Any circuit sims out there that can do metal detectors?

[brucester]

Hi,

I have been teaching myself electronics and was wondering if anyone knew of a circuit simulator that included magnetic induction from coils in a circuit?

For example, a piece of software that can simulate a metal detector for trying out circuit designs.

Thanks

[Davor]

You can use LTspice as I do and simulate the most of the circuits. Magnetic coupling is expressed by coupling "K". Non-ferromagnetic targets are expressed as a parallel combination of inductance and resistance to mimic target eddy current response. For example, you can use 150uH in parallel with 1ohm to mimic 150us response of copper.

[Qiaozhi]

Hi,

I have been teaching myself electronics and was wondering if anyone knew of a circuit simulator that included magnetic induction from coils in a circuit?

For example, a piece of software that can simulate a metal detector for trying out circuit designs.

Thanks

As Davor said, you can use LTSpice, which has the major advantage that it's FREE to download and use.

In the book "Inside the METAL DETECTOR" I used SPICE to simulate the Raptor design, although the details of how this was actually achieved were beyond the scope of the book. Chapter 9 (Fig. 9-15) shows the complete response of the GEB and DISC channels. Since it is quite difficult to actually simulate the targets themselves, I used an idealized target response as a stimulus for each channel. The correlation between simulation and physical reality was close enough to prove that the design would work in practice.

[6666]

Here is a little trick
get a copy of Ltspice
get it running
it makes schematics that end in .asc
now in the search bar at top of forum , search on .asc
that will throw up some threads that contain .asc
have a browse through those threads
for sims

[brucester]

cool, thanks guys

will have to read your book

[Qiaozhi]

cool, thanks guys

will have to read your book

The link is here ->
http://www.geotech1.com/forums/conte...metal-detector

[Tinkerer]

Here is a little trick
get a copy of Ltspice
get it running
it makes schematics that end in .asc
now in the search bar at top of forum , search on .asc
that will throw up some threads that contain .asc
have a browse through those threads
for sims

There is just one little problem with that: the .asc files do not upload. Therefore one usually uploads the files as .zip files and the screen shots as .png or .jpeg

I uploaded many LTSpice files that show the circuits with the target responses, but they are spread all over the forum.

Tinkerer

[6666]

Therefore one usually uploads the files as .zip files

Yes thats right I remembered I should have added that after the edit time of one hour expired
but point is if you search on .asc
it will take some reading of the threads ,
quite often you will see a comment saying that the .asc file is attached
as a zip or rar etc

[Tinkerer]

You can use LTspice as I do and simulate the most of the circuits. Magnetic coupling is expressed by coupling "K". Non-ferromagnetic targets are expressed as a parallel combination of inductance and resistance to mimic target eddy current response. For example, you can use 150uH in parallel with 1ohm to mimic 150us response of copper.

The RX signal is the sum of exponentials. That is, the decay curve of the coil, the decay curve of the ground and the decay curve of the target. Some targets are of resistive nature, like gold and other targets are of reactive nature, like ferrite. Iron is of mixed nature, meaning the response is resistive and reactive at the same time, but not by the same amplitude. I have posted sims that show these responses as well as the sum of these responses. These sims were designed to be as close to the observed real component and target responses as I could manage, but there are still a lot of improvements to the models possible.

Tinkerer

[Davor]

Could you seek those Fe target sims and place them in a topic http://www.geotech1.com/forums/showt...for-simulation ?

Thanks

[Tinkerer]

Could you seek those Fe target sims and place them in a topic http://www.geotech1.com/forums/showt...for-simulation ?

Thanks

It is an excellent idea to make a thread dedicated to target sims.
Looking at your target models, I see that you are modelling for a target response to a sine wave excitation. I am used to work with transient excitation. Is the response the same? Intuitively I would say NO.

Therefore we must clearly differentiate between sine wave simulations and transient simulations.

There is a lot to be learned with target simulations and if we perfect the target simulations, it will give insight that may help us design better detectors.

Tinkerer

[Davor]

Actually the same target sim will function equally well with both PI and VLF IB systems if it is correctly modelled. My Fe target is not completely correct since it should work well as a 2-port device, and I was able to place it's response into a correct quadrant only as a 4-port device.

You need a 2-port Fe model target with a coil to accommodate for a mono coil simulation. That's why I'm so interested in your Fe target sim. Otherwise a 4 port is just fine. My model is even capable of long term magnetisation, but only as a 4-port. When I use it as a 2 port it behaves as a non-Fe.

[Tinkerer]

Actually the same target sim will function equally well with both PI and VLF IB systems if it is correctly modelled. My Fe target is not completely correct since it should work well as a 2-port device, and I was able to place it's response into a correct quadrant only as a 4-port device.

You need a 2-port Fe model target with a coil to accommodate for a mono coil simulation. That's why I'm so interested in your Fe target sim. Otherwise a 4 port is just fine. My model is even capable of long term magnetisation, but only as a 4-port. When I use it as a 2 port it behaves as a non-Fe.

I am afraid that my simulations are not as good as you expect. Maybe with your help we can improve them.

I observed the target responses on the oscilloscope, with real coils and real targets and then try to obtain similar results with simulations.

For iron I use a R+X response. I noticed that the relationship between R+X change with different types of steel and with different shapes of steel targets. A thin flat steel disk will show more X response when presented vertical to the coil. At a certain angle the X+R just about cancel and when flat the R has higher amplitude.

I also noticed with real targets, that often the TC of the X response is shorter than the TC of the R response. I try to repeat that with the simulation, using TC, but the in fact I think we should use the B curve. Using the B curve, is still a guess, because we have no way of knowing what kind of steel target is in the ground.
Common transformer steel contains a certain amount of silicon, silicon steel is designed to be "soft" in magnetization, but this is considering 50 or 60Hz.
Ferrites are usually designed for high frequency, so the "TC" is very short.

As you see, I use a simplified view, but in the absence of anything better, it gives a reasonable approximation.

I posted a sim on the thread: http://www.geotech1.com/forums/showt...for-simulation

Tinkerer

[Davor]

Thanks, I seen them.
With Fe we have 2 effects present at the same time, magnetisation and eddy currents. Such target should respond for a VLF excitation by a 2nd quadrant response. I havee no first hand experience with eddies in iron/steel so my model is simply responding in a proper quadrant - nothing more.
The most interesting would be modelling the magnetisation, which is very different in various ferromagnetic materials. So in fact the model should accommodate for two constants. Mine works if you couple Tx to target coil#1 and Rx to target coil#2.

I've done some preliminary tests with such target in PI system and obtained some very similar responses to the real world ones - qualitatively. One of my conclusions for the combination of a Fe and non-Fe targets was that the PI discrimination system with sampling in several points is a bust - the same ski-like response is obtainable with a combination of two non-Fe targets, and the three points sample interpretation is as reliable as tarot.

Zero delay sample behaviour is entirely a different story.