PI DAMPING EQUATION CHALLENGE

[moodz]

Playing around with a "practical circuit" but now with "real component models" meaning that the circuit is actually buildable ( and thats the next step ). We see some interesting artifacts that will be used to make it work.

As Carl mentioned the coil is going to change inductance with ground / hotrocks / targets so the plot below is a coil stepping in 30 microhenry steps so we can see the result ( in real world I only see the coil change 2 - 3 microhenries over ground unless you put the coil on a sheet of steel etc.).
What is interesting is that we get a "pivot" point for the coil current and an "error bump" voltage when the coil is not at zero current. The error voltage is proportional to how far the coil is from damped. This is an immediate opportunity to sample the error voltage at the pivot point time and the feed back loop will be a Voltage to PulseTime conversion ( voltage to freq perhaps ?? ) to normalise the damping point when coil inductance changes.

As you can see the error voltages are big enough ( more than 1 volt per microhenry ) for an error control loop to work off.

[moodz]

Added the RX front end for feeding into the first amp ... the pic below shows the current being damped from 1 amp in around 600 nanoseconds and we have a target response starting at 680 nanoseconds. ( rather a large target coupled in at 10% so we can see it note ... this would be the input to the first amplifier )

I am a bit surprised but this is actually looking good so far ( remember we are on simulation here ... pudding tastes best on eating not on reading the recipe ).

[moodz]

...added a separate analogue ground and now getting some really nice target responses. Target coupling is 5%. Damping in 618 nanoseconds @1 AMP peak 800 volts. No amplification.

[ivconic]

As long as you eat imaginary pudding; you stay hungry!


It's time to make it and try it out in the real world.
You shoot a couple of video clips and that's it, later.
From the very beginning, still on the AMX thread; I liked the idea of such a TX.
But it should be better elaborated and explained (in one concise post and without graphics over the entire post, if possible) and all
the disadvantages of such an approach, how to avoid shut-through, how to avoid avalanche (Carl case) while at the same time one can
freely choose and slightly higher supply voltage than 2v. And of course there are a lot of more precise questions that need to be answered.
Of course, some questions can't be answered until things move out of the simulator and into the real world.
But I believe you are quite right in your expectations and assumptions.
However, details are needed, on practical implementation and application.
But ok, we'll wait for it to materialize and see how it turns out.
Things are much easier to understand if the whole thing is explained well in words alone.
Images, oscillograms, assumed situations in the simulator... are completely clear to those who do it directly.
But to the one who does not have the conditions to reproduce it all identically on his table; more detailed explanations are needed.
The thing about this TX has developed to the point that it is now very convenient material for a good and extensive PDF... not to mention a smaller book.
But without practical tests on the "material", everything will not be known.
You have my support to see it through.

[Gunghouk]

This formula does not work if you change the coil or pulse repetition rate so it works only for this set of operating parameters and is not a universal solution as the current changes with these two variables.

[Gunghouk]

Changing the coil parameters for my crappy 25 way screened cable concoction meant decreasing the repetition rate from 10kHz to 1.6kHz.
Rs=3.8
L=470u
C=403p (calculated)
Pulse period=625uS

[ivconic]

This formula does not work if you change the coil or pulse repetition rate so it works only for this set of operating parameters and is not a universal solution as the current changes with these two variables.

Among other things, I was thinking about that. If the design is conditioned by specific conditions and does not have a wider range of operation and selection of operation modes; then it's a design flaw.

[Gunghouk]

Ok I can it get to work at different rep rates but only with 1.37uS damping time with these mods (harder driven Voodoo TX cct)

[bbsailor]

I would like to offer another damping resistor issue to consider. Resistor values, depending on the resistor type and wattage rating may change values based on the temperature of the environment they are used in and the amount of energy that is being damped. Warmer resistors tend to have their values rise. This could cause a damping stability issue if calculated or tuned damping resistor value is very close close to the stability value point..

Try to choose a damping resistor value and wattage rating that does not get too warm when operating. Here is a simple test that I have used. Measure and record the damping resistor value with a good ohm meter as follows.
1. Ohm value at room temperature.
2. Hold between fingers for 2 minutes.
3. Heat with a hair dryer for 1 minute.
4. Place in a refrigerator for about one hour.
5. Place in a freezer for about one hour.

