I am trying to build an respiratory magnetometer. This device is no longer widely used even thought it was well liked. The reason it is no longer used is due to the fact that the coils used in it were custom made and the maker stopped making them. Over all with the advances in technology it seems to be much easier to produce. However I am still having problems recreating the coils. I have talked to several engineering students and contacted a few manufacturers that make custom coils and no one can help me (this may be due to my production size). There are 2 transmitter coils and 2 receiver coils. They are tuned to 5900 Hz and 4300 Hz.

The description in the patent (which has long since expired) is "Encased in a round plastic disk, each coil has its winding axis oriented parallel to the skin surface, and is roughly rectangular in shape, being about 3/4 inch (19mm) along the axis, and about 7/16 inch (11.1mm) by 3/8 inch (9.5mm) across the axis. The coils have inductance of 50 mH." There is a Q of 20. I can not figure out how they built the coils at such a small size with the 50 mH inductance, but this is not my area of expertise. Any help on how to build this or where to buy one would be greatly appreciated.

Another option could also be to use a hall effect sensor instead but many of the people I have spoken with did not think that this would be a good option at these distances. Any advice is welcomed.

Here is a link to the patent: http://www.google.com.ar/patents/US4258718

  • \$\begingroup\$ 50mH seems quite a lot for small dimensions but I guess you can achieve that with many turns, does the patent mention that? Odds are that inductance is not very important, as the other parameters. You probably want to understand the principle and build a similar apparatus with what you can find. \$\endgroup\$ Jan 17 '15 at 22:29
  • \$\begingroup\$ As a very quick guess, lots of turns of extremely fine magnet wire, which is probably quite challenging to work with. Likely it was done with a combination of specialized machinery and experience-honed operator skill. You might look at various inductor formulas and think about what kind of turns count and wire size would reach your goal. \$\endgroup\$ Jan 17 '15 at 22:53
  • \$\begingroup\$ I'm not surprised the manufacturer stopped making the coils. My rough calculations indicate about 6000 turns of #38 wire - in a very small volume. Wire length is about 500 feet and resistance is ~300 ohms. \$\endgroup\$ Jan 17 '15 at 23:09
  • 1
    \$\begingroup\$ Could the coils have been fewer turns on a relatively high permeability core, like a ferrite rod antenna? 300 ohms and Q=20 look unlikely together, and I calculate Xl=1350 at 4300Hz for a Q of something like 4. \$\endgroup\$ Jan 18 '15 at 10:43
  • \$\begingroup\$ @VladimirCravero The patent does not mention much else. I have been told that this is often the case that patents do not normally give you all of the information only just enough to receive the patent. They do mention the possibility of using a smaller coil (1mH) for monitoring infants. I am thinking that I am going to make a few changes to the overall design as this circuit was designed in the late 1970s and there are more options for many of the components and boards. Many parts can be replace with my computer so I think I may be able to get away with different coils. \$\endgroup\$
    – ann
    Jan 18 '15 at 23:24

The general principles are made clear enough in the patent. See "Summary of the invention".

This is an "easy enough" task to do by a number of means.
How well the method per se works is unknowable, but what the equipment needs to do is track the motion of 2 points on the rib cage relative to points on the patients back. Larger coils are liable to be no great problem.

Linear hall sensors are usable - field strengths gets small with increasing distance but some extremely capable active magnetic detectors can be obtained.

A near field inductive power transfer system would work and the ranges are within what is available commercially. Resonant transmit and receive coils and suitably high Q will give adequate range.

If this is of interest then more specific questions are welcome.

I've been looking at the specs of various linear high sensitivity hall sensors - and there are integrated modules that do vastly better at not vastly more. You can buy linear hall sensors with a 6 mG full scale range so lets say you can deal with fields down to 1 mG to start.

A modern rare earth magnet gives ABOUT 1/2 to 1 G at 1/2 the magnet length from the pole face. So if you had a say 10mm thick magnet you'd get 500 mG at 5mm (very very roughly). 1 mG is 1/500th of that. If field falls as d^2 you get to about 110mm and if field falls as d^3 you get only 40mm. At large distances field fall off is cube law but nearer its square law. (Ignore why for now). Realistically this means that you are near the edge of what is easily done off the shelf.

I measure ABOUT 250mm from back to chest front so I'd expect somewhere under 1 mG at my back with the above magnet. Larger magnets are one answer BUT a pulsed coild may do better AND you can use GMR effect sensors (don't even ask) that measure magnetic angle and these may work here and you could probably make a simple front reference harness that would greatly reduce distance

There are MANY possible solutions and I have little doubt that a few days of playing my a competent engineer would produce something usable.

HOWEVER THE big problem here is that we don't know enough and there is a great danger of dying the death of 1000 cuts / answerlets / questionlets here as we slowly edge towards understanding what you really want. If you are willing and able to provide a decent overview of what you are trying to do and why it would help heaps.

If you are trying to produce a commercial solution and do not want to say too much then you could look for competent people who seem liable to be able to help and take this offline. If you just want an answer and do not care who hears the details then this is an excellent place to ask.

There are various other methods that would allow you to track a say 3mm dia sphere in a space several meters on a side with mm's of resolution - but you don't need that. Lower res versions of that would do. Cost effort budget aim market ... need to be known. [Longish ago researchers managed to not only track a flying blow fly BUT were able to see which way it turned its head.! What you want is trivial compared.].

  • \$\begingroup\$ Where is the best place to start? Because so much of the original design can be replaced/improved with new technology I think the best place to start is with the available coils and their properties. I had looked into some coils used for wireless charging but I did not get very far as it seemed that they would not work at these distances. Most of the information I have found is for systems that either cover a smaller distance or a lot larger distance (with coils that are too large). Are there certain coils/systems currently available that you recommend looking at first? \$\endgroup\$
    – ann
    Jan 19 '15 at 21:33

The papers I can track down using the method all seem to be from right before the time the Polhemus company came out with the "Flock of Birds" magnetometer 6-degree-of-freedom position sensors. It wouldn't surprise me if this is the stuff that resulted in the formation of the involved company. It looks like the product is still available, though the product seems to have changed hands a few times over the years: http://www.5dt.com/products/pfob.html

I've used the Flock of Birds. The tech is fairly bulletproof, though linearization can be a bit tricky if your magnetic field is distorted (it involves mapping out the whole field). Sensitivity is pretty good, but the sample rate is a hair low for many things. It's thousands of dollars, but not insanely expensive for what it does, and I've often thought it would be suitable for many of the "how can I measure where something is"-type questions we get here.

So, I haven't directly answered your question, but I have pointed you to a commercial products that uses a similar, if not identical, technology, that can perhaps give you a boost in your research.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.