A brief introduction to the past and current history of basic levitation setups from contributors

By Meredith Lamb    12/22/2002

The history of the use of graphite in tiltmeters/seismometers goes back in time to the decades of the 1960's and 1970's, for the most part,  with the A.D. Little Company with their variety of instruments.  At the time they used high purity graphite rods with alnico magnets.  Their has been a number of U.S. patents issued since then, which all seem to use the various material, designs and dampening and/or feedback components of the times.  Their is no known diamagnetic seismometers with patent applications pending.  All such patents are interesting and often still somewhat relevant to current efforts on a non-commercial and amateur level of development.  Currently, the main effort is for a fairly simple seismometer with a minimum number of parts; however it turns out that there is quite a variety of amateur designs around that can work within their design layout limitations.  Present day amateur diamagnetic graphite mass designs are very light weight and size limited; they are vary from a portion of a gram, up to ~ 3 grams, and are also very frail and delicate.

With the advent of certain specific shapes of stronger neodymium magnets in the last decade, that critical component alone, has really been a main part of being able to make diamagnetic graphite levitation devices.  Spectrographic graphite rods (high purity) has been around for quite some time, and still offer potential as part of the mass in a seismometer.  In more recent years, a new type of graphite has come on the scene called pyrolytic graphite (PG).  Unlike spectrographic rods, the PG has too be orientated to it sensitive plane to get results, but when its been done so, the results have been quite abit better than spectrographic rods, in its lift capacity ability, for added dampening material and/or flags for optical readouts.  Their is a varietys of PG that isn't of any value, like the HOPG (Chemical Vapor Deposition) variety (mostly glass like, and not diamagnetic).  Other useful PG varietys include "substrate nucleated", or "continously nucleated" PG.  The specific variety of pyrolytic graphite mentioned pertains to the gas diffusion deposition variety (natural gas or methane), where heat and layers of graphite are deposited over a long period of time onto a surface, by the manufacturer.  Usually this graphite is outragiously expensive when bought via the manufacturer.  Now, with at least one web retailer selling such; the price per piece has significantly dropped into the realm of affordibility.   See; this pyrolytic graphite material works very well.

Perhaps the main interest in diamagnetic approaches, is due to the fact that the small suspended mass is usually without any source of friction (except for air); which greatly enhances its potential use.  Another plus, is usually they can be quite small in size.  Useful graphite when levitated can also be contained (stabilized) within a magnetic influence design.  For the moment, most of the designs are relatively short period in their normal undampened oscillation cycle (period).  However, like all normal horizontal seismometers, they are also quite susceptible to any displacement influence as occurs with any quakes with the "p" and "s" phases.  They can, but don't always register the "L" wave of very large quakes.  I do think their is a limit of quake epicenter distance and magnitude sensitivity with these units, which could perhaps be related to their very small mass and lack of substantial enerta, like that found in a larger regular seismometer mass.  They can be quite sensitive to mechanical tilt influences in their respective environments; as are most other horizontal seismometers.   They may in the end be only appropriate for certain minimum magnitude quakes within a specific mile/kilometer radius, in a normal amateur noisy setting; but its unknown how they could perform in a quieter location.

As with all subjects of interest, it should be emphasized that there has been alot of good people participating or contributing over time with this diamagnetic graphite levitation seismometer/s aspect in a variety of ways.  To name afew; Robert E. Lamb (deceased), David V. Lamb, Martin D. Simon, Dr. John Lahr, Chris Chapman, James Spottiswoode and others; so the honor of their contributions should be known and denoted in the amateur diamagnetic seismometer help, design and applications realm.

To tell it via my own personal tale, I'd have to go back to a local group of people that were invited to a local PSN (Public Seismic Network) type meeting in Golden, Colorado; that was sponsered by John Lahr, around September of 1999.  John had a variety of diamagnetic repulsion demonstration items.  One of which, was a balanced top that when accelerated in rotation and lifted above a magnetic base, the top could be made to levitate in space.  For myself, the sight of the top spinning in space some 6-8" in air by itself was fascinating!  It also immediately conjured up the thought of using levitation as a approach with seismology.  Of course, the spinning top by itself, wasn't ideal for a "quiet" seismometer application.  Just the demonstration and its curiosity provoking thought alone sparked further pursuit.

