Has any one tried making a mag sensor using 555 and library tag?

[Atul Asthana]

Here is a mag sensor made using library tag and NE555.
https://youtu.be/dP84iJlQCDI
Has any one tried making it?
1. What is the accuracy/linearlty around earth's mag field strength?
2. What is the repeatability?
3. Can a better/more stable oscillator (other than 555) be used, with lower jitter and lower temperature drift? And sinusoidal waveform instead of a high stress pulse/square waveform?

If anybone has made it, can you please share tech details and measured specs.
Thanks in advance.

[Altra]

I played with different circuits a while back using metglas mu metal. These were all self oscillating. Never did any specific testing.

You can see details in this thread

https://www.geotech1.com/forums/show...ating-Fluxgate

[Carl-NC]

I really like this idea, very simple and elegant. With 2 sensors you could make a BFO-style gradiometer. However, I like the idea of a feedback coil and PLL to force a constant frequency. Then you can ADC-capture the control voltage and use a micro for DSP. Or, use a capture-compare channel and a DAC channel to make a digital PLL. Lots of possibilities.

Atul, not sure what you are asking in #1. If repeatability means the difference between two sensors I'd guess it's not that great. And stability is very dependent on the sensor itself so a better oscillator likely won't make much of a difference. Most fluxgate mags I'm aware of use pulse drives.

[Carl-NC]

I played with different circuits a while back using metglas mu metal. These were all self oscillating. Never did any specific testing.

You can see details in this thread

https://www.geotech1.com/forums/show...ating-Fluxgate

I had forgotten about this... need to review. Thanks.

[Krzysztof]

I know 3 sensors that are suitable for DIY construction.
1. FGM3, digital output (Fig8 diagram) here:
https://www.geotech1.com/forums/atta...9&d=1361396864
2. FLC100, analog output
3. The PNI 11096 ASIC, 3-axis system with internal signal processing, sensitivity selection with SPI output
Arrangement https://www.youtube.com/watch?v=dP84iJlQCDI requires a coil 100mH for stable operation, works well.
All these sensors work as magneto-inductive (MI) generators.
They do not use the measurement of the second harmonic excitation generator, so they cannot be treated as fluxgate despite such a popular name.
Only here was the real measurement of the original fluxgate second harmonic used:
https://www.geotech1.com/forums/show...093#post290093
See Carl's design:https://www.geotech1.com/cgi-bin/pag...fmx1/index.dat
Virtually every MI generator works as analog, regardless of the switch used (transistor or 555), its temperature drift depends mainly on the magnetic material, it should be thermally insulated as much as possible, then the repeatability of the results is very high.
On the sensor coil itself there is always a sinusoidal signal but distorted (flattened) due to the saturation of the core (already above 0.5Vpp) by the Earth's magnetic field.

[Altra]

Hi Krzysztof, Do you happen to know which mu metal was used in the Speake FGM3 or FLC100? I know it was not Metglas it was lower permeability. I have a strip from a opened sensor, but no way to analyse it.

[Atul Asthana]

Thanks friends, your replies are of great help. I plan to get the sensors made, as against easier route of buying FGM/FLC (relatively expensive), since this is a learning exercise + useful outcome.

I am in touch with some engineering colleges, and have offered to help them (specially the electronics, communication/signal processing, embedded students, actually many other branches too, but those are not of concern here) do useful and research/development based projects, so that the students can learn the real value of their studies and can get into the industry/jobs much more easily.

I am planning to get (similar) sensors+freq counters made by these engineering college students, and monitor the local geo-magnetic activity, if this/similar oscillating sensors turn out to be good enough.

Once, I have the oscillating frequency captured digitally by a microcontroller; then temp compensation, feedback (force balance), filtering, linearisation etc. will be easier to achieve, provided the sensors turn out to be intrinsically stable and have good enough resolution.

I guess, I need to first figure out/test the accuracies, resolution and temp stability of these sensors in their raw form, before any signal processing can be taken up.

Thanks friends, I will keep coming up with similar questions/doubts and will expect help from you.

