Fast coil idea

[ivconic]

From induction cooker.
0.4 ohms.
0.07mH (70uH).
This would be indeed fast coil.
How to drive it properly?

[Elliot]

[ATTACH]50517[/ATTACH]

From induction cooker.
0.4 ohms.
0.07mH (70uH).
This would be indeed fast coil.
How to drive it properly?

How to drive it properly? Put it back in the cooktop I'd suggest.

How much faster than 5uS do you want? Just curious.

[ivconic]

Experimenting...

[ivconic]

I am not familiar with igbt's so i guess i'll had to read some articles.
Here is possible driving circuit.

Problem is; igbt will not tolerate common impulse that we can see on many PI designs.
Maybe all this is wrong.
Maybe i should try to use this coil only for Rx.
But what's better way to learn than to burn to ashes few expensive components?!

[ivconic]

[Tinkerer]

How to drive it properly? Put it back in the cooktop I'd suggest.

How much faster than 5uS do you want? Just curious.

As you say, it is a matter to drive properly. And then, of course is still the damping.

Attached is a scope picture of a "fast, slow coil".
As you can see, it is perfectly damped at about 2us.
The red trace is the Mosfet gate pulse OFF.
The blue trace is the Flyback, about 760V

The diameter is 200mm

Here is the challenge: Who dares to guess the answers?

How many uH does the coil have?
How many turns does the coil have?
What is the DC resistance?

Hint, this is no ordinary PI. I would call it advanced.

[bbsailor]

For those who tinker (pun intended) with coils, here are a few thoughts to consider when making "fast coils".
1. A fully stimulated target needs a coil discharge time constant (TC) five times faster than the targets TC. The coil discharge TC is based on the coil inductance divided by the damping resistor value and any other resistors such as the op amp input resistor being grounded when the clamping diodes are on.
2. A fully charged target will loose its eddy current charge in five target TCs. That means that the delay must be short enough to capture some target eddy currents before they die out.
3. Once delays get below10 uS in the range of 5 to 10 uS the characteristics of the coil wire and coil design will limit how low you can get the delay to go. Coil wire may retain some eddy currents beyond a low delay setting.
4. Total capacitance of the coil includes the wire turn to turn capacitance and is affected by the wire insulation thickness and insulation type due to its dielectric constant. Adding a diode in series with the MOSFET and coil minimizes the MOSFET COSS as seen by the coil. The coil shield will also add capacitance to the coil and this is controlled by the coil shield area, spacer from the coil windings and dielectric constant of the coil spacer separating the coil and shield. The coax that connects the coil to the circuit will add about 30 pf per foot. With fast coils you want a creative way to make the coil to coax joint far enough away from the coils sensitive area so this joint is not seen as a target.

Here is how to visualize the ripple that you see when observing the coil damping. More capacitance as seen by the coil represents the energy in the ripples that needs to be damped with lower damping resistor values thus changing the coils discharge TC. So, as you can see, making fast coils requires making many tradeoffs, especially when you get the delays very low.

Here is an example. A 300 uh coil trying to detect a 2 uS target has these characteristics.

The optimum coil discharge TC to fully stimulate the target is 2/5 or .4 uS and would need the damping resistor to be 750 ohms (300/750 is .4). If the delay is set at 10us, the stimulated target eddy current energy will be zero at this delay so a delay of 7 or 8 us would be needed to detect this target. As targets get smaller, tinkering with coil designs need to keep these things in mind (no pun intended).

Joseph J. Rogowski

[green]

Interesting trace. Usually the first half of the decay would be faster than the last half. Only way I can get a trace that looks similar, is to underdamp and let the return to zero go to -.6V across the MOSFET. Would be hard to see with 200V/div. Wondering how you get the mirror image critical damped.

[Altra]

[ATTACH]50517[/ATTACH]

From induction cooker.
0.4 ohms.
0.07mH (70uH).
This would be indeed fast coil.
How to drive it properly?

