Honeywell magnetic sensors

[J_Player]

The magnetometer project section has links to dated designs referencing the FGM-3 sensor, which is no longer available [Admin note: FGM sensors are still available through Speake, see post below]. However, a new sensor technology has appeared which surpasses the performance of the FGM-3 sensors. Honeywell has released the HMC series of magnetic sensors which use miniature on-chip magnetic film sensors with the same resolution as the FGM-3 sensor, but use only 6 ma for 2.7nT resolution, without the bulk or long cycling times needed by other types of sensors. These new sensors can be used in the same basic designs as the current projects, but with simpler circuitry and fewer components. The HMC1001 and HMC1002 seem to have the best resolution in the group, and cost less than $25.


Here is an overview of magnetic sensor products from Honeywell:
http://www.globalspec.com/FeaturedPr...ensors/25860/0

Here are some honeywell schematics for a simple gradiometer circuit:
http://www.magneticsensors.com/appli...adiometer.html
http://www.honeywell.com/sites/servl...F-0D4D3586AC87

Application notes, tech article and software downloads for these products here:
http://www.honeywell.com/sites/porta...ature&theme=T5

You can buy a 1-axis, 2-axis or 3-axis sensor ICs starting at about $15 here:
http://www.newark.com/jsp/level5/mod...n_US/70806.xml
http://shop.ssec.honeywell.com/shopd...nd+Accessories

PNI also sells small boards with magnetic sensors with a resolution of 15nT for about $45.
http://www.pnicorp.com/family?nodeId=c1c
These PNI sensors use a different technology with inductors that sense the field.

[BillA]

just for perspective:

a G-856 has an accuracy of 0.5nT with 0.1nT resolution

a G-858 has a sensitivity of 0.05nT @ 0.1 sec cycle rate
and 0.01nT @ 1.0 sec cycle rate

for a loop in the pavement to detect a car to open a gate, 15nT is ok
never gonna see anything faint, sorry

[J_Player]

Perspective

You are correct. The G858 is a very sensitive instrument often used for making grid surveys over a large area, and costs about $5000. The Honeywell sensors are not intended to compete with these high-end gradiometers. The best of the Honeywell sensors is rated at 2.7nT, which is slightly better than the un-available sensors specified on the fluxgate magnetometer project, and costs maybe half as much to build a similar magnetometer using the Honeywell sensors. http://geotech.thunting.com/cgi-bin/...fmx1/index.dat

This is not a project to build the highest performance magnetometer, but to build a cheap hand-held device that an average detectorist could toss in the back of the car along with his metal detectors when he goes relic hunting. I have not built one of these magnetometers, but the attraction for me is that it can be built to a size and weight smaller than an average metal detector, and can be used to locate iron objects buried in relatively shallow soil. While the resolution is around 1/100 of a G858, the ability to locate shallow buried ferrous objects is much better than 1/100. It could be particularly useful in locating things like tanks, vehicles, cannon balls, old weapons, tools, and possibly steel containers if not buried too deep. And these magnetometers seem extra useful because they can run on a 9-volt battery for a long time and weigh less than the shaft of an average metal detector, all for less than $100.

Now, if I wanted to survey a large area for deeper buried objects and soil anomalies, I would definitely prefer a G858. And I would also keep in mind that I would not be carrying the G858 in the other hand to use later while I was hunting with my metal detector.

[BillA]

a minor point:
the G-856 is ~$5k
the G-858 is ~$15k (or 20 ?); the G-858 is SERIOUS overkill, but an instrument of choice as a gradiometer in archaeological surveying (0.25m reading stations with DGPS) - add $5k

I would pose a question: Has anyone actually found say a cannon ball with one of these kits ?
(In years of browsing the forums I do not recall reading such ??)

[J_Player]

Not so minor...

You are correct again. I had my model numbers confused. The cost is as you stated, and more when you add on the accessories. From what I read the G858 is an excellent tool for locating geological and archelogical structures under the ground as well as metal targets. The G856 is gives less resolution, but is more affordable. Personally, I would spend my money on a collection of metal detectors before I spent $5000 - $7000 for a G856 magnetometer/gradiometer.

