Ideas to tinker

[Tinkerer]

What if?

It happens now and then, that we come across an idea that seems new.
What if we tried it this way or that way, would it work? would it be better?
Sometimes we give it a try. We progress to a certain point and then we meet a snag, or even a brick wall. Then we give up.
Often years later that idea comes back. In the meantime we have learned more. New and better components might make it possible to make that old idea work.

This thread it to bring up new and older ideas.
Look at the merits of the ideas and how to make them work or improve.

Not for beginners.

For people who have some understanding of at least part of the functioning of a metal detector.

So let's start with the TEM method.

Eclipse, You gave it a try. What problems did you find?

Can we find solutions for these problems?

[eclipse]

Awesome!

Lets start with this simulation file if you like it (see the attached files).

This steers away from the co-planar coil and the reverse TX coil controlled with a P-ch mosfet but I think it's the one thing it got people scared of.

It's with a regular DD coil with TX coil 300uH. The RX is made with bonded bifillar wire used to create center tapped RX (bought 0.25mm wire of this type so it's actually available for purchase and you don't have to built it yourself). The preamp is THAT1512 - for ease.
Protection diodes in the RX my observations where not needed (KISS).




TEM_0.1.zip

Next step would be subtract the RX (the method you describe in your topic called Breakthrough!!!!! Full discrimination with PI) but lets start with with basics.
Thoughts?

[Tinkerer]

Excellent.

Are you satisfied with the result?

[Tinkerer]

Awesome!

Lets start with this simulation file if you like it (see the attached files).

This steers away from the co-planar coil and the reverse TX coil controlled with a P-ch mosfet but I think it's the one thing it got people scared of.

It's with a regular DD coil with TX coil 300uH. The RX is made with bonded bifillar wire used to create center tapped RX (bought 0.25mm wire of this type so it's actually available for purchase and you don't have to built it yourself). The preamp is THAT1512 - for ease.
Protection diodes in the RX my observations where not needed (KISS).


[ATTACH]50891[/ATTACH]

[ATTACH]50893[/ATTACH]

Next step would be subtract the RX (the method you describe in your topic called Breakthrough!!!!! Full discrimination with PI) but lets start with with basics.
Thoughts?

eclipse,

I have taken the liberty of changing your simulation to fit better for the small nuggets that you are looking for.

the TX L2 you had is the bucking coil for a concentric induction balanced RX coil. You don't need it for the DD coil.

I also added 3 targets, 1us, 5us, 10us, TC's, to show the target reaction on the TX pulse.

[eclipse]

The coil (47uH) is inductor for not allowing high current flow back to the battery supply or other circuit sections as per Aziz TEM designs (aka magic choke).


In my version 220nF cap reduces the flyback to ~100V and the RX sees a lot less so protection diodes were not needed.
But with lower tuning capacitor the voltage goes up and protection diodes will be needed.
You've changed the op amp with one that's rated for up to +-35V on the inputs (and the simulation runs a lot faster than before)


I've built the tx/preamp section with real components and it worked (my simulation).
I've picked this preamp because it's ready for Bi-Polar TX without modifications needed.


I've experimented a bit with the next thing I wanted to do but gotten not very far - figure out how to filter the TX voltage
in a way that it will work for wide range of TX flyback 100-250V for example, with minimum adjustment needed so that it can be used to subtract the RX
( as per your post a while back called Breakthrough!!!!! Full discrimination with PI).
I don't know if you have made further experiments of this but this was the direction I wanted to dive into, I think this topic has been overlooked.

[Tinkerer]

The coil (47uH) is inductor for not allowing high current flow back to the battery supply or other circuit sections as per Aziz TEM designs (aka magic choke).


In my version 220nF cap reduces the flyback to ~100V and the RX sees a lot less so protection diodes were not needed.
But with lower tuning capacitor the voltage goes up and protection diodes will be needed.
You've changed the op amp with one that's rated for up to +-35V on the inputs (and the simulation runs a lot faster than before)


I've built the tx/preamp section with real components and it worked (my simulation).
I've picked this preamp because it's ready for Bi-Polar TX without modifications needed.


