OverUnity Research

Hello All,


I have been experimenting with a device that I will need some assistance with .
The basics are working but my health is very intermittent and my concentration for writing code is suffering and I believe that this project is important for any one who is interested in sensitive magnetometers and how they could be an essential tool for working with magnetic fields in3 dimensions .

If there is anything to using push pull (both positive and negative) (important)magnetic fields to perhaps behave in away that is unexpected
when rotated in a specific way  then this tool will be essential to be able to work with strong and weak fields both near and far .

When lightning strikes nearby the magnetic field is not 2 dimensional

One magnetic sensor gives very sensitive readings in 3 dimensions but of course from a single point in space . A very useful tool to begin with as the sensor is small and can be on plastic goose-neck or wooden stand .

These sensors are cheap and one particular model the mlx90393 can run many units in one bus .

I suggest that arduino uno or nano as the base processor knowing that there are better choices but as clunky as arduino can be its cheap and common and is up to the task at this early stage .

A goal of this device is to rotate a 3d object on screen in its simple form rounding things to be 0-360 degrees .
Calibration is important ans these units were designed for compasses  the z scale is not proportional to the x and y . and was designed to compensate for declination and inclination but is still as sensitive .

Further ahead many sensors can be spatially oriented and integrated to get the bigger picture .
I am of the understanding that the whole thing will shut down if its field becomes too strong ..means to an end .
I would imagine that a few monitors may be needed but the  1 in front of you is a good start

I currently use an environment called processing for the display but its very hit and miss depending on the system .

There has to be a better way and who here would not at least be curious what changes take place around anything that uses magnetic fields ?

The magnetometers can run at kilohertz if tweaked correctly and just think there is one in every phone nowadays .

There have been recent published studies about the effect lighting has on magnetic fields so there is a possible cause and effect reversal possible at the very least .

Search for arduino bunny rotate just as an example of a useful display.

I am asking for help and possibly offering something of great interest in return but I ask that you please stay on the subject of making the device work .

Some of you may know that we need a kind of oscilloscope for magnet fields and this I believe is it .
There are some very smart guys here whose input I will appreciate and perspective I respect but this is all that is really left to study regarding doughnut shaped devices like the ftpu we...(I) was lead away from back in 2006 .

Health issues have me on this swansong that I believe will at least lighten the load by offering you what I have available by specific omission and some lose specific information..(i hope you get that).
Please can we avoid the cloak and dagger stuff My life is not what it used to be and i don't really want to discuss anything but this device, its use  and hope that some progress will lead to more progress .

Here is some  basic arduino code ...note only common libraries used (preferable as many libraries have the same names in arduino)

// Distributed with a free-will license.
// Use it any way you want, profit or free, provided it fits in the licenses of its associated works.
// MLX90393
// This code is designed to work with the MLX90393_I2CS I2C Mini Module available from ControlEverything.com.
// https://www.controleverything.com/products

#include<Wire.h>

// MLX90393 I2C Address is 0x0C(12)
#define Addr 0x0C

void setup()
{
// Initialise I2C communication as MASTER
Wire.begin();
// Initialise serial communication, set baud rate = 9600
Serial.begin(9600);

// Start I2C Transmission
Wire.beginTransmission(Addr);
// Select Write register command
Wire.write(0x60);
// Set AH = 0x00, BIST disabled
Wire.write(0x00);
// Set AL = 0x5C, Hall plate spinning rate = DEFAULT, GAIN_SEL = 5
Wire.write(0x5C);
// Select address register, (0x00 << 2)
Wire.write(0x00);
// Stop I2C Transmission
Wire.endTransmission();

// Request 1 byte of data
Wire.requestFrom(Addr, 1);

// Read status byte
if(Wire.available() == 1)
{
unsigned int c = Wire.read();
}

// Start I2C Transmission
Wire.beginTransmission(Addr);
// Select Write register command
Wire.write(0x60);
// Set AH = 0x02
Wire.write(0x02);
// Set AL = 0xB4, RES for magnetic measurement = 0
Wire.write(0xB4);
// Select address register, (0x02 << 2)
Wire.write(0x08);
// Stop I2C Transmission
Wire.endTransmission();

// Request 1 byte of data
Wire.requestFrom(Addr, 1);

// Read status byte
if(Wire.available() == 1)
{
unsigned int c = Wire.read();
}
delay(300);
}

void loop()
{
unsigned int data[7];

// Start I2C Transmission
Wire.beginTransmission(Addr);
// Start single meaurement mode, ZYX enabled
Wire.write(0x3E);
// Stop I2C Transmission
Wire.endTransmission();

// Request 1 byte of data
Wire.requestFrom(Addr, 1);

// Read status byte
if(Wire.available() == 1)
{
unsigned int c = Wire.read();
}
delay(100);

// Start I2C Transmission
Wire.beginTransmission(Addr);
// Send read measurement command, ZYX enabled
Wire.write(0x4E);
// Stop I2C Transmission
Wire.endTransmission();

// Request 7 bytes of data
Wire.requestFrom(Addr, 7);

// Read 7 bytes of data
// status, xMag msb, xMag lsb, yMag msb, yMag lsb, zMag msb, zMag lsb
if(Wire.available() == 7);
{
data[0] = Wire.read();
data[1] = Wire.read();
data[2] = Wire.read();
data[3] = Wire.read();
data[4] = Wire.read();
data[5] = Wire.read();
data[6] = Wire.read();
}

// Convert the data
int xMag = data[1] * 256 + data[2];
int yMag = data[3] * 256 + data[4];
int zMag = data[5] * 256 + data[6];

// Output data to serial monitor
Serial.print("Magnetic Field in X-Axis : ");
Serial.println(xMag);
Serial.print("Magnetic Field in Y-Axis : ");
Serial.println(yMag);
Serial.print("Magnetic Field in Z-Axis : ");
Serial.println(zMag);
delay(500);
}






These sensors (mlx90393) are more expensive than others but not by much and the others seem to have errors in high gauss situations 15 bucks aud here

Thanks for reading
Computers are the best tools for some things and we never had such sensitive and small  magnetometers back in 1997 did we ?

Over to the brains trust ...


3D Magnetics

Thanks 3D M for the post.

if we were to make say a cube structure of these magnetometers to maps a cube of space, how big and how many do you think we would need.

I would have thought that the pcb that these are soldered to may affect the field around the chip?

Peter

inductance in wire will always create some distraction



As The hall sensor is very small and calibration can be quite extensive with many hundreds of calibration points per axis.

that's the beauty of the sensor area being small and digitised close by.

this wiki page is agood read
https://en.wikipedia.org/wiki/Ball_lightning

I am not seeing anything too difficult to do, I am interested in a minimum system design to start, say maybe 3 layers of 3 x 3 sensors, so a system initially using  27 sensors in a 3x3 cube.

I think the way to do this is design a pcb that is long and thin and place chips periodically along it's length 3 in the above case, then just keep using more pcbs linked together to form the 3x3x3 cube, then plug the cube using flying wires into a cpu board.

A raspberry pi would have been a bit better than an arduino because it has an onboard graphics processor for 3d visualization using a tv or monitor.

EDIT
Just starting to look at the technicals, first thing to note is that the uno or nano are 5V systems, the MLX90393 is 3V6 max, there are other Arduino systems based on 3V like the Due again the PI is 3V3.

Also there are speed constraints with I2C at 400khz clock frequency, a minimun of 108 clock cycles to read x,y,z data giving us a sample speed of 3.7khz per device (This may take longer as I have only glanced over the details), with 27 devices to sample this would allow us to do 137 complete samples a second which seems inline with the chip sampling speed, infact due to magnetic sampling noise it is suggested to operate around 50-100ms sample time 10-20Hz sample Speed.

If i were to choose a design, i would use a dedicate pic to continually sample a bank of 3 and relay this data back to the processor board, when done this way it would be possible to initiate all 27 chips to sample at the same time, to take a snap shot in time of the mag field, otherwise we would need to initaite each chip to sample one by one with quiet a delay during sample time, i dont think this would give a true mag field picture.
This would mean a long thin pcb with 1 pic micro and 3 or 4 mag chips.

Perhaps thinking that the simpler the better would allow more experimenters .