At that time, the only very limited diamagnetic levitation devices around were those mostly constructed by afew individuals, and they generally used impure "shotgun" bismuth metal, that levitated a magnet.   Eventually over time, I was able to obtain purer bismuth pellets for experiments.  In the end though, it wasn't the bismuth, but certain varietys of graphite that worked better for levitating magnets.  Of course alot of common graphite like in motor brushes was not of any value with it being varying degrees of useless paramagnetic graphite; but some dubious diamagnetic graphite worked to a degree.  Eventually I obtained some surplus spectrographic rods from a surplus store, that far surpassed any previous bismuth or other
impure graphite for diamagnetic levitation of magnets.  Jumping briefly ahead, spectrographic graphite (and some pencil leads) are quite pure, and remain to this day as a good material for levitation.

I eventually made a "S-G" (hanging pendulum in escense) type diamagnetic levitation instrument, where a magnet was levitated by enclosing it mostly in diamagnetic graphite.  I used a Hall type sensor with it.  The trouble with levitating a magnet is that it is sensitive to all magnetic influences like passing cars, buses, and even a occasional geo-magnetic storm produced by a solar sun flare.  It seemed to be a short range seismic instrument occasionally; but all the other "noise" it picked up wasn't desireable.  In one instance, a nearby running refrigerator motor induced visual oscillations with the levitated magnet.  Unless they are enclosed, they also have a tendency to point in the earths North-South directions, exactly like a compass.

There was also at about the same time, a newer variety of graphite that appeared in the form of a diamagnetic levitation stand, that was made at UCLA, and designed by Martin D. Simon.  The device used a new form of diamagnetic graphite called "pyrolytic graphite" (PG), which is a gas diffusion form of deposited graphite.  This PG was the best of any graphite I've seen for levitating magnets.  Some of the discs went into the above paragraphs, "magnetometer/seismometer/tiltmeter".  Unfortunately, this UCLA diamagnetic stand is no longer being made presently for the costs was a very good design for illustrating diamagnetics.  Along with email communication, Martin forwarded some other leftover pieces of PG for experiments, which also came in useful for a variety of later experiments and actual use.  See Martin Simon's web site for a variety of text, photos, and more technical presentations on diamagnetic levitation:

At the time, the idea of flip-flopping the arrangement; where the graphite floated and the magnets were stationary was of high interest, but the means were not apparent to do this.  There was no web reference to doing this outside of earlier patented designs with their variety of complicated mechanical arrangements using alnico magnets. early January 2000, Robert, David and I set about to come up with way to use neodymium magnets, with the UCLA PG discs to levitate graphite on a simpler scale than previously done or known.

The breakthrough came on January 17th 2000, when it was found that 4 rectangular neodymium magnets with a high gauss rating, when connected in a square magnetic attracting outline layout  would do the trick, with the UCLA PG discs.  It was a exhilarating moment at last!  Only later on, with further U.S. Patent examinations, did I fully realize it wasn't exactly new, the SRI International, company illustrated the magnet pattern in one of there U.S. patents.  In
retrospect, its the spaces inbetween the magnets that provides the greatest levitation lift; just like all the rest of the patents exhibited, but didn't really verbally present to my understanding at the time.   So, there was nothing really new in this, but now, we could proceed further with obtaining a variety of neodymium magnets, spectrographic rods and PG for other experimental trials.