Based on how much the damping resistor value changes higher, select a resistor value that you can mount in parallel with the damping resistor to bring the value back to the initial room temperature value. Use a small toggle switch to turn on the parallel resistor if the metal detector becomes unstable. Remember, the optimum goal is to choose a damping resistor type, accuracy, and wattage that keeps its value while operating for several hours in the search environmental temperature. The switched in parallel resistor would be a value to lower the drifted damping value. A 1000 ohm damping resistor with 100,000 ohm parallel resistor would now be a 990.099 ohm value using the R1 X R2 divided by R1 + R2 value formula.

Here is a simple metal model to keep in mind when designing and building a pulse induction metal detector. When the TX pulse raises to about 95 percent of max current at about 3 coil time constants, the TX pulse turns off and a voltage/current oscillation will occur that is based on the total capacitance the coil sees that must be fully damped too allow the RX circuit to turn on. The amount of energy in these oscillations will govern the amount energy the damping resistor will absorb and be the source of potential damping resistor value changes.

Joseph J. Rogowski

[moodz]

Changing the coil parameters for my crappy 25 way screened cable concoction meant decreasing the repetition rate from 10kHz to 1.6kHz.
Rs=3.8
L=470u
C=403p (calculated)
Pulse period=625uS

Your coil works perfectly with the formula .... There is no time term in the equation so repetition rate has nothing to do with it and is theoretically being damped in 750 nanoseconds.

[moodz]

Among other things, I was thinking about that. If the design is conditioned by specific conditions and does not have a wider range of operation and selection of operation modes; then it's a design flaw.

How many exams have you been to where people dont read the question properly and consequently dont get the right answer

[moodz]

I would like to offer another damping resistor issue to consider. Resistor values, depending on the resistor type and wattage rating may change values based on the temperature of the environment they are used in and the amount of energy that is being damped. Warmer resistors tend to have their values rise. This could cause a damping stability issue if calculated or tuned damping resistor value is very close close to the stability value point..

Try to choose a damping resistor value and wattage rating that does not get too warm when operating. Here is a simple test that I have used. Measure and record the damping resistor value with a good ohm meter as follows.
1. Ohm value at room temperature.
2. Hold between fingers for 2 minutes.
3. Heat with a hair dryer for 1 minute.
4. Place in a refrigerator for about one hour.
5. Place in a freezer for about one hour.

Based on how much the damping resistor value changes higher, select a resistor value that you can mount in parallel with the damping resistor to bring the value back to the initial room temperature value. Use a small toggle switch to turn on the parallel resistor if the metal detector becomes unstable. Remember, the optimum goal is to choose a damping resistor type, accuracy, and wattage that keeps its value while operating for several hours in the search environmental temperature. The switched in parallel resistor would be a value to lower the drifted damping value. A 1000 ohm damping resistor with 100,000 ohm parallel resistor would now be a 990.099 ohm value using the R1 X R2 divided by R1 + R2 value formula.

Here is a simple metal model to keep in mind when designing and building a pulse induction metal detector. When the TX pulse raises to about 95 percent of max current at about 3 coil time constants, the TX pulse turns off and a voltage/current oscillation will occur that is based on the total capacitance the coil sees that must be fully damped too allow the RX circuit to turn on. The amount of energy in these oscillations will govern the amount energy the damping resistor will absorb and be the source of potential damping resistor value changes.

Joseph J. Rogowski

Hey Joe this is a valuble insight and there are thermal drift issues to be considered. However what we are trying to achieve in this thread is damping of coil energy at the point in time that during a coil turn off ( and the LC tries to oscillate ) the coil at a point in time transfers all its energy to the C capacitance at peak flyback and at that precise time ( nanosecond resolution ) time we discharge the C into a small resistance which converts to heat energy.
So we are not talking about a conventional damping resistor.

moodz

[moodz]

Hey Joe this is a valuble insight and there are thermal drift issues to be considered. However what we are trying to achieve in this thread is damping of coil energy at the point in time that during a coil turn off ( and the LC tries to oscillate ) the coil at a point in time transfers all its energy to the C capacitance at peak flyback and at that precise time ( nanosecond resolution ) time we discharge the C into a small resistance which converts to heat energy.
So we are not talking about a conventional damping resistor.

moodz

Also I should mention that the resistor we are talking about will vary in proportion to I^^2 or current squared in the coil ... so its not a regular resistor.

[ivconic]

How many exams have you been to where people dont read the question properly and consequently dont get the right answer

The real question is how many videos have I seen of your work?

[Gunghouk]

Hi Moodz, can you post your original LTspice model please ? I'm new to LTspice and there must be an error in mine somewhere or its my attempt to use available thru hole devices.
Thanks for the stimulating topic