To do magnetic resonance imaging, you need a static magnetic field that is uniform and a switchable magnetic field that has a uniform gradient. How do you produce such fields? In this post, I explain one of the simplest ways: using a Helmholtz coil and a Maxwell coil.

Both of these are created using circular coils. The magnetic field produced by a circular coil can be calculated using the law of Biot and Savart (see Chapter 8 of *Intermediate Physics for Medicine and Biology*)

where μ 0 is the permeability of free space (the basic constant of magnetostatics), *I* is the coil current, *N* is the number of turns, *R* is the coil radius, and *z* is the distance along the axis from the coil center.

## The Helmholtz Coil

The Helmholtz coil consists of two circular coils in parallel planes, having the same axis and the same current in the same direction, that are separated by a distance *d*. Our goal will be to find the value of *d* that gives the most uniform magnetic field. By superposition, the magnetic field is

To create a uniform magnetic field, we will perform a Taylor expansion of the magnetic field about the origin ( *z* = 0). We will need derivatives of the magnetic field. The first derivative is

(The reader will have to fill in the missing steps when calculating these derivatives.) At *z* = 0, this derivative will go to zero. In fact, because the magnetic field is even about the *z* axis, all odd derivatives will be zero, regardless of the value of *d*.

The second derivative is

At *z* = 0, the two terms in the brackets are the same. Our goal is to have this term be zero, implying that the second order term in the Taylor series vanishes. This will happen if

or, in other words, *d* = *R*. This famous result says that for a Helmholtz pair the coil separation should equal the coil radius.

A Helmholtz coil produces a remarkably uniform field near the origin. However, it is not uniform enough for use in most magnetic resonance imaging machines, which typically have a more complex set of coils to create an even more homogeneous field. If you need a larger region that is homogeneous, you could always just use a larger Helmholtz coil, but then you would need more current to achieve the desired magnetic field at the center. A Helmholtz pair isn’t bad if you want to use only two reasonably sized coils.

## The Maxwell Coil

The Helmholtz coil produces a uniform magnetic field, whereas the Maxwell coil produces a uniform magnetic field gradient. It consists of two circular coils, in parallel planes having the same axis, that are separated by a distance *d*, but which have current in the *opposite* directions. Again, our goal will be to find the value of *d* that gives the most uniform magnetic field gradient. The magnetic field is

The only difference between this case and that for the Helmholtz coil is the change in sign of the second term in the bracket. If *z* = 0, the magnetic field is zero. Moreover, the magnetic field is an odd function of *z*, so all even derivatives also vanish. The first derivative is

This expression gives us the magnitude of the gradient at the origin, but it doesn’t help us create a more uniform gradient. The second derivative is

This derivative is zero at the origin, regardless of the value of *d*. So, we have to look at the third derivative.

At *z* = 0, this will vanish if

or, in other words, *d* = √3 *R = *1.73* R*. Thus, the two coils have a greater separation for a Maxwell coil than for a Helmholtz coil. The Maxwell coil would be useful for producing the slice selection gradient during MRI (for more about the need for gradient fields in MRI, see Chapter 18 of ).

## Conclusion

Below is a plot of the normalized magnetic field as a function of *z* for the Helmholtz coil (blue) and the Maxwell coil (yellow). As you can see, the region with a uniform field or gradient is small. It depends on what level of accuracy you need, but if you are more than half a radius from the origin you will see significant deviations from homogeneity.

Russ Hobbie and I never discuss the Helmholtz coil in *Intermediate Physics for Medicine and Biology*. We don’t mention the Maxwell coil by name, but Problem 33 of Chapter 18 analyzed a Maxwell pair even if we don’t call it that.

The Maxwell coil is great for producing the magnetic field gradient needed for slice selection in MRI, but how do you produce the gradients and needed during MRI readout and phase encoding? That, my friends, is a story for another post.

*Originally published at **http://hobbieroth.blogspot.com**.*