I've experimented a bit with the next thing I wanted to do but gotten not very far - figure out how to filter the TX voltage
in a way that it will work for wide range of TX flyback 100-250V for example, with minimum adjustment needed so that it can be used to subtract the RX
( as per your post a while back called Breakthrough!!!!! Full discrimination with PI).
I don't know if you have made further experiments of this but this was the direction I wanted to dive into, I think this topic has been overlooked.

This is basically a flexible PI-VLF hybrid.

You can demodulate like a PI
You can demodulate like a VLF
You can demodulate like a PI and a VLF at the same time

You can also make the circuit behave more like a PI or more like a VLF.

There are many, many ways to change and adapt this basic circuit to the various uses of detectors.

My choice of Instrumentation Amplifier was just a lazy choice to find something to fit your circuit, because I could not remember how to make the LTSpice see the THAT that you used.

[green]

This is basically a flexible PI-VLF hybrid.

You can demodulate like a PI
You can demodulate like a VLF
You can demodulate like a PI and a VLF at the same time

You can also make the circuit behave more like a PI or more like a VLF.

There are many, many ways to change and adapt this basic circuit to the various uses of detectors.

My choice of Instrumentation Amplifier was just a lazy choice to find something to fit your circuit, because I could not remember how to make the LTSpice see the THAT that you used.

I was thinking to act like a PI Tx current needs to be constant(not changing). Tx current your reply#4 looks like a saw tooth. Is my thinking wrong?

[Tinkerer]

I was thinking to act like a PI Tx current needs to be constant(not changing). Tx current your reply#4 looks like a saw tooth. Is my thinking wrong?

thank you for your input.

What we traditionally call a TX pulse in PI, is about similar to a sawtooth. When we pulse the PI ending with a constant current, we call it Flattop PI pulse. Most of the denominations we use are just common use, there are no exact rules that I know of.

After the TX pulse we have a "listening period" where we traditionally like the coil current to be 0.

With the circuit above, being a hybrid, it is never a 100% PI nor a 100% VLF. It is always a bit of both. The good thing is that we can change it to be more of the one, or the other.

I used to call one variation I used as 2 TAU, ie. a fast ramp and a slow ramp. The excitation of the targets being more ideal for short TC targets and long TC targets respectively.

The Tinkerer TEM_02 simulation shows a mostly PI response to the target.

[eclipse]

Ideas on how to modify this to become a bipolar TX?
I've made it with a simulation but it requires bifilar TX coil wire and another transistor - the coil wire requirement is a bit constraining.
H-bridge solution didn't seems to work (on a working h-bridge circuit I swapped the damping resistor with capacitor and result was ringing all over the place).
I can post these 2 circuit if it would help.

[6666]

Ideas on how to modify this to become a bipolar TX?
I've made it with a simulation but it requires bifilar TX coil wire and another transistor - the coil wire requirement is a bit constraining.
H-bridge solution didn't seems to work (on a working h-bridge circuit I swapped the damping resistor with capacitor and result was ringing all over the place).
I can post these 2 circuit if it would help.

Could you post them please.

[eclipse]

Here's the bipolar version with dual TX wire coil & 2 mosfets.
Note the lower coupling factor of TX/TX2. It's not tested in real life.

TEM_0.3.zip

[eclipse]

Here's an experimental circuit. What do you think ?



experiment.zip


And the bipolar circuit.
It can be made to become the THS.
Can it be made to be TEM ?

[eclipse]

Bipolar TSW with the bipolar schematic posted above after playing some of the values

[eclipse]

just figured why not do this - we could have both (multi-frequency) type deal (half sine + half truncated sine)

maybe focus on processing more .. rx/preamp sampling etc.

[Tinkerer]

Getting interesting