenting with resonance and a function generator. This is what I have so far...

The PCB coil:
Tuned resonant frequency with .15uF cap in parallel, 31kHz. Calculated inductance based on resonance and known capacitor in parallel, 176uH. Q of tank circuit, about .55.

Hand wound coil with 26 gage magnet wire:
Tuned resonant frequency with .1uF cap in parallel, 33.5kHz. Calculated inductance, 226uH. Q of tank circuit, about .9.

Hand wound coil with 26 gage wire with teflon insulation (about .042" OD) producing about .027" spacing between copper windings:
Tuned resonant frequency with .15uF cap in parallel, 31kHz. Calculated inductance, 176uH. Q of tank circuit, about 1.1.

The caps are the same type, and so their Q should be similar. I believe this would mean that the difference in Q of the entire circuit is largely due to the difference in Q of the coil... and since Q is a measure of losses in the coil... of which resistance is a major factor in the PCB... it all seems to point back to resistance. There's only so much I can do to make adjustments on the PCB, so I'm leaning toward just making the traces as wide as possible, making the traces as thick as is reasonable, and the gaps as narrow as is reasonable and being done with it.

 

[rickford]

reducing your waste stream volume

What is waste stream volume?

 

[MikeHalloran]

There are other ways, but most PC boards are made by a subtractive process.
You buy a sheet of FR4 clad entirely with copper.
You use a photographic process to mask the copper that will remain.
You use an etchant to dissolve the remainder of the copper.
The etchant plus copper becomes a toxic waste, which you must pay to have recycled or discarded, and you must keep detailed records tracking the stuff.
You don't have to pay for disposal/recovery of copper left on the board.
So the board _may_ get cheaper, and you get environmental creds, by maximizing the copper area on the finished board.



Mike Halloran
Pembroke Pines, FL, USA

 

[rickford]

Ok, that makes sense.

Well, thank you all that contributed. I'm about ready to commit to a layout so I guess here goes...