Building a Fluxgate Magnetometer based Current Probe

RE: Building a Fluxgate Magnetometer based Current Probe

jw0752 — Fri, 01 Sep 2017 06:25:20 GMT

Hi Shabaz,

I made some more observations. The base level voltage I was reading can be completely reversed to a negative maximum by rotating the unit by 180 degrees. I suspect I am reading the Earth's field which is suppose to be 25 to 65 uT at the surface. My calculations gave a prediction of 25 uT. My maximum reading is 122 mV and the minimum is -122 mV. I used the formula in the EV kit manual to calculate this value. This is just too cool. Thanks again for blogging about the fluxgate.

John

RE: Building a Fluxgate Magnetometer based Current Probe

jw0752 — Fri, 01 Sep 2017 04:59:38 GMT

Hi Shabaz,

My Fluxgate Magnetomer arrived today and I have begun to experiment with it. It is a really interesting sensor. I was able to verify that it can detect my watch at close proximity. Just as you observed the oscilloscope trace will react upwards for 1/2 second and then down for 1/2 second. I also took a powerful neodynium magnet and played around with seeing the range of the magnetic field as well as the effect of the magnet's orientation relative to the fluxgate. Then I got down to business as I really wanted to observe how sensitive the fluxgate would be to a single conductor crossing in contact with its surface. Here is my experimental set up:

[View:/resized-image/__size/620x465/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-7f7252a6-0e0d-48ce-b61c-c7ff32af435f/7633.contentimage_5F00_185761.jpg:620:465]

and a slightly different angle:

[View:/resized-image/__size/620x465/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-7f7252a6-0e0d-48ce-b61c-c7ff32af435f/7462.contentimage_5F00_185762.jpg:620:465]

A positive flow through the circuit produces a magnetic field from the right to the left and a reverse polarity produces a field from left to right. I began looking for a response using the oscilloscope to read the output of the evaluation module I started in the uA range but quickly moved to the mA range before I began to see results that were above the noise level. At this point I decided to use the Fluke multimeter to read the output. This produced a much less noisy reading and I was able to dependably detect a 1 mA current in the wire. To highlight this I would watch the Fluke stabalize with no current and then I would apply a 1 mA current and watch the Fluke jump approximately one mV higher. Then I would switch the polarity of the power supply and once again let the Fluke stabilize before apply the reversed current and watch the Fluke drop by approximately 1 mV. Here are pictures for one of the many times I repeated this experiment:

[View:/resized-image/__size/620x465/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-7f7252a6-0e0d-48ce-b61c-c7ff32af435f/6866.contentimage_5F00_185763.jpg:620:465]

No Current Applied - base line

[View:/resized-image/__size/620x465/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-7f7252a6-0e0d-48ce-b61c-c7ff32af435f/1108.contentimage_5F00_185764.jpg:620:465]

Positive 1 mA flow upward past the fluxgate.

[View:/resized-image/__size/620x465/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-7f7252a6-0e0d-48ce-b61c-c7ff32af435f/0513.contentimage_5F00_185765.jpg:620:465]

Positive 1 mA flow downward past the fluxgate.

While conducting the experiment I was painfully aware of the sensitivity of the fluxgate. Just the movement of my body in the proximity of the test setup would cause the neutral baseline voltage to move a few mV one way or another. At one point I took the neodynium magnet and positioned it a foot away. Rotating it slightly I was able to zero the output to the millivolt meter. I suspect that with the proper noise management and differential amplifier the fluxgate could be used as a current probe just by laying it perpendicular to circuit board traces. This is way beyond my abilities at this time but it is fun to experiment and speculate.

RE: Building a Fluxgate Magnetometer based Current Probe

shabaz — Wed, 23 Aug 2017 18:59:27 GMT

This should really be a separate topic, but since the watch/clock stuff is already mentioned here now, here are a couple more 'scope captures of the watch.

This first capture shows the alternated magnetic field every second at the top (zoomed out to show about 4 seconds worth), and then the main part of the screen shows the trace zoomed into the transition from high to low.

[View:/resized-image/__size/620x372/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-7f7252a6-0e0d-48ce-b61c-c7ff32af435f/1602.contentimage_5F00_185758.png:620:372]

If I zoom in further, this is the result:

[View:/resized-image/__size/620x372/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-7f7252a6-0e0d-48ce-b61c-c7ff32af435f/5482.contentimage_5F00_185759.png:620:372]

It is really interesting. I *think* (not sure, I have no way of knowing without breaking open the watch! and soldering test connections to it [View:/resized-image/__size/16x16/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-7f7252a6-0e0d-48ce-b61c-c7ff32af435f/1524.contentimage_5F00_1.png:16:16]) that the current pulse was for about 9msec. Those funny Bart-Simpson-like spikes on that pulse I'm guessing could be due to the inductance of the coil inside the watch, and any capacitance, causing a LC circuit to be formed. All speculation though : (

Eventually the bart-simpson effect is not visible, and the metal is magnetized in the other direction, and that could be occurring on the downward slope perhaps?