 First levitation

The next quite noteworthy design was made by David Lamb around February of 2000.  His design was (and still is), the most simplest levitation design I've ever seen or known.  It consisted of a ring magnet (rated at 35) and a cylinder rod magnet (rated ~40).  The cylinder magnet poles ends were inverted into the larger ring magnet hole.  The UCLA PG disc simply and readily levitated atop the surface of the magnets!  One didn't need any complicated support frame, but the levitation did improve when it was laid flat atop some steel/iron base.  David added some silver plate atop the PG, and used a laser and photodiode on the edge of the discs to get a signal.  It was the first known amateur diamagnetic seismometer!!    There was/is problems with this approach, as the disc tended to freely rotate, and being as the setup acted like a inverted pendulum, and with weight distribution (tilt effects) of dampening material, it did tend to create problems.  The PG in itself also exhibits a tendency to present magnetic node areas with the total number of magnets prefers rather random areas to wander too, dependent on environmental influences.  The instrument was a rather short period instrument; perhaps 1-3 seconds if I recall correctly.  One very obvious positive aspect of this instrument was the tenacity of the PG to remain over the magnets, even with moderate induced shaking.  I've liked the basic design myself; but perhaps there is a way around the problems yet....

Davids ring and cylinder magnet levitation

Robert Lamb came up with the idea of cleaving the PG into pieces of various thickness's, in a vice jig in early 2000.  The idea, and actual use of the various disc thickness pieces thereafter were very important as it demonstrated both the problem of the PG thickness (weight influence), and the minimum efficient diamagnetic thickness of the PG, to carry a load of dampening aluminum, copper, or silver plate.  The smaller the thickness of the PG, the easier it levitated; and even levitation with lower grade neodymium magnets like the grade 35.  Generally, there was no tested thickness's, but, one might say, that a thickness of 0.050" and below is "adequate", while a thicker piece of PG, then tends to be less effective, as its carrying "unused" PG.  PG can only react diamagnetically at a certain distance from a magnet; any more on the back side of the PG, then becomes a unused PG weight burden.  One needs all the weight they can get for dampening and/or a optical flag sensor.   With the advent of cheaper PG, one can now simply sand the material to various thicknesses.  A thin disc or PG shape is usually more desireable, as long as shape is flat on both sides, relative to the magnet/s used.  Robert had alot of ideas, and one of the more unusual ones was immersing some of the diamagnetic setup material into a fluid for dampening purposes.  Bob also suggested using a hanging magnet on a string, to test the diamagnetic repulsion "strength" of various materials; which was often done in the earlier days with a large variety of material.

 Immersion in a fluid

The next significant instrument design came with the use of cylindrical magnets that were made by Wondermagnet/Forcefield.  These were 1" long x 1/4" diameter, and were polarized through the diameter (not on the ends).  It was found that one could lay out pairs on steel where they are attracting side by side, while putting on other lengthwise additional pairs could lenthen the area for a spectrographic rod to levitate and stabilize.  The magnets were a grade 45, about the maximum field thats (very rarely) made by retail outlets.  One could adjust the length of the
graphite rods, and shape it into various lengths (triangle, half moon, quarter moon etc.), with various results.  Here, the magnets pairs either attract or repell each other, which is helpful with
dampening.  The end period of these generally doesn't get much above a maximum of 6 seconds, but they can also sense the horizontal displacement phases of bigger quake "L" phases (with a longer period than normal for the seismometer) occasionally.  As I understand it, one must write a email to Forcefield for current availability and price.  John Lahr made a replica of the design and suggested that spacers inbetween the pairs could help enhance the diamagnetics; he was right; it does.  With a appropriate sized spacer of ~ 0.022" thickness, the gauss field and diamagnetics seems to peak as measured with my gaussmeter.   James Spottiswoode's version of this basic design also brought up the previously unknown fact that another source of graphite could also be used for levitation, and that is that certain brands of relatively common refill pencil leads, are apparently nearly as pure as spectrographic rods, and alot cheaper!  It was a big oversight for many; but a very significant contribution!  On John Lahr's web site below, there is a variety of diamagnetic related designs, ideas, and interesting notes and graphics, along with a current list of pencil lead results.  Included are some notes and drawings from Chris Chapman on various design setups.  James web site is also a "must see", and denote all the photos, grams, and the basic circuit he uses.

