Ivconic Magnetometer/Gradiometer

[ivconic]

Dont worry, 3D printer is not so scarry, because it can not grab your arm/leg - while you do not pay enough attention - and throw/press you by 80cm/s like bigger portal 3D mill can do without notice

What a coincidence!
You are Czech, and a friend who is having "Deckel Maho" was living and working in Brno exactly with those large beasts in main company there.
Later on he bought several large machines and moved here to run small private enterprise.
I know Czechs are very good in machining and mechanics! As he told me.

[sled]

Hey ivconic,

did you build a calibration coil to approximate the field strength / frequency relationship? I’m building something similar for hobby purpose, still have some original FGM-3 sensors from Speake which I have characterized a long time ago. Now I want to order the new sensors, however I discovered they have a non-linear behaviour, see here: https://docs.wixstatic.com/ugd/7034a...8c8c49afd0.pdf

what is your experience with the new sensors?

[ivconic]

Hey ivconic,
did you build a calibration coil to approximate the field strength / frequency relationship? I’m building something similar for hobby purpose, still have some original FGM-3 sensors from Speake which I have characterized a long time ago. Now I want to order the new sensors, however I discovered they have a non-linear behaviour, see here: https://docs.wixstatic.com/ugd/7034a...8c8c49afd0.pdf
what is your experience with the new sensors?

No i did not.
But thanks a lot for pointing that out!
Thanks for the link too!

"...still have some original FGM-3 sensors from Speake which I have characterized a long time ago..."

Can you tell here more details ?

[ivconic]

"...The output signal frequency transformed to analog value through LM2917 frequencyto-voltage converter..."

Could it be this (the conversion) to additionally affects the linearity?
I can only assume and be wrong, of course, since there is no schematic presented in the document.

[Hyena]

What a coincidence!
You are Czech, and a friend who is having "Deckel Maho" was living and working in Brno exactly with those large beasts in main company there.
Later on he bought several large machines and moved here to run small private enterprise.
I know Czechs are very good in machining and mechanics! As he told me.

Really? Glad that somebody has luck and effort for sure to make his own bussiness! Fingers crossed to keep it that way. I tried my own company, but lazy salesman destroyed it in the beggining :-/
Then I moved from machine to CAD. Less mess but more stress Also less fun. If you will ever need any help with that virtual 3D rubbish, just PM me. I can also mill, but I am sure you have somebody closer

[ivconic]

Really? Glad that somebody has luck and effort for sure to make his own bussiness! Fingers crossed to keep it that way. I tried my own company, but lazy salesman destroyed it in the beggining :-/
Then I moved from machine to CAD. Less mess but more stress Also less fun. If you will ever need any help with that virtual 3D rubbish, just PM me. I can also mill, but I am sure you have somebody closer

I tried to use Autodesk Fusion 360 for a skim works. But it looks too complicated for a fast drawing so i left it for this winter to become more familiar with it.
Than i stumble across the Tinkercad and it is doing the job along with Freecad and 3DToool. Mixing all together and i get what i want.
Of course it is far from any professional use, but it is quite alright for simple and skim tasks.
Recently i got Corel CAD 2018. but hadn't time to check it out.
Will have to delay it for winter.

[sled]

"...The output signal frequency transformed to analog value through LM2917 frequencyto-voltage converter..."

Could it be this (the conversion) to additionally affects the linearity?
I can only assume and be wrong, of course, since there is no schematic presented in the document.

Good point, I will check the datasheet. We calibrated and characterized the original Speake sensor, as far as I remember the linearity was okay, we used a frequency counter. However we noticed a temperature drift and added some styrofoam insulation around the sensors.

Do you use a gradiometer setup or single sensor magnetometer?

PS: We used a Helmholtz coil for calibration due to the better uniformity of the field. You can adjust it so it cancels the Earth’s magnetic field and compare that to known published values for your location.

PPS: It seems the author of the PDF did the initial measurements without Freq. to Voltage conversion, by directly measuring frequency on the oscilloscope, so I think the non-linearity is inherent to the sensor.

[Hyena]

I use autoCAD 2014 for fast 2D sketches. I do not like their new ribbon design and I always switch to classic old enviroment. I started with SolidWorks, its user friendly and able to handle even little bit complicated designs while keeping it reasonably parametric, also has build in CAM modules for direct machining. Then I moved to Catia, because of automotive But it is far from user friendly... still can handle most complicated models I have ever worked with.
Looking forward to finished design

[ivconic]

Good point, I will check the datasheet. We calibrated and characterized the original Speake sensor, as far as I remember the linearity was okay, we used a frequency counter. However we noticed a temperature drift and added some styrofoam insulation around the sensors.

Do you use a gradiometer setup or single sensor magnetometer?

PS: We used a Helmholtz coil for calibration due to the better uniformity of the field. You can adjust it so it cancels the Earth’s magnetic field and compare that to known published values for your location.