Its primary purpose is to aid in the visualization of magnetic fields.
It would be just a visualization aid ..not a precise instrument.

so visual feedback giving our tiny brains enough info to see cause a effect directly rather than exact angles.
Aural feedback ..brain / eyes  \i hope thats the right term.
perhaps an audio channel could come into play (7.2 of course )

Its easy to over think this  ....I know

First steps please ?

just one sensor is important to begin with dont you think?


Edit ..link for step1 but we are only interested in the magnetometer outputs .

https://learn.adafruit.com/ahrs-for-adafruits-9-dof-10-dof-breakout/magnetometer-calibration

I like your enthusiasm though.

There are some very skilled individuals here but considering it an artistic device might be a better context.

To promote a shared understanding , the simplest of tools at first is important .
That said the best of brains are required to refine it as proves useful ....or not .
the one i use can  sense a magnet from well over a meter away . and gain is not at maximum.

As this proves or disproves shaping and controlling magnetic fields to be useful and I think it will ,then many hands make light work .
And the there is my own situation which to date had been not exploring this ..in fact being led away may be more accurate .

Get some screen objects spinning and see if it inspires you it certainly has  not raised as much curiosity as i had anticipated .

With vr technology the possibilities are endless but walk first .

Peter, the answer to you question is 6  but integration would go well from there

surrounding your self with 6 and one in hand (portable)

All arduino have 3.3 volt and 5v ..no hitches to get to first base
3k per device is fine and latency compensation is common .

having the processors near the reference points does make any noise more controllable .
only useful serial data is then possibly affected in stronger fields .
Adruino nano or adruino micro are barely bigger than the sensor boards .
I politely ask that we stay simple untill when ...iffff    a few people become involved with basic units.

It may well be a slow moving weaker field at distance that we need to detect .

perhaps get same compasses and surround your self with them while moving a magnet or a coil and battery .
it is impossible to see six at once and needle inertia makes them almost useless ..magnetometers dont do that.

It may be that a set 10 metres away would be useful ... Blank canvas .
I cant find any prior art on this except a few un finnished lighting studies and a 3d scanner on hackaday which only excite me more .
Nobody has to my knowledge made one of these for this purpose .

All way too busy flying their drones with their smartphones

3D Magnetics



added link

https://www.physics.ncsu.edu/physics_ed/Articles/LunkBFieldArticle.pdf

Hi 3DM
Sorry for the absence not been too good the last few days, dam flu.

Yes I get it, simple is fine.

I have been unable to find your referenced magnetometer board on https://www.controleverything.com/products.

Either I am missing something or it is discontinued there ?

context ,sorry ,

That link is within the arduino code and it is dead.
I always post the whole code as respect for the writer

The links i pasted  work at this time .

The Arduino Nano has an onboard 5V regulator which supplies the processor, the USB driver chip can provide this 5V supply when connected to a USB PC port, the USB driver chip does also have a 3V3 supply output but this is not used on the nano but does get fed out on a pin for external uses, this could be used to power the MLX magnetometer.

The MLX datasheet says max pin voltage is 4V so it is not possible to connect the MLX to the Nano without a bus converter.
The Specs of the Nano clearly state it is 5V here
https://store.arduino.cc/arduino-nano

The Uno is the same, Specs are here
https://store.arduino.cc/arduino-uno-rev3

It is possible to convert a nano to work on 3V3, this involves replacing the onboard 5V regulator to a 3V3 version and also altering the clock speed from 16MHz to 10Mhz see here
https://www.ba0sh1.com/blog/2013/03/30/tutorial-3-3v-hacking-for-arduino-nano/

probably easier just to add a bus translator to the MLX board

Here's a typical bus translator for the arduino boards
https://www.adafruit.com/product/757

There is also an Arduino Pro Mini that is cheap and 3V3 but it is TTL serial not USB so a RS232 to TTL converter would be needed.
https://store.arduino.cc/arduino-pro-mini

Quote from: Peterae on 2018.02.04, 18:20:02
The Arduino Nano has an onboard 5V regulator which supplies the processor, the USB driver chip can provide this 5V supply when connected to a USB PC port, the USB driver chip does also have a 3V3 supply output but this is not used on the nano but does get fed out on a pin for external uses, this could be used to power the MLX magnetometer.

The MLX datasheet says max pin voltage is 4V so it is not possible to connect the MLX to the Nano without a bus converter.
The Specs of the Nano clearly state it is 5V here
https://store.arduino.cc/arduino-nano

The Uno is the same, Specs are here
https://store.arduino.cc/arduino-uno-rev3

It is possible to convert a nano to work on 3V3, this involves replacing the onboard 5V regulator to a 3V3 version and also altering the clock speed from 16MHz to 10Mhz see here
https://www.ba0sh1.com/blog/2013/03/30/tutorial-3-3v-hacking-for-arduino-nano/

probably easier just to add a bus translator to the MLX board

Here's a typical bus translator for the arduino boards
https://www.adafruit.com/product/757

There is also an Arduino Pro Mini that is cheap and 3V3 but it is TTL serial not USB so a RS232 to TTL converter would be needed.
https://store.arduino.cc/arduino-pro-mini

Peter,

How about the Arduino Teensy 3.2 or Teensy 3.6.  These are both 3.3v boards with USB that are based on the NXP M4 ARM series mpus.  You can download a module that allows these to be programmed with the Arduino IDE and they are library compatible.  Performance is high with clock speeds of 72Mhz and 180Mhz respectively plus the Teensy 3.2 also has 5v compliant inputs.

Regards,
Pm

Thanks partzman, indeed they are 3V3 systems and cheap, see what 3DM thinks of this, I don't mind sticking with the nano and using a bus converter, but definitely easier using a 3V3 system ;)

Every arduino i have worked with interfaces with no issues uno nano micro ..havent tried teensy but they all have 3.3 volt and the listed magnetometer and others interface to all with no problems .

just using raw data and a  screen object is eye opening as you can move it about .

I do urge use of the 90393 for many good reasons after being burned with falsely labled other types.
and poor settings implementation.

its not until you see that inertia free 3d magnetic compass displayed by a simple on screen 3d object that you will get inspired .

One unit with links above and a few local tweaks is enough to see some light where you were almost blind previously.

You may have to slow down your push pull drive to see if you have some harmony....small steps.
Push Pull is important because the surrounding field has at least 3 dimensions . Push pull is 2 and coil orientation/phasing makes another one .

All basic stuff really  but seeing the fields in some way is all we need to navigate

Hi 3DM

Yes I totally agree, it will be very interesting to be able to view what is actually happening in 3D, such a crucial device for anyone trying to understand a kick or any sorth of magnetic field interaction, hell even Tesla used magnetometers during ligtening strikes to understand what goes on.

Thanks
Peter

north south east west
inclination declination for all points

But is there an easier way of expressing these values for common ground ?

Hi 3D M
Thanks for your posts ;)

This subject is quiet complex, Still trying to digest the wealth of information you have posted in your previous posts.

I have ordered 2 MLX90393 boards.
https://www.ebay.co.uk/itm/222692173738

and a Teensy V3.2 to get me started.

The Magnetometer calibration article you posted was interesting, the code is available for download, hopfully with some tweaking it can be made to work with the MLX, i like the Spherical bubble representation for calibration, very clever.

I need to have a play with the setup and think about this a lot more yet, It would be nice if some of the other brains in the forum can think this out as well.

Yes it was the calibration sphere that turned a few more lights on in my partially fried brain.

Calibration in our case is to create 3d close to linear response on all axes as opposed to a compass.
and 360 degree interpolation plus additive and subtraction data to record and alert to any continuous spin( i hope i explained that right)



In other words xyz need to have even response form each axis .

Think of a cube with 6 faces, one in the centre of each face ...corners if you like ,plus a centre sensor. A centre object will rotate as you will soon find with one sensor .
Then the hard part..the display back ground needs to be interfaced with the outer sensors, perhaps a ring that spins in 3 planes and changes color with common mode changes ....a zero button will be important .

Recording and playback of our data streams is also essential so that we can share results and move forward ..

As you see we need the likes of Smudge, Ion, Whatsup, Centraflow just to name a few of the  great minds here who may have interest in creating a tool for a greater understanding of field interactions to develop.

I am lead to believe that being able to control a spinning 3 d magnetic field has merit and potential for discoveries.
MRI operation is one area where my interest was sparked further after  some kind words from someone who would know .
Of course MRI is reaction based rater and designed to see into tissue for  medical diagnostics.