 The Gold Rod setup

From this period on, the designs and changes created by David Lamb became, many...literally!  He came up with such a variety since 2000, that it often became hard to follow all of them.  Most of them did involve more complicated mechanical arrangement/s, and stronger than normally found neodymium magnets; with the use of spectrographic graphite rod material.  David has been recording results of his various seismometer designs since early 2000.

 David's U setups

Currently, David has designed and is operating acouple horizontal diamagnetic seismometers of a more effective and truer response nature.   A major part of this design use, is that the single length magnet avoids all the mulitple magnetic fields that can interact adversely like in the "gold rod design".  David calls them, the "H" design; as the word and the shape of the magnets, and the carriage he uses, does overall reflect a type of H form.  Bascially, a "H", is a two channel graphite levitation arrangement with a cross bar connection.  Here, the diamagnetic graphite is the PG with the gas diffusion manufacture.  It is capable of lifting additional weight of some ~ 5 times the weight of the graphite in ratio; which is quite a departure from using spectrographic/pencil leads, which are very limited in the weight they can carry.  This design is the first utilizing this type of graphite that I know of.  This graphite needs to be shaped for its specific magnet gap/s; which can be somewhat time consuming, but quite possible to make.  Most significantly here, David is probably using ~ grade 35 neodymium magnets, and, they seem to be consistently for sale (available) from Alltronics.  Having a magnet readily available seems to be a consistent problem with a number of past diamagnetics designs presented; of which their isn't any ready solution.  David currently operates two units as of December 2002; and improvements have been made occasionally.  Davids current "H" magnets are available via; and is listed on their web "magnetics" product page as "2000 gauss super magnet" for $5.95 each; which for the size is quite reasonable.  They are big enough to cause damage to fingers; I'd suggest one wear heavy leather gloves when handling or arranging them.
Actually, the magnet/s would also likely serve well in a conventional coil/magnet and/or eddy
current dampening seismometer, just from their physical size.

 David Lamb's H seismometer

Concurrently with Davids "H" design, was the very significant contribution by Chris Chapman in relation to the suggested added amound of sheet iron atop the magnets, which do effectively help guide the gauss fields to the magnet spacing/gap (iron spacer) separation areas.  It does help alot!  It seems that 0.0325" thick sheet iron is about ideal, but its unknown exactly as to what thickness is the best.  I've tried 0.1" thick iron and it seems to degrade the levitation more than it helps.

Also concurrent with David's "H" design, has been the idea of using the same PG, but with strips of such orientated both horizontally (for levitation), and vertically (for H frame containment within the magnet gap areas.  This specific PG is only effective in one plane.  The idea has been noted from David and John over time.  The idea does work, per my current incomplete "H" version, very well.  As for myself, I am still working on the subject of damping the "H"; none of the previous used or suggested approaches seem to work as well as I'd like to see.

David's "H" setup in reality probably represents the best in a basic setup, for its ingenious simplicity.  The "specific, selected" magnets are simply laid out on a very flat piece of iron/steel with the same magnet poles upward (this can be varied with a "string of magnets" lengthwise).  Inbetween the magnets are placed a specific size of iron spacer (cut steel/iron, keystock works well for some versions).  It is this upper channel or air gap, within which diamagnetic levitation takes place.  The narrower the gap, the greater the diamagnetic repulsion effect with graphite.   I "think" that the iron/steel spacer height, works best where it covers a complete pole thickness, but not the opposite magnet pole....but this can be varied, and one can still get fairly good results where it extends into the opposite poles area somewhat.

Perhaps in the truest sense of the words; all diamagnetic seismometers are perhaps still a "work in progress", but, nevertheless, quite interesting for their potential.   This web page does not try to illustrate the finer details of various diamagnetic setups; it is meant only to introduce the overall subject in general.  It is quite fun stuff, in the sense that there are still alot of unknowns to pursue.

Take care all,   Meredith Lamb

 Back to Index page