But will these cheap Honeywell sensors detect a cannonball? In fact I don't know. These sensors are fairly new in the field of magnetometry, and no detectorists have used them that I know of. These sensors are used mostly in compass applications for vehicles and in instruments. But a quick google shows that there is a lot of promise for them:

1. A test was conducted by Kyushu University 6 months ago using the HMC2003 3-axis sensor (4nT resolution) to verify their theories in mine detecting. They built a dual sensor in the same configuration as Carl's fluxgate magnetometer project. They scanned a grid at a height 1/2 meter above a 2.4kg piece of iron that weighed 2.4kg (smaller than a hockey puck), and charted the results in all three axes. As you can see in the chart, any of the 3 axes will tell you where the mine is.

Chart: http://wl.ifastnet.com/HMC2003_test.jpg

I suspect that if a real cannonball was buried say, a foot under the ground, and a 3-foot rod holding the sensors was used, an average detectorist would stop his grid searching as soon as he heard the first anomaly, and would zero in to pinpoint the target. (See the full Kyushu University study at https://qir.kyushu-u.ac.jp/dspace/bi...3/1/66-3-1.pdf ).

2. Another study at the University of Virginia reports that a HMC1002 sensor detects slowly moving passenger vehicle at 8-10 ft. http://www.cs.virginia.edu/papers/tr...-mobisys04.pdf

3. The Honeywell sensors have also been used in a number of satellites in recent years: http://journals.tubitak.gov.tr/engin...1-5-0505-1.pdf

4. Quantum Magnetics, Inc tested several types of gradiometers including the Honeywell based sensors. They made the following conclusions about fluxgates and the Honeywell sensors:
● They are adequate for most applications. Such gradiometers can be used for numerous applications including vehicle tracking, detection of buried UXO and mines, detection of underground structures, and concealed weapons detection and tracking.
● They are easier to adapt to adapt to various applications.
● It is pointless to use expensive sensitive sensors.
http://www.dtic.mil/ndia/2002mines/kumar.pdf

Even after reading these reports, I could still be wrong. Maybe these sensors are useless as gradiometers for locating buried ferrous objects. But it only costs $100 to find out.

[BillA]

metal detectors are going up as well, along with their capabilities
for a single coil unit I just dropped $4k on a used Minelab GPX 4000 (LOTS of extras involved)

I am not sure a cheap mag would be better than a good 2 box/loop detector.

[Geophyz]

What exactly was the resolution of the FGM-3? I don't recall ever seeing that published (not that I looked all that hard). And when did they stop making them?

[99thpercentile]

There is a huge amount of potential for GMR sensors such as the the Honeywell sensors. I think the numbers quoted above are a little bit generous since the last time I spoke with anyone at the GMR companies, September 2006, the sensitivities were about 7nT. The cutoff for reasonable field use is around 1nT. I use a G-858 on a regular basis and use a G-856 as a base station. Right now several GMR manufacturers are getting large amounts of funding from DARPA to improve the sensivities of their sensors. The Navy in particular wants units we better than 1nT sensitivity. Wait two or three years and you should be able to get these sensors for $25 each.

The upside is that you can build array sensors to actually image targets instead of having one measurement every few centimeters in-line and every half meter between lines. There was a good paper at one of the UXO Forums by Alison Chaiken in 1997. I have read several PhD dissertations on the subject as well. These are one of the most exciting developments in geophysical sensor technology in a number of years.

[BillA]

99thpercentile
I am looking for a (second) G-858 sensor, any idea who might have a used one ?
(Geometrics does not, only new for sale.)

You may be interested to know that Geometrics is finally developing a Li-Poly battery pack for the G-858; I made a prototype almost 2 years ago.