Atul

[Atul Asthana]

This is a great idea of using a pll to force balance, but looking at the oscillator's response on the oscilloscope, I guess, the frequency variation is far beyond the capture range of the plls I know of, namely the 4046 varieties.
However, for a narrow range, it may work perfectly, or you may have to design a very wide band pll.

[Atul Asthana]

Atul, not sure what you are asking in #1. If repeatability means the difference between two sensors I'd guess it's not that great. And stability is very dependent on the sensor itself so a better oscillator likely won't make much of a difference. Most fluxgate mags I'm aware of use pulse drives.

Repeatability: I meant to say that if I measured the same parameter of same strength and in same physical/environmental conditions at different times, what will the statistical distribution look like and how close will the measurements be to mean value? Its basically the standard deviation around the mean (though, some times, even modal value may be the reference).

[Krzysztof]

Hi Krzysztof, Do you happen to know which mu metal was used in the Speake FGM3 or FLC100? I know it was not Metglas it was lower permeability. I have a strip from a opened sensor, but no way to analyse it.

Look at this picture:
https://magnetometer-kit.com/wp-cont...tometerkit.jpg
It looks like it's a core strip from an anti-theft detector, fits the length and width.
Some non-detachable detectors had only metglass itself, maybe someone used it without disassembling them.
In any case, the reaction of the core depends (magnetic permeability) on the amount of material assuming the same conditions of current control.

[Altra]

Look at this picture:
https://magnetometer-kit.com/wp-cont...tometerkit.jpg
It looks like it's a core strip from an anti-theft detector, fits the length and width.
Some non-detachable detectors had only metglass itself, maybe someone used it without disassembling them.
In any case, the reaction of the core depends (magnetic permeability) on the amount of material assuming the same conditions of current control.

Thanks! The reason for asking. I made various sensors, most worked well with two exceptions. One if the metglas ribbon was glued to a pcb or encapsulated in epoxy. In both cases the sensitivity was very low. Later when I came across the YouTube video with the NE555 sensor, the man states that the Metglas ribbon needs to be free to vibrate.

https://m.youtube.com/watch?v=dP84iJ...ature=youtu.be

After that I realized why gluing or potting ruined the sensors by damping the mechanical vibration. That brings us to the Speake FGM3. When you open an FGM3 the mu metal is attached to the pcb and then potted in epoxy. Not sure why the Metglas needs to vibrate and the speake mu metal does not?

[Atul Asthana]

Look at this picture:
https://magnetometer-kit.com/wp-cont...tometerkit.jpg
It looks like it's a core strip from an anti-theft detector, fits the length and width.
Some non-detachable detectors had only metglass itself, maybe someone used it without disassembling them.
In any case, the reaction of the core depends (magnetic permeability) on the amount of material assuming the same conditions of current control.

Long time ago, I recall having seen some notes/emails/posts of xrays photos of one of the commercially available sensors, possibly an FG series, where everything was traced, including the circuit, which used one transistor in the oscillator and one transistor for converting the oscillation to logic levels.

May be, some one still has that photo.

[Krzysztof]

Read post 5, there is a diagram of FGM3.
This picture was not very readable.

[Altra]

Adding to the discussion, FLC-100 schematic. This sensor is driven vs self oscillating.

[Krzysztof]

One if the metglas ribbon was glued to a pcb or encapsulated in epoxy. In both cases the sensitivity was very low.

I had the same problem recently, after flooding, the resin got into the hole of the carcass with the core - the result, the frequency dropped 5 times, the sensitivity was poor.
Metglass stuck between the papers worked properly, but when the resin did not get into the carcass hole.
I don't speak English, it was hard for me to notice it on YT (I'm more of a francophone). English for programming only.
In the photo from post #10, apparently, metglass is in a plastic case, and they duplicate FGM3 technology - maybe there is some know-how here. My various bulk sensors have always worked predictably. Only once did I stick metglass under the PCB and wound the coil directly around the PCB and nothing came out.