For those who just like some drama and entertainment another "possible" reference is " the bell island boom"...but I can assure everyone that pootube hits has no place here .

The first step is to make a tool so that  magfield interaction..of which we really know so little despite calculation and equations that boggle the mind .

3d analog display ...but digital in operation,analog in use .

I remember that Lester Hendershot was trying to create a better compass according to the remaining stories at least.

I once took a well versed geologist to a very remote place where I had found a substantial quantity of gold  ...he said that he had no idea how the gold was there  and would never have predicted it ....BUT ..after a drilling program and assay he could predict other areas nearby with some certainty.
Which he did .
3D

There is little nice chip called esp8266 and a bunch of ready pcb like Wemos with battery shield. A device could send data via wifi to computer for analysis. Just an idea. Such device will be not harder to make then original, just sensors and Wemos with battery shield and battery. No cable, handy one. Consider it.

Quote from: forest on 2018.02.10, 09:14:46
There is little nice chip called esp8266 and a bunch of ready pcb like Wemos with battery shield. A device could send data via wifi to computer for analysis. Just an idea. Such device will be not harder to make then original, just sensors and Wemos with battery shield and battery. No cable, handy one. Consider it.

Thanks for raising this .
I thought I had made the arduino point earlier and well enough for understanding why .
Here is a longer version
Yes there are many versions of controllers that can achieve the same end but why not stay simple and on the same page with code .
I have trouble with the word shield ..it a bloody circuit not a defensive weapon .
and sketches are programs while we are at it , and many options are available . There are better micros I know ,if you think that many people working with different devices will help then I fail to understand the purpose of suggesting yet another micro ..it has been covered and I'm sure that different micros will eventually be used as and if advanced systems develop
Arduino is the most available ,very cheap , and up to the task .

Wifi can be added to any arduino easily and cheaply if so desired ....later

This is not a complex device and adding another em field to the equation early on would be un helpfull.


That said it will be complex to integrate the sensors code to detect the outer field.

But it will seem simple once the algos have been integrated where would be written in assembler on a custom micro but we are far from that point .

The sensor I selected was not done lightly and not without trying a few with time consuming (wasting ) results , it has both addressable i2cbus and an spi buss along with a few other attributes that are important should the gauss become more intense .

Of course there are many paths ..how about a common path ?

If it is so hard just to be on the same page with a basic tool imagine how difficult it would be all having different tools .
I would add that having this same tool in many hands may allow safer voltages to be used and more simple experiments using lower frequencies .
Having said that, it is help and collaboration that is required but perhaps nobody sees the relevance of this at this time..then the joke is on me !
Perhaps it always was ?
Which makes this whole initiative pointless and not for any of us at this time  .

I would agree that staying with the Arduino platform initially is a good idea, unless someone could definitely say that there is not enough processing power with that device. Standardization can be a good thing.

Regards

Quote from: ION on 2018.02.11, 05:22:20
I would agree that staying with the Arduino platform initially is a good idea, unless someone could definitely say that there is not enough processing power with that device. Standardization can be a good thing.

Regards

Yes,
processing can take place in the host display program if power becomes an issue .

Quote from: ION on 2018.02.11, 05:22:20
I would agree that staying with the Arduino platform initially is a good idea, unless someone could definitely say that there is not enough processing power with that device. Standardization can be a good thing.

Regards

What limited info I know about the NXP K20 processor in the Teensy 3.2, it should be more than capable to run the software for the MLX90393.  For example,  even though  it does not have an FPU for floating math operations (these operations would therefore be on the slow side), it does have an MPU which allows one to use say 32 bit (16/16) integer/fractional algorithms for high speed math.  Also, the external oscillator can be run up to 2x without any ill effects if more execution speed is needed.

The problem I see and perhaps 3DM can shed some light on this, is the latency rating of the MLX90393 per axis of 192us minimum and 66.56ms maximum.  This parameter is programmable depending on the amount of digital filtering required, so, how does this realistically affect the realtime sampling of our moving magnetic fields?

Regards,
Pm 

partzman

I too was questioning the ability of these magnetometers and also the cpu Arduino board for suitability, but I then realized that for the moment it is immaterial, I think the important point is that at the moment we have no tools to help visualize what happens to the field when we pulse coils, so we need to develop tools to visualize the field and the interactions with the surrounding structures and layouts, any tool is better than none and you have to start somewhere right ;)

Anything after then can be refined in a number of ways.

Quote from: Peterae on 2018.02.11, 20:41:03
partzman

I too was questioning the ability of these magnetometers and also the cpu Arduino board for suitability, but I then realized that for the moment it is immaterial, I think the important point is that at the moment we have no tools to help visualize what happens to the field when we pulse coils, so we need to develop tools to visualize the field and the interactions with the surrounding structures and layouts, any tool is better than none and you have to start somewhere right ;)

Anything after then can be refined in a number of ways.

Hi Peter,

Yes I agree that tools are needed.  3DM has apparently used the Arduino with the MLX90393 so I would be interested in his evaluation of even a relatively slow moving magnetic field.  What might be a solution would be to use the MLX magnetometer for DC to very low frequency waveforms and to use a 3-axis arrangement of sense coils for the higher speed measurements.  The Arduinos have enough A-D to handle this and the math could be done on of off board.  The 3-D display software should be nearly identical for both types of measurements IMO.  Just some thots!

I plan to build the Teensy3.2/MLX combo but I need to order the parts.  I will also give the sense coils a try as well.

Regards,
Pm

Good points,
Latency would need to be compensated and therefore increased.
As this is a visual aid and not a control instrument it will not matter .

A good example is I play a midi keyboard via a computer via a vst and afew other things in the signal chain  and latency of 100 ms is the point where playing a keyboard is out of sorts.
If i play for a while the brain adapts to this.

however as it can be allowed for anyway it should not be an issue other that to have all the  sensors  synchronized.

If it were a control instrument then huge issues would be present .But these are  visualisation units  will not have any direct connections to the item under test or be required to generate feedback ...other than a visual reference to the alleged Coriolis type disturbances that have been coined as a "turbine effect". Which apparently do have a lot of latency so even if we had 500ms heaven forbid.. it would still be useful as out brains will compensate anyway.

That does not rule our using magnetometers for control ...who knows ,but you circuit coils or whatever you use to manipulate the field would be electrically and physically (bar the field) isolated from the visualization device .

It wont be long before all these magnetometers are hidden behind other code and processors as they become more refined and less accessable ..like many things these days .

last time i went camera shopping i could not find a single one that had a phone in it ...i digress

@Partzman
the sense coils arangement has had much interest from me as well and who knows as this develops there may be a simple way like that.  I hope so ...but as you will see the magnetometers are so very sensitive and the fact they they are a pin point in size and precision made  is an advantage .

I wonder if spherics audio labs were making spherical speakers which lead to spherical fields ?
Mono surround sound?

Quotelast time i went camera shopping i could not find a single one that had a phone in it

O0

In my opinion, an Arduino Nano makes for a great first system platform...perhaps for the end system depending on the functions. If we're talking bench results from spinning pulse motors, it's ideal. If we're talking  meters or feet of space between large systems then that same Arduino code can normally be directly ported over to ESP8266/ESP32 devices. MESH's Nodes the whole 9 yards could then be implemented under the same system base code. But not at first, am in complete agreement there.

Using the Arduino IDE may not suit everyone, but really, pretty much anyone who's used microcontrollers will be familiar with it. Doing it some other way may bring in platform problems/OS problems....your code I believe is in Processing, but will work in the regular IDE. There could also be many many problems with folks not being used to something and dropping out...so it needs to be simple yet highly effective for collaborative purposes. I'm good on flexibility, but you won't catch me near a Windows program running Visual C lol.
Am also unfamiliar with Processing, but build gadgets with screens on them daily...so you will see bias in my first post here on the subject ! we all do it.

My preference for direction would be in a hand held unit, perhaps simply sat on the bench, with an ili9341 or ST7735 screen showing the output - see, how long did that bias take. A cheap and cheerful 2" or so colour graphing real time solution.
How we would expect to see the magnetic properties is important - what would make the most sense to the average person looking at it and, perhaps most importantly at our general non M.I.T. levels, how peers can view and understand the information too.
Well, as I read this thread (thanks Itsu for the heads up) what i'm looking at is the determination of the op to give us a solution to view magnetic fields as they happen. An electronic 3D version of that green flexible sheet stuff.
There is merit in several people presenting their version of the idea, there is merit in 1 avenue with the leadership of the op being the boss.