Bill

[J_Player]

Some answers:

Geophyz: What exactly was the resolution of the FGM-3?
The FGM-3 is a fluxgate sensor manufactured by Speake Sensors. According to Stan's project page and the manufacturer's page, the sensor resolution is about 50nT: http://geotech.thunting.com/pages/ma.../fmx1/fmx1.pdf http://www.speakesensors.com/

page 7 of this PDF file shows where an experimenter sensed cars passing at up to 75 feet away using a FGM-3 sensor: http://www.directivesystems.com/PDF/MAGTALK.PDF

Geophyz: ... And when did they stop making them?
They did not stop making them. The FGM-3 and a couple of smaller variants are manufactured by the people at this link: http://www.speakesensors.com/ But the only supplier link in the USA says he is closed: http://www.fatquarterssoftware.com/ [Admin note: they are open again]

99thpercentile:
The resolution and sensitivity of the Honeywell sensors varies within the HMC group. The highest resolution is from the HMC1001 and HMC1002, which is published on the Honeywell data sheets here: http://www.ssec.honeywell.com/magnet...1-2&1021-2.pdf

The Honeywell data sheet states the resolution is 27 microgauss = 2.7nT. the sensitivity is typically 3.2mV/V/Gauss.
However, this is the resollution of the sensor. The circuitry which processes the sensor signal can degrade the resolution in order to gain a wider range or for other reasons. Some of the other sensors in this HMC family have lower resolutions, but contain on-chip signal processing. These sensors are built in a microscopic wheatstone bridge configuration, which inherently has room for improvement. The materials for the magnetic sensing film can be improved for better sensitivity as well as the on-chip configuration and manufacturing tolerances. I suppose with enough development work, these sensors could be made into complete magnetometers lacking only the power amplifiers to send the signals for the desired application.

The concept of imaging arrays is just around the corner. When these sensors are capable of better resolution, the next thought of the designers will probably be to construct arrays and write software to make magnetic survey maps of an area. I imagine that sometime in the future we will be able to see images of what is under the ground as well as see magnetic maps of a plot of land after having walked in a grid with one of these lightweight sensor arrays.

Bill A: I am not sure a cheap mag would be better than a good 2 box/loop detector.
I am not sure if it would be better either. But I still wouldn't mind having a gradiometer that weighs 1 pound for under $100.

Bill A: metal detectors are going up as well, along with their capabilities
for a single coil unit I just dropped $4k on a used Minelab GPX 4000 (LOTS of extras involved)

The minelabs GPX 4000 is considered one of The best nugget hunters in the world. But as we speak, less expensive competition threatens to make it obsolete. The "Pulse Devil" is about to be tested before final release. It is described as a PI detector that has all the depth capability of the GPX 4000, but without the "hole" where it misses certain size nuggets, and it also has the ability to discriminate as well as any VLF detector. From what I heard Dave Emery thinks the price of his Pulse devil will be substantially less than the Minelabs GPX 4000. If this is true, then it is an example of newer, better detectors becoming less costly (a rarity in today's world). A second PI detector the "Goldsweeper" is also scheduled to be demonstrated soon which can discriminate various metals. However, neither of these detectors is available for purchase yet.

[BillA]

J_Player
appreciate your info on the Honeywell sensors; I suspect that if the government has an interest (per 99thpercentile), the research is already ongoing
(this is quite related to IEDs)

re metal detectors (straying off-topic)
yes, am aware of the buzz (hype ?)
being 61 and having bought a number of Eric Foster's machines over the years, I would opine:
the new machines will cost less and be ~ as good in some situations and a bit worse in others
i.e. a lower cost alternative
with so many interactive variables a designer is forced to choose between an infinitely adjustable machine, and one that can be understood and used in the field
- can't kiss all the girls, at least at the same time

Bill

[Crowbar]

About FGM-3 resolution: it's near 10nT but not 50nT. About MR sensors: I made a fluxgate gradiometer(on FGM-3) and a MR (on HMC1021)gradiometer and want to say that MR grad is better due to simlple sensor align and as I think linearity. But such gradiometers goes only as a support instrumuments(I used them to determine a calm place for 3-axis mag)

[Unregistered]

Whats the resolution of the FGM-3h

I know it'll saurate north/south but just wondering?

any low cost 0.5nT 0.1nT sensors?

[efkern]

FGM-3 sensors ARE available

I don't know who posted the message that Speake sensors are no longer available.

This is NOT true.

If you are in the US, please go to www.fatquarterssoftware.com to order or make an inquiry. In the UK or Europe, please see Speake's website at
www.speakesensors.com to place your order with Speake. Both firms have FGM-3 sensors, the Speake SCL-007 chip, a kit of parts for the PC board, and the gradiometer PC board designed by Carl Moreland.

Best Regards,

Erich Kern
Fat Quarters Software
Murrieta, California