Both versions.
I finished both "basic" versions.
Hopefully those will be available as kits very soon.
Now i am adding more functionalities to prepare more advanced version.

[ivconic]

I use autoCAD 2014 for fast 2D sketches. I do not like their new ribbon design and I always switch to classic old enviroment. I started with SolidWorks, its user friendly and able to handle even little bit complicated designs while keeping it reasonably parametric, also has build in CAM modules for direct machining. Then I moved to Catia, because of automotive But it is far from user friendly... still can handle most complicated models I have ever worked with.
Looking forward to finished design

Main reason i gave up from Fusion 360 is because it is choking my system much with online crap.
Hopefully it is not because of the system resources but it is because of the sudden madness of everybody pushing the online and "server dependable" software.
Huge BS!
I am already having MS Visual Studio, full Corel and Atmel Studio, latest updates, running pretty smooth on the same system. And only Autodesk software is choking it, how come?
They don't sell you software anymore; instead they only try to lease it for a limited time, so it seems.
I am not that rich to buy new computer on every few months just to please their madness.

[Qiaozhi]

The first thing you should print is Benchy - the 3D torture test -> http://www.3dbenchy.com/about/
It's a small test piece that will show the capabilities (or otherwsie) of your 3D printer, and confirm that you have everything set up correctly.

[ivconic]

The first thing you should print is Benchy - the 3D torture test -> http://www.3dbenchy.com/about/
It's a small test piece that will show the capabilities (or otherwsie) of your 3D printer, and confirm that you have everything set up correctly.

I will do that as soon as i receive material. Ordered few hours ago.
Only few meters of plastic filament with debatable quality came with printer.
Probably testing piece. I already wasted it on few small items.
I am getting "hairy" results with 100% speed.
Last item printed with 80% of speed is giving much better result.
I ordered better quality of PLA 1.75mm filament, two colors, 1kg each.
Looking forward to see the difference.
Things are bit more complex than i expected. It has heating on bed too.
I must experiment with adjusting temperature on nozzle and bed separately, to find best combination for given filament type.
Supporting PC software is having much more options to adjust than if printer is used with SD card as standalone.
But i tested it today only with SD card as standalone.
In few days, with proper material; i will test it more seriously.

[Qiaozhi]

I will do that as soon as i receive material. Ordered few hours ago.
Only few meters of plastic filament with debatable quality came with printer.
Probably testing piece. I already wasted it on few small items.
I am getting "hairy" results with 100% speed.
Last item printed with 80% of speed is giving much better result.
I ordered better quality of PLA 1.75mm filament, two colors, 1kg each.
Looking forward to see the difference.
Things are bit more complex than i expected. It has heating on bed too.
I must experiment with adjusting temperature on nozzle and bed separately, to find best combination for given filament type.
Supporting PC software is having much more options to adjust than if printer is used with SD card as standalone.
But i tested it today only with SD card as standalone.
In few days, with proper material; i will test it more seriously.

Be prepared to sacrifice quite a lot of filament during experimentation. Getting the nozzle and bed temperatures correct is important, and particularly the gap between the nozzle and bed. What surprised me was how little filament is actually used to create a model, and how long it takes to print. Printing from SD card is the most reliable.
Repetier-Host software can be used to slice the stl file, and shows how long the printing will take and how much filament is required. You can also adjust the printer settings from there.

[ivconic]

Be prepared to sacrifice quite a lot of filament during experimentation. Getting the nozzle and bed temperatures correct is important, and particularly the gap between the nozzle and bed. What surprised me was how little filament is actually used to create a model, and how long it takes to print. Printing from SD card is the most reliable.
Repetier-Host software can be used to slice the stl file, and shows how long the printing will take and how much filament is required. You can also adjust the printer settings from there.

True, i was also surprised to see how little filament is used.
When i saw that piece of filament they packed with printer; i thought it will not be enough not even for a simple test.
Instead; i printed 4 small items yesterday, changing adjustments to see the differences.
Gap between the nozzle and bed is really important. I must readjust it, because i left it bit wider. I did what people suggest on Youtube. But i think it can be done better.
Software that comes with is named "Cura". It comes on a SD card with printer.
I installed it and on first start it looks up for upgrade and offers me new version. So i updated it.
Funny thing is that version from SD card carries "15.04.2" version number and update is "3.4.1" !?? Defies any logic?
Yet i kept both, since those looks quite different.
Here are the pics from last item. Not finished because there was no more filament.

[ivconic]

Until filaments arrive; i am "studying" the details on advance printing.
And more i realize that this is very serious story!
In meantime i installed "Simplify 3D" software and found it is very helpful in preparing the gcode for printer.
Cura is also alright. Having all the features.
But it is good to have "second opinion" from another software too.
Right now i am exploring several ways to add "supporters" ("helpers") to the design.
Bottom line; i had to correct my drawings, for which i though are ok, and i was wrong!

Still much to learn before start wasting material.