I like it, though presently have no spare monies for the MLX90393. It does remind of gyro's though and am very comfortable with those (self made quads and mini aircraft systems etc).
The whole 3.3V problem isn't one, just use a level shifter: https://www.ebay.com/itm/5Sets-4-Channel-Bi-Directional-Logic-Level-Shifter-Converter-3-3V-5V-For-Arduino/202151474158
5 of them for $1.29
If powering multiple sensors a 1117 3.3V regulator or similar will supply more than enough juice: https://www.ebay.com/itm/NEW-10pcs-AMS1117-3-3-LM1117-3-3V-1A-SOT-223-Voltage-Regulator-FC/152815072274
10 of them for less than $1
(am in the USA, but am from England btw, in case my language and units are all over everywhere)
For completeness, should you wish to go for a portable screen solution -
2.4" ili9341 based touch screen for < $5: https://www.ebay.com/itm/2-4-TFT-LCD-Display-Shield-Touch-Panel-ILI9341-240X320-for-Arduino-UNO-MEGA/192163706724
Note also the SD card slot...which could be used for recording a session and transferring online or to any database.

Just  finally, while I totally agree about the Arduino, an ESP826 solution afterwards would allow for live data streaming to ThingSpeak or another host.

A very worthwhile endeavour that you have with the idea and I would hope to assist however possible.
Who knows, make the lot Open Source, but sell completed units to folks for a great hobbyists solution. It's certain to further magnetic fields knowledge, whether 100's were sold or you throw your hands in the air and make just 1 of them.

Thank you for your input Slider,
There are cheaper chips for 5 bucks but the 14 dollar 90393 is important as it will handle high gauss and has excellent internal calibration settings and can be i2c addressed. and later SPI  as more response may be required.


The code is what runs on the arduino nano to send the xyz data to the pc .

The PC displays the graphic and does need to be large.
A different processing program on the pc displays the bunny .
,calling for the xzy values from the usb com port.

The first step is a single unit in the centre (hand held or mounted in the centre of the device being tested /built.
Bunny rotate is the best example

The implications for all kinds magnetic research are obvious
Anybody curious as to how a coil or a series of coils may interact with the field that is already there should get a lot just from making a single unit .

It is my contention that it will enable intuitive research ,a tool that has never been available before .
I would suggest that we have been working in the dark ,having said that ,this would not need to be calibrated ,rather orientated.

You could have a beeper sound when the 6 external field sensors behave in a certain way ...so many possibilities and


sensing field rotation at a distance around the unit being tested will show the kind of influence that SM and Spherics has indicated as important.
There would be no doubt that your generated field is rotating and would avoid boom and bust scenario as well being able to use safe voltages ,as 99% of people, for their own good reasons avoid .

It would encourage analogue  wave forms  from multiple sources  allowing a control unit to be built in the neutral space in the middle

These things are exactly like gyro chips in the way they output data xyz..


All of this assumes that is possible to"tune into magnetism "
I like to imagine that it may be possible to sail on the field or in the field
And any body  cares to  see the invisible world of mag fields ,rather than a hand and thumb  like in kinder-garden.

Does a spinning/rotating /modulated  mag  field have inertial properties under certain conditions?


Here is some study material that is relevant
https://hackaday.io/project/11865-3d-magnetic-field-scanner

http://www.rotoview.com/sensor_kinetics.htm

member Hope at Stefan's shared this today

here http://overunity.com/17633/smart-phone-magnameter-app/msg517785/#new

[sorry if similar was already mentioned]

to note, I have been told that there are other portable devices available to do this ,view the magnetic field [not certain of cost]

will have to ask for more info.

respectfully
Chet K

A good subject Chet,
Dumbing it down to a 2 dimensional  APP would be the wrong direction as we need to develop external integration of multiple sensors as the comments suggest.
A small screen would not achieve the goal ...but should inspire what i propose here in the first post.
That is not to say  a vr environment will not apply later as / if we learn .
Like so many things we tend to shoot for the stars when we know so little of what is under our feet .
3D is key ..Im sorry  i have to repeat it . The screen in front of you will do it .single raw magnetometer data to start with.
Shopping for apps might be  fun however and that is the problem in this age of delightenment.
All of this is probably too late as everything is packaged and upsold.
I am not harping on about high voltages and electron tubes at least !
This is also not a control circuit but a means to get to first base in a collaborative way ,which protects everybody .

This is about building one ...please respect the aim and intent of building one here .
There are too many cases of people having momentary success yet not knowing what is happening .

WE do know that rotation of  fields are important
I will assume that noisy signals are unpredictable even in one dimension


some relevant material is here if you are interested in participating physically
https://hackaday.io/project/11865-3d-magnetic-field-scanner

Have bought this one:
https://www.ebay.com/itm/CJMCU-90393-MLX90393-Digital-3D-Hall-Sensor-Displacement-Angle-Rotate-3D-Positio/112460263873
O0

It's going to be 3 weeks or so before it arrives, but there are several projects here in front of it all.
Am going to d/l Processing and have a look...(for the several years i've built stuff with uc's it hasn't featured).
Oh, of possible related relevance - system here is a tower PC, Q6600, Linux Mint 18.2
All in all, yes this needs to happen and it needs to be at a price point that all can afford.
Do you have a Github space set up ?
The need will be to share, manage, improve code and to have test environments in various locations that can verify results. Well, I know you've thought this through, am just looking at pitfalls that often stop development, as stated.   

Quote from: Slider2732 on 2018.03.06, 21:22:21
Have bought this one:
https://www.ebay.com/itm/CJMCU-90393-MLX90393-Digital-3D-Hall-Sensor-Displacement-Angle-Rotate-3D-Positio/112460263873
O0

It's going to be 3 weeks or so before it arrives, but there are several projects here in front of it all.
Am going to d/l Processing and have a look...(for the several years i've built stuff with uc's it hasn't featured).
Oh, of possible related relevance - system here is a tower PC, Q6600, Linux Mint 18.2
All in all, yes this needs to happen and it needs to be at a price point that all can afford.
Do you have a Github space set up ?
The need will be to share, manage, improve code and to have test environments in various locations that can verify results. Well, I know you've thought this through, am just looking at pitfalls that often stop development, as stated.

Thank for your input Slider ,
Those are exactly what I am using .
I would rather we did everything here as the code we will use will be easy to paste or zip up .
there will only be 2 main packages ...the arduino code ,to communicate  and the host code (processing or python)to display
It should not matter pc or linux from what i understand .

I know that we would have to agree on the libraries used and there are so many libraries with the same name it could be confusing for some ..i have been caught out too many times .

I would rather steer away from using 3rd party repositories and keep it all here with forum members at least for the development period. And it would be open to all members ,even the public.
a zip package with libraries and running code should not be too large at first.
It may be possible not to use anything but the common wire library? My skills here are fairly low so no, i have not though this through beyond knowing that this tool will be required for continuity when testing coil arrangements .
As far as coding is concerned I can make things work but very untidy an as my health is poor my brain function varies randomly.
I once wrote a program that was 3 pages long (commodore 64) i showed it to my brainiac friend and he reduced it to 12 lines .
I worked the same but was not ever critical on memory or execution speed.

RE my health
Chronic fatige syndrome is the best that the medical profession can do ,but I keep looking even if they don't /can't.

I even wonder if screwing about with hv and spin fields may be bad for health and there is some evidence of this .
Many people have contacted me over the years warning of health issues .

Otto's death was one where the timing cannot be excluded.

Gk made a ring that scared him enough to want to put it in a microwave oven .

SM had throat cancer and also questioned the health implications .

That is why it is important to be able to work with low velocities and low voltages . And to be able to see the field of influence .

Whatever we do it can not be in the magnetic darkness we currently work in.

There may be medical implications here both positive and negative .

There are many articles on NMR which are inspiring and educational but focused on body imaging only.

But it is help i /we require and can I ask what the pitfalls of keeping it all here might be  ?
As we move forward just keep ver1. 1.01 1.02 etc so that any body can go back or forwards.
I would possibly get out of hand but we could change it as the problem develops .
For now it is small and we need to think of something better than a rabbit spinning ...though it is easy to see.