In a way it is all like miniaturized warfare on that mechanism inside the watch, like an electromagnetic pulse (EMP) on a micro scale : ) occurring every second : )

RE: Building a Fluxgate Magnetometer based Current Probe

balearicdynamics — Wed, 23 Aug 2017 18:03:05 GMT

Another great project and very very original Shabaz ! Really like this.

Enrico

RE: Building a Fluxgate Magnetometer based Current Probe

shabaz — Wed, 23 Aug 2017 13:17:45 GMT

This is quite interesting, but easy for something with such sensitivity. Similar to what [mention:8fb233e8974445138ef0af2b63b49f94:e9ed411860ed4f2ba0265705b8793d05] discussed above concerning a non-contact probe for wandering around a PCB.

I took off the coil, and placed the PCB (still connected to the amplifier) onto a mobile phone, and powered it up. The trace below shows the result. A lower value means higher current.

The two red dots on the left side show what happened when I briefly pressed the power button (i.e. not long enough to power on the phone). When I pressed and held it for a long time, then it started booting up. The red dot on the right side shows the point in time when the password on the phone had been entered. I didn't capture the entire startup after that.

[View:/resized-image/__size/620x371/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-7f7252a6-0e0d-48ce-b61c-c7ff32af435f/4338.contentimage_5F00_185755.png:620:371]

I can also see things like on the phone lock screen, if I don't enter a password, after a while I can see the current consumption decreases, even though the display doesn't dim so early. Similarly, with applications running, it doesn't seem to be related to any interference from the screen, so I suspect it is indeed related to current consumption measured indirectly by the magnetic field off the phone.

It could be an interesting tool for testing mobile phone apps to see if they have inefficient areas consuming resources.. although the mobile phone development environments are sophisticated enough to handle all that.

I'm going to try to convert this into a general-purpose portable tool, a "fluxgate magnetometer for all".

It is crazily sensitive. it can pick up the magnetic field from a small alarm clock (the usual kind with a single AA cell) just ticking away, 10cm away : ) I can see each second on the oscilloscope.

RE: Building a Fluxgate Magnetometer based Current Probe

mcb1 — Wed, 23 Aug 2017 07:56:40 GMT

Nice job [mention:b0bc65b9ecdc4307bd967592f00e340a:e9ed411860ed4f2ba0265705b8793d05]

This is certainly stretching the DIY Test Equipment way out into the commercial prototype model.

I presume that reducing the size of the toroid has no effect on the results, just the ability to wind it.

I also interested in how you did cut it.

My first thought was a 1mm blade in an angle grinder (my favourite use for grinders) but that doesn't leave much.

You could probably bolt it down using a metal washer, and just sacrifice the washer in the process.

Irecall having these at my first place of work.

I recall using them occasionally but we not really in the digital realm back then.

http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1976-12.pdf

Cheers

Mark

RE: Building a Fluxgate Magnetometer based Current Probe

DAB — Tue, 22 Aug 2017 19:00:00 GMT

Excellent and timely post.

I am planning some experiments soon that will need this type of sensor, so now I have a good design to work from.

Thanks

DAB

RE: Building a Fluxgate Magnetometer based Current Probe

rachaelp — Tue, 22 Aug 2017 08:28:42 GMT

Hi Shabaz,

This is a great project and as usual I thoroughly enjoyed reading your write up. For low current applications, this solution would seem to be pretty perfect so long as you don't need to monitor higher frequencies. I'm very impressed, I might have to get one of those TI eval boards and give this a try myself. It wouldn't work for the SMPS applications I need a current probe for but it's very sensitive at low currents so possibly better suited to very low current applications than a more general purpose solution which can also handle the high currents. I think this is something that's going to also end up in my toolkit [View:/resized-image/__size/16x16/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-7f7252a6-0e0d-48ce-b61c-c7ff32af435f/contentimage_5F00_1.png:16:16]

Best Regards,

Rachael

RE: Building a Fluxgate Magnetometer based Current Probe

Workshopshed — Tue, 22 Aug 2017 07:25:27 GMT

Fantastic write up Shabaz and a very interesting sensor.

RE: Building a Fluxgate Magnetometer based Current Probe

jw0752 — Tue, 22 Aug 2017 05:38:45 GMT

Hi Shabaz,

That is a really cool sensor and very interesting what you did with it. How did you determine the proper orientation for the external field? Do you think the inductive properties of the external coil could be disruptive to the circuit under test? Would there be any merit to just looping a coil around the fluxgate magnetometer and not using a ferrite core?

John