En route for the detection of a force from the vector potential

[F6FLT]

Hi All,

I started to test these new and recent components for me: Hall effect sensors and tunnel magnetoresistances.
They are used in industry for magnetic field detection.

I tested the SS496A and TMR2005 in a basic setup, to familiarize myself with how they react, and especially to get a practical idea about their sensitivity.

The SS496A (3 pins) is powered by a voltage V in the range of 4 to 7V, and the transverse Hall output voltage related to the deviation of electrons by the magnetic field is amplified and presented on the third pin.
Taking the mid-point of the power supply as a reference, the output varies from 0 to ±V/2 depending on the direction and intensity of the magnetic field. I made the middle point thanks to a simple resistance network with a multi-turn potentiometer. The voltmeter is connected between this mid-point and the output (see diagram). I used an HP3468A multimeter, which allows me to see µV.

The TMR2005 (4 pins) consists of a bridge of 4 magnetoresistances mounted in such a way that by circulating a current between two opposite nodes, the voltage between the other two opposite nodes that constitute the output varies according to the magnetic field. The most difficult thing was to solder the wires (component size <3 mm, see photo)! A magnifying glass is essential.

Note that the background noise is not negligible, I think there is noise in 1/f, and there is also the 50 Hz component of the mains current here whose magnetic field is everywhere. I have improved things a little by using a 4.5 V battery instead of my mains power supply, but it is still far from enough. The voltage of the voltmeter fluctuates by 50 to 200 µV.

Result in static:
I used a stack of 5 rectangular neodymium magnets of 50x20x2mm, and observed the output voltage as I turned them over at some distance, to reverse the polarity felt by the sensor.
At a distance of 2 meter, the output voltage of the SS496A varies by about 100µV, this is perfectly visible. At 3 meters, we no longer distinguish the useful signal from the background noise.
With the TMR2005, the sensitivity is much better, I have the same signal at 3 m as I had with the Hall effect sensor at 2 m. At 4 m, the signal is still slightly visible. And the TMR2005 is not the most sensitive of the two magnetoresistances I bought! I'll have to test the TMR2104.

These components are therefore extremely sensitive. I improved the range a little by sandwiching the SS496A between 2 ferrite rods in line.

Result in dynamics:
With a variable field, how do they react? This is a disaster. I used a 16 cm diameter coil, 100 turns, powered by a sine signal. Above 500 Hz, we begin to see the decrease of the signal. At 10 Khz, the signal disappears in the background noise, both with the SS496A and the TMR2005. Don't expect them to react to the electromagnetic field of a radio wave! 

Why these tests?
In particular, the analysis of the Marinov device shows that a spatial gradient of the vector potential should be at the origin of a force on moving electrons. This force should have an effect similar to the Lorentz force, but without a magnetic field, including on the electrons of a Hall effect sensor. Hence the idea of using this type of sensor to avoid the mechanical problems we have with Marinov (but there are others, like the B field leak, that's the problem I'm on now). This is the first step.

[ion]

The SS496A must have really good internal initial balance that it only requires a fraction of a percent trimming. I was surprised at this.

edit: reading further I see what you are doing, finding a midpoint of the power supply.

I played with some giant magnetoresistive sensors many years ago when samples were first available. The output was giant, but the size of the devices was so tiny I may have lost them. They were the size of tiny chip resistors.

Best of luck with your new research in this area, looks interesting.

[F6FLT]

...
I played with some giant magnetoresistive sensors many years ago when samples were first available. The output was giant, but the size of the devices was so tiny I may have lost them. They were the size of tiny chip resistors.

Best of luck with your new research in this area, looks interesting.

Thanks. The SS496A has the size of a small transistor but the magnifier is needed to deal with the 3 mm TMR.
There are more and more of these SMD components. I couldn't find a "big" TMR. I had already had the problem with wideband HF IC amps, they became tiny. It worries me about the future of the electronic handyman.

I wonder if it is possible to supply the Hall effect sensor with DC + an alternating component of the same frequency as the magnetic signal to be detected, to obtain by a feedback a kind of lock-in amplifier. The idea is to move the DC component to the frequency of interest. It's just an idea on the margins, I have little hope a priori, I think it would be known.
If you have experience with these components, maybe you can tell me more, especially if there are Hall effect sensors that reach the MHz (I don't have an explanation for their really very low frequency limit).

[ion]

From F6

If you have experience with these components, maybe you can tell me more, especially if there are Hall effect sensors that reach the MHz (I don't have an explanation for their really very low frequency limit).

My experience with these devices is limited to the basics, so I probably can't be much help, but will follow your thread and see what transpires, interjecting only when I have a useful comment.

Regards

[Peterae]

Not sure if this is of any help to you but if the are ratiometric and you use 2 stacked devices back to back you get a + increase in the output of one device while the second device gets a- decrease in voltage equal but opposite, measure between both outputs and you have doubled your output sensitivity.

[Itsu]

Something like verpies proposed here:

https://www.overunityresearch.com/index.php?topic=2194.msg33906#msg33906


Itsu

[F6FLT]

Not sure if this is of any help to you but if the are ratiometric and you use 2 stacked devices back to back you get a + increase in the output of one device while the second device gets a- decrease in voltage equal but opposite, measure between both outputs and you have doubled your output sensitivity.

The sensitivity is already very good. Today I am discovering that the Earth's magnetic field is perfectly detectable. I think I'm detecting it, unless it's the ferromagnetic masses of the surrounding furniture and appliances. I'm thinking of testing in the garden away from disturbing sources.

It is the frequency response that I would like to improve (but for reasons other than the subject of the potential vector so I will not dwell on it).

Something like verpies proposed here:

https://www.overunityresearch.com/index.php?topic=2194.msg33906#msg33906

Itsu

This discussion is not accessible to me.   

[Itsu]

Those private threads......


I will ask verpies if he allows publication here.

Itsu

[Peterae]

Ah they are all pretty slow, the article that itsu refers to is a differential setup like I suggested, back to back so will not solve your speed problem either.
Yes I made a compass using 2 3503 hall sensors years ago for the kids, hooked the output to a vco.

You can get 3d magnetometer chips that supposedly go up to a few MHz, there was a page started by 3dmagnetics to build a device using a PCB module.

[F6FLT]

...
You can get 3d magnetometer chips that supposedly go up to a few MHz, there was a page started by 3dmagnetics to build a device using a PCB module.

Thanks for this information. I'll save that for later. For the moment I am trying to highlight a Hall voltage when the current is made through a gradient of the vector potential rather than a magnetic field.
Near a toroid where a field is confined, we should have such a gradient.
I incorporated 2 neodymium magnets into a magnetic core, so that B is looped. There is a small magnetic leak on the magnet side, noticeable by a small force on a screwdriver. But even on the rounded side between the two magnets, the sensors detect a significant field, and I don't know if it's the one expected from ∇A, or the one related to leaks of B field.

(On the right, this is the magnet stack I used to test the remote detection).

[Itsu]

Those private threads......


I will ask verpies if he allows publication here.

Itsu

Got he OK, from verpies,  so below is his design, but i understand from Peter that it probably is not doing what you want.

The text to go with it was:

"The transmitting op-amp gain is set with the ratios of R2/R1 and R4/R3 - the two ratios must be made equal (e.g. by precise trimming). "



Itsu