The outer field ring as a multi coloured torus ?...six rabbits wont do it but it it would be the starting point.
I mean 7 counting the middle one.

While im at it, i can explain why 6 .
Imagine a cube with six faces all with an arrow pointing a single point any where within the cube .
It takes 6 to be accurate ...except we are detecting the field disturbance from that virtual single point which make it possible to spin tilt and tumble, raise and lower, the outer ring in a virtual space ..the computer screen .
It is simple yet  complex and easy to over think .
There are many iterations of reference that will grow as understanding progresses ...but simply it is 7 sets of xyz axies.
A common zero button will be important .

The spacing of the 6 sensors will be up to the user but at few meters they detect a neo magnet easily
,and can be gain adjusted for all axies,they would be top/bottom/left/right/back/front ...luckily they all have different first letters......oh and c for centre.



I know it takes time to get these things in the mail ,one lot took 5 weeks (aliexpress seller)
I also know that one you get one up and running ...you will share in my enthusiasm.

Thanks
3D

Perfectly fine to post the code here....I prefer it, hopefully others too.
It's just the amount of things that seem to be on Github, like Sourceforge used to be. Knowing you wish to post here is perfect for the project, because it allows for a history, for the latest code to be the last version posted and can show folks in the future how it progressed to whatever point it's at when they take an interest.
Very often in my own coding, a section may begin to give troubles after a feature is added and i'll load up a previous version, to redo whatever it was from that point. The new code then works and gets the newest version number. The thread might well work in that way too.

Sorry to hear about the cronic fatigue. Am not sure if I have something similar. About 5 years ago something bit the lower of one leg and it all went rock solid on the calf to about the diameter of the base of a large pills bottle. It lasted for about 2 months and slowly faded. After that, even getting to the mailbox can leave the legs completely tired out and so sluggish that it's like dragging them using the hips. But, other times a half mile run may be possible (am 48). Best guess is some form of tick got me, or a spider.

For magnetic issues, yep, can relate there. Personally I have troubles if messing with neodymiums when perhaps rewinding an RC motor. Strong field changes and exposure that's above a few minutes, which I would imagine is a common problem. 

A library may offer a shortcut and the Adafruit ones are normally good and reliable. But am simply thinking about no need to reinvent the wheel.
Speaking of which: https://github.com/tedyapo/arduino-MLX90393
There seem to be a couple of different libraries.
A library could be used as is, or studied and part replicated. Am thinking of the MIDI library for an example of that, where only parts of it may need to be used. Whereas a TFT screen library may need everything.

I found some cheaper mag sensors

https://www.aliexpress.com/item/CJMCU-90393-MLX90393-digital-3D-Holzer-sensor-displacement-angle-rotation-3D-position/32829326091.html?spm=2114.search0104.3.2.5f2db7357gnvm2&ws_ab_test=searchweb0_0,searchweb201602_4_10152_10151_10065_10344_10130_10068_10342_10547_10343_10340_10548_10341_10084_10083_10618_10307_10131_10132_10133_10313_10059_10534_100031_10103_10627_10626_10624_10623_10622_10621_10620,searchweb201603_37,ppcSwitch_3&algo_expid=b3ac6cd9-cf2e-4e98-9905-10947e761d7d-0&algo_pvid=b3ac6cd9-cf2e-4e98-9905-10947e761d7d&transAbTest=ae803_5&priceBeautifyAB=0


Just in case that link is rubbish   "aliexpress mlx90393" should find it "
6 bucks each

The MLX90393 has arrived.
Will be having a look at it over the next couple of days :)

Good news Slider,
Im not very with it of late but will be better soon i hope.

I have excellent information that a cause or effect was a coriollis type magnetic effect that reduced to null in the middle and fell off at distance .

It was never discovered for certain which was cause or effect but the structure of out known world suggests both.

Im sure you will get inspired once you start seeing how responsive thse things are .

I would like to pose a basic scenario to the brains trust here in hope that there may be some better equipped mentally than myself.

Does a transformer driven by an ac or dc signal of single phase and source from single supply, saturate and desaturate the same as one that is driven from split supplies or  polarity reversed single supplies like a h bridge?

My question relates to wether we are in fact failing to interact with our surroundings by not understanding the difference and more importantly not applying it .

I could be wrong, as well as niave ,but perhaps we are missing one comlete half of what is important?
push pull tecniques were emphasised by SM on several occasions .
As was field rotation .

3D asked:

QuoteDoes a transformer driven by an ac or dc signal of single phase and source from single supply, saturate and desaturate the same as one that is driven from split supplies or  polarity reversed single supplies like a h bridge?

Do you mean symmetrical saturation / de-saturation?

I would think a single ended supply source of AC should saturate and de-saturate a transformer symmetrically if the amplifier or oscillator output driving the transformer is capacitance coupled to the transformer.

I think in any case it boils down to current drive and drive reduction both forward and reverse required to saturate/de-saturate the transformer,

however a transformer will automatically de-saturate if the driving single ended supply is disconnected or if forward current is reduced.

If the supply is slowly brought to zero current, there may be some remanence in the core when the transformer comes out of saturation.

This and other factors can result in "flux walking" which can cause early saturation on subsequent cycles.

We have not yet touched upon loading effects on the secondary of the transformer and how this affects saturation.

Maybe if you state the idea or what you have in mind in another way or draw a simple set of schematics we can pitch in some ideas.

What is it would you like to do or try or have in mind? We can test it for you.

Regards

I was looking at this thread from the start and was intriged by this.
I bought the sensor and have some arduino's, so was kind of waiting for the thread to develope, but things stalled
rather quickly after a promising start.


So i was looking on the net for some sketches which could be used.
But almost the only one is the one mentioned in post #1 which can be found here:

http://www.arduinoprojects.net/sensor-projects/connecting-mlx90393-magnetic-field-sensor-arduino-uno.php#codesyntax_1


This website shows the connections and how to display the outcome,
so thats what i did, see video: https://www.youtube.com/watch?v=gCGvBwM5n3I


So the seriel monitor show repeatedly the values of the 3 axis, so now we need to build on that further to
get a more sophisticated visualisation.


Any programming guru's still interested?


Regards Itsu

interesting stuff.

If you were to make a cube matrix, then maybe have the sensors just light 3 color leds, local to each sensor position. Like those cube matrix led kits. Once you get one sensor module working then make a bunch.  No programming. real time. Watch the field of a coil and magnet when pulsing or varying coil input.

The color changing view would give a kind of different way of seeing the fields rather than seeing them as lines. May be helpful. And could be programmed to do so as well where the program could take averages between sensors to show a solid field on a screen.

Mags

Quote from: Itsu on 2018.12.29, 14:51:11
So the serial monitor show repeatedly the values of the 3 axis, so now we need to build on that further to
get a more sophisticated visualisation.

Any programming guru's still interested?

Regards Itsu

I don't know about guru, but I whipped together a quick demo for you that *should* work with the setup you're describing, assuming you have a relatively standard Arduino and computer set up.


For efficiency, you can modify your arduino code as follows:
- Change Serial.begin(9600);  to  Serial.begin(57600);
- Change "Magnetic field in X-Axis:" to just "X : "      (likewise for Y and Z)
- Change delay(500);      to delay(20);

I don't have a magnetometer chip handy so you will have to be my beta-tester to see if it works right (or at all) ^-^ O0

If all works well, it should be fairly easy to modify and expand to whatever purposes we're looking for.  (LED visualizer, etc)


Mag Field Visualizer prototype code:
https://drive.google.com/open?id=17QSrOkvh3JUlxviBcByrqdJENGHbY4x0

Hi Mags, Reiyuki,


Thanks for the interest.

I have made these changes to the code:

- Change Serial.begin(9600);  to  Serial.begin(57600);
- Change "Magnetic field in X-Axis:" to just "X : "      (likewise for Y and Z)
- Change delay(500);      to delay(20);


Now the serial monitor shows:

Z : -37
X : -82
Y : -3
Z : -34
X : -76
Y : -3
Z : -36
X : -74
Y : -7
Z : -36
X : -79
Y : -5
Z : -33
X : -76
Y : -2
Z : -38
X : -79
etc.  (had to adjust the monitor speed to 57600 also).