Not really metal detector related but search "ZVS induction heater circuit" . These circuits are self oscillating. You can heat/melt ferrous metals. With a graphite crucible you can melt gold, silver, copper. Takes a lot of current.

https://www.youtube.com/watch?v=EujaZ_mNt7g

[Tinkerer]

For those who tinker (pun intended) with coils, here are a few thoughts to consider when making "fast coils".
1. A fully stimulated target needs a coil discharge time constant (TC) five times faster than the targets TC. The coil discharge TC is based on the coil inductance divided by the damping resistor value and any other resistors such as the op amp input resistor being grounded when the clamping diodes are on.
2. A fully charged target will loose its eddy current charge in five target TCs. That means that the delay must be short enough to capture some target eddy currents before they die out.
3. Once delays get below10 uS in the range of 5 to 10 uS the characteristics of the coil wire and coil design will limit how low you can get the delay to go. Coil wire may retain some eddy currents beyond a low delay setting.
4. Total capacitance of the coil includes the wire turn to turn capacitance and is affected by the wire insulation thickness and insulation type due to its dielectric constant. Adding a diode in series with the MOSFET and coil minimizes the MOSFET COSS as seen by the coil. The coil shield will also add capacitance to the coil and this is controlled by the coil shield area, spacer from the coil windings and dielectric constant of the coil spacer separating the coil and shield. The coax that connects the coil to the circuit will add about 30 pf per foot. With fast coils you want a creative way to make the coil to coax joint far enough away from the coils sensitive area so this joint is not seen as a target.

Here is how to visualize the ripple that you see when observing the coil damping. More capacitance as seen by the coil represents the energy in the ripples that needs to be damped with lower damping resistor values thus changing the coils discharge TC. So, as you can see, making fast coils requires making many tradeoffs, especially when you get the delays very low.

Here is an example. A 300 uh coil trying to detect a 2 uS target has these characteristics.

The optimum coil discharge TC to fully stimulate the target is 2/5 or .4 uS and would need the damping resistor to be 750 ohms (300/750 is .4). If the delay is set at 10us, the stimulated target eddy current energy will be zero at this delay so a delay of 7 or 8 us would be needed to detect this target. As targets get smaller, tinkering with coil designs need to keep these things in mind (no pun intended).

Joseph J. Rogowski

bbsailor,

thank you for the thorough explanation of the classical PI coil considerations. For many years this was my guideline.
Then I thought: what if we would try to think "outside the box"? Could there be a different way of doing it? Could there be a different way of looking at it?
If I lie under the table and look up, I still see the same table as I saw from above, but it looks totally different.
So this is what I tried to do. DIFFERENT. It ended by being as different as the table from underneath. Different in every way.
And it works.

The boss said: Why didn't somebody think of that before?

[ivconic]

bbsailor,

thank you for the thorough explanation of the classical PI coil considerations. For many years this was my guideline.
Then I thought: what if we would try to think "outside the box"? Could there be a different way of doing it? Could there be a different way of looking at it?
If I lie under the table and look up, I still see the same table as I saw from above, but it looks totally different.
So this is what I tried to do. DIFFERENT. It ended by being as different as the table from underneath. Different in every way.
And it works.

The boss said: Why didn't somebody think of that before?

Nicely put.
With coils taken out from inductive cooker; i am trying also to think outside the box.
But i can't claim success yet, it is too early, just started.
Same as you; i also hesitate to disclose the idea and details.
Because of the cases like this one is:
https://www.geotech1.com/forums/show...hp?25676-Leech

[Tinkerer]

Interesting trace. Usually the first half of the decay would be faster than the last half. Only way I can get a trace that looks similar, is to underdamp and let the return to zero go to -.6V across the MOSFET. Would be hard to see with 200V/div. Wondering how you get the mirror image critical damped.

You make some very good observations here. I will try to take a detailed picture of the zero level for you.
In the meantime, a picture of the coil, before assembly.

[Tinkerer]

Nicely put.
With coils taken out from inductive cooker; i am trying also to think outside the box.
But i can't claim success yet, it is too early, just started.
Same as you; i also hesitate to disclose the idea and details.
Because of the cases like this one is:
https://www.geotech1.com/forums/show...hp?25676-Leech

I congratulate you for your search for innovation. This is the way to go.

For many years I have posted all my ideas on the forum, even if I did not know how to explain what I was trying to do. Or did not even know what I was doing. I still don't know what I am doing half of the time, but I try to understand the best I can.

Now I can not give all away anymore, because it is not mine to give.

But I can say that much: keep trying, there is much out there. There is a whole new world outside the box.

[ivconic]

Your previous attachment is invalid. I can't open and see it.
...
Yes i am thinking outside the box for quite some time.
But i got smarter in years, not to immediately disclose all what's on my mind and post here on forum.
Because i get nothing in return from being so naive and enthusiastic, i get only occasional "b.tch-slap" from some highly educated fachidiots.
And whats worse; i get attacks from several totally illiterate ignorant ... all the way!
I learned this; worst thing to be in life is to be either fachidiot either illiterate ignorant, two very ugly extremes.
...
I still don't know what I am doing half of the time too!
But that's the main beauty in whole story!
Cheers!

[Tinkerer]

OK, here is the picture of the fast-slow coil, TX&RX again