About the "MagFieldVisualizer", i have downloaded and unzipped the files.
MagFieldVisualizer starts ok, but without changing anything in the "input", no movement is seen (by moving the mouse).
But the big arrow should probably be the result of the x, y, and z inputs combined to show the major direction i guess.

Guess we now need to use the serial monitor data (x, y, z) as input for your programm.
Using Com5 at the moment.

Thanks,   Itsu

Quote from: Itsu on 2018.12.30, 10:32:07
Hi Mags, Reiyuki,


Thanks for the interest.

I have made these changes to the code:

- Change Serial.begin(9600);  to  Serial.begin(57600);
- Change "Magnetic field in X-Axis:" to just "X : "      (likewise for Y and Z)
- Change delay(500);      to delay(20);


Now the serial monitor shows:

Z : -37
X : -82
Y : -3
Z : -34
X : -76
Y : -3
Z : -36
X : -74
Y : -7
Z : -36
X : -79
Y : -5
Z : -33
X : -76
Y : -2
Z : -38
X : -79
etc.  (had to adjust the monitor speed to 57600 also).

About the "MagFieldVisualizer", i have downloaded and unzipped the files.
MagFieldVisualizer starts ok, but without changing anything in the "input", no movement is seen (by moving the mouse).
But the big arrow should probably be the result of the x, y, and z inputs combined to show the major direction i guess.

Guess we now need to use the serial monitor data (x, y, z) as input for your programm.
Using Com5 at the moment.

Thanks,   Itsu

Thanks for testing (I hoped we'd get lucky on the first run).
Only thing now is to try having the arduino plugged-in before starting the program.

I'll make some sanity-checking and automation and tweaks if/when I get some time later today. ;)

I had/have the arduino plugged in and running when starting your program.

So do i understand it correctly that your program already should work?
No need to change any input?
I do not see any reference to the arduino or its serial monitor (where the data is) in your program.
What do i use to start your program?  Using MagFieldVisualizer.exe  now.

Anyway, take your time, i can test whenever you are ready  O0


Itsu

Interesting, very interesting. You could feed that data into excel, matlab, or Unity 3d for a 3d picture. Even that one sensor could display a dynamic point with a trace. Just leave the previous xyz vector plotted while plotting the next one and you would see a spirograph type picture.Post a file and lets have a look, shall we?

 (https://www.google.com/search?q=multidimensional+graph+in+excel&client=firefox-b-1&tbm=isch&source=iu&ictx=1&fir=8CfdCLLXitEUKM%253A%252C5Pt_auC66lyQdM%252C_&usg=AI4_-kR_nwnIsGoN1ALzbfusFibyI-zZ2w&sa=X&ved=2ahUKEwj7tOjxosnfAhUj0YMKHXacA1sQ9QEwF3oECAMQDA#imgrc=8CfdCLLXitEUKM:)https://www.google.com/search?q=multidimensional+graph+in+excel&client=firefox-b-1&tbm=isch&source=iu&ictx=1&fir=8CfdCLLXitEUKM%253A%252C5Pt_auC66lyQdM%252C_&usg=AI4_-kR_nwnIsGoN1ALzbfusFibyI-zZ2w&sa=X&ved=2ahUKEwj7tOjxosnfAhUj0YMKHXacA1sQ9QEwF3oECAMQDA#imgrc=8CfdCLLXitEUKM:
(https://www.google.com/search?q=multidimensional+graph+in+excel&client=firefox-b-1&tbm=isch&source=iu&ictx=1&fir=8CfdCLLXitEUKM%253A%252C5Pt_auC66lyQdM%252C_&usg=AI4_-kR_nwnIsGoN1ALzbfusFibyI-zZ2w&sa=X&ved=2ahUKEwj7tOjxosnfAhUj0YMKHXacA1sQ9QEwF3oECAMQDA#imgrc=8CfdCLLXitEUKM:)

 (https://www.google.com/search?q=multidimensional+graph+in+excel&client=firefox-b-1&tbm=isch&source=iu&ictx=1&fir=8CfdCLLXitEUKM%253A%252C5Pt_auC66lyQdM%252C_&usg=AI4_-kR_nwnIsGoN1ALzbfusFibyI-zZ2w&sa=X&ved=2ahUKEwj7tOjxosnfAhUj0YMKHXacA1sQ9QEwF3oECAMQDA#imgrc=8CfdCLLXitEUKM:)Once the display format is hammered out then an application can be generated. There is the resonant operating frequency of the coil and the read speed of the sensors that has to be determined up front so we can get a match for the display parameters.

 (https://www.google.com/search?q=multidimensional+graph+in+excel&client=firefox-b-1&tbm=isch&source=iu&ictx=1&fir=8CfdCLLXitEUKM%253A%252C5Pt_auC66lyQdM%252C_&usg=AI4_-kR_nwnIsGoN1ALzbfusFibyI-zZ2w&sa=X&ved=2ahUKEwj7tOjxosnfAhUj0YMKHXacA1sQ9QEwF3oECAMQDA#imgrc=8CfdCLLXitEUKM:)Only if we had the mag sensors on an LED type strip then we be cookin really good.

 (https://www.google.com/search?q=multidimensional+graph+in+excel&client=firefox-b-1&tbm=isch&source=iu&ictx=1&fir=8CfdCLLXitEUKM%253A%252C5Pt_auC66lyQdM%252C_&usg=AI4_-kR_nwnIsGoN1ALzbfusFibyI-zZ2w&sa=X&ved=2ahUKEwj7tOjxosnfAhUj0YMKHXacA1sQ9QEwF3oECAMQDA#imgrc=8CfdCLLXitEUKM:) I cant seem to get the line feeds or html code in my posts correctly. Dont know when this update happened but is sure messes up the posts. (https://www.google.com/search?q=multidimensional+graph+in+excel&client=firefox-b-1&tbm=isch&source=iu&ictx=1&fir=8CfdCLLXitEUKM%253A%252C5Pt_auC66lyQdM%252C_&usg=AI4_-kR_nwnIsGoN1ALzbfusFibyI-zZ2w&sa=X&ved=2ahUKEwj7tOjxosnfAhUj0YMKHXacA1sQ9QEwF3oECAMQDA#imgrc=8CfdCLLXitEUKM:)

https://www.sign-in-global.us/products/26281/encoder_strip_linear_encoder_board_sensor_for_roland_sj_540_sj_740_fj_540_sj_740_sc_540.html?ex=USD&utm_source=Googleshopping_us&utm_medium=Googleshopping&utm_campaign=Googleshopping&gclid=EAIaIQobChMI_PrO66bJ3wIVC4NpCh3qhQC7EAQYBCABEgK5NvD_BwE

 (https://www.sign-in-global.us/products/26281/encoder_strip_linear_encoder_board_sensor_for_roland_sj_540_sj_740_fj_540_sj_740_sc_540.html?ex=USD&utm_source=Googleshopping_us&utm_medium=Googleshopping&utm_campaign=Googleshopping&gclid=EAIaIQobChMI_PrO66bJ3wIVC4NpCh3qhQC7EAQYBCABEgK5NvD_BwE)https://ams.com/as5306 (https://ams.com/as5306)

Thanks GK,

i was looking to get this serial monitor data imported into EXCEL, and found "PLX-DAQ".

This seem to work as PLX-DAQ / EXCEL seems to "see" the data coming from my COM5 port (close the serial monitor first).

But it needs some control statements in that data to know what to do with it.

The format is something like:   "DATA,", DEC Val1, ",", DEC Val2, CR

But the trick is to seperate in the data stream these control statements from the axtual X, Y and Z data.

This website contains info and a link to PLX-DAQ:
https://robottini.altervista.org/arduino-and-real-time-charts-in-excel


Itsu

Great addition to Arduino and Excel!
Actually cant you spit out the Arduino data to the terminal in the x,y,z format you want then just scrape the terminal window for the file?
This way you have three columns to match the input of the excel graphing.
I mention this way as a precursor than going whole hog into developing an app so you don't get painted into a corner like you just found with the capture package you are using?
Just create your own csv format i.e. x,y,z data only. Get to the graphing as quick as possible to nix out any gotchas, as in minimal effort.
One can always make what works pretty but making something pretty work can be like marriage. :-* >:-) :'( LOL. I had to go there.

What am I doing wrong with my posts. I keep having to enter html codes to get line feeds, spacing and carriage returns? I feel like i am using a teletype on the internet. Is there a forum option to turn on formatting?
Testing format change. I think I found the formatting change. WYSIWYG option unchecked. Left this here for prosperity sake.

OK,  just changed the Arduino code to modify the Serial.print statements to print the X, Y and Z on one line and
incorperate the PLX-DAQ control statements:

// Output data to serial monitor
Serial.print("DATA, TIMER");
Serial.print(", ");
Serial.print(xMag);
Serial.print("   , ");
Serial.print(yMag);
Serial.print("   , ");
Serial.print(zMag);
Serial.println("  ,CR");

The output now (from PLX-DAQ Excel) looks like: 

TIMER           X-axis       Y-axis    Z-axis

90:33:45        -57,00      -111     30
99:00:00        -56,00      -107     29
99:11:15        -58,00      -107     25
106:52:30      -53,00      -104     30
111:28:07      -55,00      -108     30
111:28:07      -59,00      -107     28

Using a simple graph it shows the below kind of output (moving a magnet around the sensor):
My Excel does not have sophisticated 3D graps functions, but one can buy some 3D graps/plots add-on's like linked above by Giantkiller. 

Looking into that now.

Itsu

@Itsu, Sweet, dude! Oh yeah. Life is good!
Happy New Year!
Tis a grand time to be alive on this little rock in the sky!
Keep rollin...

So the next data point is the magnet sits at 0,0,0 then the plots will transverse around by their individual measurement data. And we can see the true spatial volumetrics of the field. Bam!
Of course it would be nice to have each side of the measurement cube with 64 sensors each. Ah yes, the future unfolds before our very eyes...
That equates to 64 times 6 which is 384 measurement points. And with the link that 3Dmagnetics posted is 384 * 6.25 = $2400.00. We need strips of these on a roll!
That number of sensors then needs an addressable bus assembly.

I repost link: https://www.aliexpress.com/item/CJMCU-90393-MLX90393-digital-3D-Holzer-sensor-displacement-angle-rotation-3D-position/32829326091.html?spm=2114.search0104.3.2.5f2db7357gnvm2&ws_ab_test=searchweb0_0,searchweb201602_4_10152_10151_10065_10344_10130_10068_10342_10547_10343_10340_10548_10341_10084_10083_10618_10307_10131_10132_10133_10313_10059_10534_100031_10103_10627_10626_10624_10623_10622_10621_10620,searchweb201603_37,ppcSwitch_3&algo_expid=b3ac6cd9-cf2e-4e98-9905-10947e761d7d-0&algo_pvid=b3ac6cd9-cf2e-4e98-9905-10947e761d7d&transAbTest=ae803_5&priceBeautifyAB=0 (https://www.aliexpress.com/item/CJMCU-90393-MLX90393-digital-3D-Holzer-sensor-displacement-angle-rotation-3D-position/32829326091.html?spm=2114.search0104.3.2.5f2db7357gnvm2&ws_ab_test=searchweb0_0,searchweb201602_4_10152_10151_10065_10344_10130_10068_10342_10547_10343_10340_10548_10341_10084_10083_10618_10307_10131_10132_10133_10313_10059_10534_100031_10103_10627_10626_10624_10623_10622_10621_10620,searchweb201603_37,ppcSwitch_3&algo_expid=b3ac6cd9-cf2e-4e98-9905-10947e761d7d-0&algo_pvid=b3ac6cd9-cf2e-4e98-9905-10947e761d7d&transAbTest=ae803_5&priceBeautifyAB=0)

And a mounting frame with variable spatial control. Well lookee here!: https://www.google.com/search?q=expandable+cage+ball&client=firefox-b-1&tbm=isch&source=iu&ictx=1&fir=4MBZuZ3QJkoD4M%253A%252CYMoZbvOxvuU18M%252C_&usg=AI4_-kRI9Gme8esAxOUs7mSvk1pSUwIPLA&sa=X&ved=2ahUKEwjkzsvshM3fAhVV_oMKHbgSAL4Q9QEwA3oECAUQCg#imgrc=4MBZuZ3QJkoD4M: (https://www.google.com/search?q=expandable+cage+ball&client=firefox-b-1&tbm=isch&source=iu&ictx=1&fir=4MBZuZ3QJkoD4M%253A%252CYMoZbvOxvuU18M%252C_&usg=AI4_-kRI9Gme8esAxOUs7mSvk1pSUwIPLA&sa=X&ved=2ahUKEwjkzsvshM3fAhVV_oMKHbgSAL4Q9QEwA3oECAUQCg#imgrc=4MBZuZ3QJkoD4M:)
This cage ball has to be what looks like 32 green axial connections. A good place to mount a sensor. Cuts cost down to $1200.00 in sensors. Just throwing out data here to track the development growth process.

The expandable area can then be used to balance amperage levels with resonance / impedance.
Now imagine if you will using this measurement system to track the nodes and antinodes of power reflections along a run of FE wire.
Sounds like harmonic destructive forces to me.

I smell a GOFUNDME project for saving the world.

To reduce costs even further the system can incorporate a rotation platform for either the test object or the sensor array.
A 360 degree single rotation is all that is needed as the platform axial rotation direction can be reversed and the data log can produce a Y axis directional data point also.
Now the sensor array can be reduced to 8 units which equates to ~$60.00.
Put a button switch on the Arduino caputer circuit and rotate the platform by hand incrementally. This would initially save the costs of incorporating a platform rotation drive system.
https://www.instructables.com/id/Desktop-CT-and-3D-Scanner-With-Arduino/?utm_source=newsletter&utm_medium=email (https://www.instructables.com/id/Desktop-CT-and-3D-Scanner-With-Arduino/?utm_source=newsletter&utm_medium=email)
This then defines the resolution. The speed of the sensor detection, the frequency of the field and the Y axis rotation speed / incremental steps.
With this approach the cage ball frame would not be necessary as the distance of the sensor and the test object can be increased or decreased by a simple act of placement.
The final model now has 8 sensors in a curved brace and a manual lazy Susan type platform. This is really cheap, like under $75.00!
This final model also negates the single turn platform. I am thinking of  the test object as a powered, wired device, like a coil.
The test object is spun to 359 degrees then rotated back to zero to start the incremental snapshot process all over again.
Each button press generates a new line of data.

Once you gain the appreciation of how fast and sensitive these devices can be made to be It will open the posibilities.

it is important to set the devices sensitivity to match the field intensity of interest  .
mine can measure a neomagnet influence  from 2 metres away

I used processing for the gui but there are python and a few other ones out there ...all abit of pain to configure .

real time instant display response with one chip is thre  first step and axis calibration can be done with the gui then inputting the numbers back to the arduino . Search for arduino magnetometer calibration .

It is very interesting and far far superior to using compasses which have inertia making them almost useless.


thanks for the interest guys . we need to be on the same page with programming environments for both the arduino and the
display.
Processing was what has worked for me  but there are so many versions .the earlier versions are often more compatible than the later ones I suggest arduino 1.5 to start with . and processing 2.2

Of interest and possibly of importance to this envevour.

https://en.wikipedia.org/wiki/Barkhausen_effect

and

https://pdfs.semanticscholar.org/06cc/5bec5a13f2dd333388f69a7c5ec58984ec5d.pdf

But lets not get ahead of things too much as having common configurations is of more importance at this stage .

If a few of us can have just one device spinning a 3d object using the same code ,that will be a good beginning .

Im under the impression that being able to see the field rotation in the centre of a device and from a few metres surrounding it will be
useful ... no essential to manipulating any effect created .

The  xyz values only  need to be changing relative to each other and don't saturate the sensor .
Its is one vey good reason to use the 90393 .



Cheaper sensors do work but everyone on the same page with this is important as there are a lot of internal programmable conditions
that are complex ,to me at least.

Its wonderful to have these devices and it is surprising that one has not been developed as it has other applications.

Thanks for the info guys, but i am afraid that this is as far as i can bring it.

My knowledge on how to visualize the magnetic field is almost zero, and i lack the energy to start a study on this.

Therefor my question earlier in this thread for knowledgeable people to step in.

I have the X, Y and Z outputs in either the arduino serial monitor or in an Excel spreadsheet and i am, up to
now, unable to locate a sophisticated 3D graph add-on for excel that does what i (we) want (show dynamically the
present magnetic field).

Some add-ons crash my Excel, some do not do what i want, and some are asking money.


So i will be looking for a simple solution, but i don't expect any breakthrough soon.


Thanks,   regards Itsu 

Quote from: Itsu on 2019.01.03, 10:44:15
Thanks for the info guys, but i am afraid that this is as far as i can bring it.

My knowledge on how to visualize the magnetic field is almost zero, and i lack the energy to start a study on this.

Therefor my question earlier in this thread for knowledgeable people to step in.

I have the X, Y and Z outputs in either the arduino serial monitor or in an Excel spreadsheet and i am, up to
now, unable to locate a sophisticated 3D graph add-on for excel that does what i (we) want (show dynamically the
present magnetic field).

Some add-ons crash my Excel, some do not do what i want, and some are asking money.


So i will be looking for a simple solution, but i don't expect any breakthrough soon.


Thanks,   regards Itsu

you will need to install processing ver 2.2
then run the bunny rotate code on processing while the arduino send the serial data,processing on your pc receives the xyz and rotates the object .

I hope this is helpful

Hi 3D,

no idea what "processing ver 2.2" is,  have to look that up.
Also "bunny rotate code" does not ring a bell.

thanks,  itsu

I am looking into Unity 3d. I have a number of years experience in this product.
The first analysis step is 'Is the data stream expedient to where this can be done in real time?'
I will wait on the Process v2.2 and the Bunny answer.
I don't want to throw any more shoes in the gears on the current push.
My additions in  this thread were to expand on the technological aspects while trying to deter any rising costs, in time and money.

I wait for the Process and Bunny answers as these seem to incur no further costs at this time and from 3Ds post might satisfy the demand of the next step.

Now imagine if you will a large sensor envelope the size of a room: https://www.instructables.com/id/Interactive-LED-Dome-With-Fadecandy-Processing-and/?utm_source=newsletter&utm_medium=email (https://www.instructables.com/id/Interactive-LED-Dome-With-Fadecandy-Processing-and/?utm_source=newsletter&utm_medium=email)
This uses Processing too:

QuoteProcessing

Processing is a visual programming language. Just like how you'd draw a rectangle in Photoshop using the mouse, you can draw a rectangle in Processing by writing code like this:

rect(100,80,10,50);

That would give you a rectangle starting 100 px in, 80 px down, 10 px wide and 50 px tall.

If you're not familiar with Processing, I greatly recommend Daniel Shiffman's tutorials on Youtube which are as entertaining as they are informative.

I also realize that posters can not just throw out answers hoping to get others to do the boots on the ground labor. But in my years of data analysis for the D.O.D.  I found that if everybody is allowed to throw the requests over the wall then a group can derive a lot of answers that can percolate to the forefront without any greater labor expenditures. Whatifs solve more problems than they create.

And the thought occurred to me based upon the mentioning of a handheld probe in prior posts. Using two sensors, one as a base location and the other in the hand could be used in the x,y settings of a scope to start a dynamic visualization process. This would let the tester see lissajous patterns and extrapolate them into an initial 3d spatial viewing platform. This would really drop the introductory pricing model. With what has been promoted here in this thread all that is needed another sensor($6.25), a DAC shield(if the current processor has none) and some code tweaks for the extra data. Conversion of the data in an DAC shield would permit direct connection of scope probes. This would then inject straight into FFT scope processes. Just trying to keep the headaches at a minimum and the process moving forward. And this point there is a usable test device.

Quote from: Itsu on 2019.01.03, 14:43:21
Hi 3D,

no idea what "processing ver 2.2" is,  have to look that up.
Also "bunny rotate code" does not ring a bell.

thanks,  itsu

https://processing.org/download/

its is one of many programming environments...but its not my strenght.

and a new find which may be of more use regarding the bunny ...or whatever object it may be .

https://www.electronicdesign.com/industrial-automation/rapid-prototyping-era-arduino-mikrobus-and-processing

this example used a different sensor but they all can be programmed to output x,y,z , ...h,p,b, or whatever variable you chose.

There is a learning curve which in my case has been interupted by a chronic health condition.

Perhaps there is a better display program ? I know processing can be awkward but as long as you have the right libs in the right place , the errors are fixable. The trouble happens with new windoes and file permissions at odd times but again there are forums like this that address those issues.

The simple 3d object display will immediately make clear that thiis is far superior to any scope display as we are only looking for rotation however weak.

please get to first base with a 3d object ...it will open your mind to the possibilities. I am sure of that.

with most of this stuff a bit of goggling helps and as to the visual display there are a few processing,python,ans any others like blender ...but they all are library dependant and version issues v library issues come up . One of the reasons that I am asking for "boots on the ground" here .



It may not be aproptiate as i had not considered that ,,my bad i guess

3D

Thanks 3D, 

so processing is some arduino programming clone which uses also arduino hardware.
I need to look into that further.

Itsu

Meanwhile as mentioned in his post #38, member Reiyuki "whipped together a quick demo" using Visual Basic.
His version 2 came alive and shows in real time the movement of an arrow representing the 3 axis (X, Y and Z)
of the MLX90393 sensor.

Still some improvements to be done, but his general idea is clear and seems to work.

Video here:  https://www.youtube.com/watch?v=Frj3ymDJdOc

Idea's, suggestions?


Itsu

Quote from: Itsu on 2019.01.04, 14:34:06
Meanwhile as mentioned in his post #38, member Reiyuki "whipped together a quick demo" using Visual Basic.
His version 2 came alive and shows in real time the movement of an arrow representing the 3 axis (X, Y and Z)
of the MLX90393 sensor.

Still some improvements to be done, but his general idea is clear and seems to work.

Video here:  https://www.youtube.com/watch?v=Frj3ymDJdOc

Idea's, suggestions?

I  apolagise  my brain fog had me miss that section entirely ..
Reiyuk ,that works well and is agood beginning . Thank you I will be back in my nexr well phase.


Itsu

I was trying to calibrate the MLX90393 sensor via this website:

https://learn.adafruit.com/ahrs-for-adafruits-9-dof-10-dof-breakout/magnetometer-calibration

like suggested by 3D in his post #4, but did not got it to run.


Even after adding some libraries like "NXPMotionSense" and "elapsMillis" the sketch won't compile/run.

Guess there is no compatible sensor there like the MLX90393 we use.


Itsu

Quote from: Itsu on 2019.01.06, 13:46:48

Guess there is no compatible sensor there like the MLX90393 we use.

Itsu

As we say in the industry: 'I guess we dodged that bullet!' Or 'stopped just shy of the brick wall!'
Mostly free equipment costs, some labor but no huge expenditure into a dead end project based upon a missing library.
Writing libs can be done but that is a lot of closed door, closet time in the bright light of the screen. Ya better have your ducks in a row before you jump and close the door.
When you come out, the world might have changed without ya.
https://stackoverflow.com/ (https://stackoverflow.com/) is a good place to post questions. The programmers will shoot out of the walls with the right tag lines in the posts.
I have asked some pretty esoteric questions based on a large variety of technical issues and received actual fully coded demos in C, Unity, Pic, Arduino and VB.
The people there are always willing to help.

At the moment I am involved/haunted with two other projects and can't tear loose.
I am not a believer in throwing half arsed solutions over the wall and running away.
Further analysis always can save time and money. We might be just a little ahead of the curve of normality here. But this solution will rear up again in the future if not soon based upon more searching.

I googled MLX90393 coding and plethora of hits came up.
For the next step I am wondering: https://www.google.com/search?client=firefox-b-1&q=mlx90393+arduino+code (https://www.google.com/search?client=firefox-b-1&q=mlx90393+arduino+code)

Hi GK,

i get what you mean, but before calling in the big guns i would like to give the good people on this forum
the chance to point me in the right direction by telling this programming nitwit Itsu to look there and there
and/or to do this and that first.

I mean those big guns would expect at least that one has done their homework before asking.

thanks anyway.

Itsu

@ITsu,
They are are gracious there. Just tell 'em your a noob.
Some of those guns aren't so big.

Gracious?

well,  i tried to sign on there, but that is impossible as they demand to know everything about your
work, your work responabilities, your history etc. etc.

As a simple responsableless retired person you are unable to sign up.


Ok,  Ok,  i finally got through, but now i first need "1500 reputations" before i can ask a question  :o<


Well,  nice try



Itsu