Detect a target with a TC less than mono coil TC

[green]

I've been thinking, can't but wondering if I'm wrong. Tried to make a spice simulation to see if I could. Haven't been successful, any thoughts on how to make or change the simulation so it could detect a target with a TC less than the coil TC?

[bbsailor]

I've been thinking, can't but wondering if I'm wrong. Tried to make a spice simulation to see if I could. Haven't been successful, any thoughts on how to make or change the simulation so it could detect a target with a TC less than the coil TC?

green,

Please define what you mean by coil TC? Here is what a coil TC could mean.
1. The delay time between the TX turn off and the RX turn on.
2. The discharge TC of the coil as defined as the coil inductance divided by the effective damping resistance value. This is critical to fully stimulate small, low TC targets at five times less than the target TC. Since damping resistor values are related to coil current levels and coil, coax, MOSFET and circuit capacitance as seen by the coil, reducing all these capacitance levels will collectively allow a higher damping resistor value and cause the coil discharge slope to become more vertical (faster) and better able to fully stimulate smaller target TCs.
3. During the delay time between TX turn off and RX turn on any eddy currents induced into a target are decaying. If you are sampling at a 10uS delay and your target TC is 2uS, after five target TCs or 10uS your target currents will have decreased to about 99.5 percent of peak with little left to detect. That is why with 2uS targets you want to have a sample delay of about 7uS or 8uS to pick up the tail end of this targets signal before it drops to zero or into the noise level.
4. Now, when you get the delays down below or about 8 to 10uS you need to consider if the coil wire itself is retaining any eddy currents and acting like a very small target.
5. Preamp saturation time requires a careful design to minimize it's saturation time so the RX signal can be turned on as soon as possible.

There is an interaction of variables but as delays get lower these interactions become greater and need to be considered in any analysis of tradeoffs related to:
1. Coil size (diameter)
2. Coil type (mono, DD, other)
3. TX power
4. RX gain
5. Damping resistor value
6. Coil design techniques
7. Full target stimulation
8. Rx signal integration and sweep speed
9. Optimum detection depth or strength relative to target size,, target TC, coil size, coil type, soil type and sweep speed.

I hope this helps.

Joseph J. Rogowski

[dbanner]

There is so much scattering of good information on the forum, it makes it hard for young players to take their knowledge of PI technology to the next level.
Add to that Mr Green's bode plots and log/lin, log/log graphs and we see a picture of bewilderment in the making.

It really does take some dedication if you want to grasp all the variables to have a more "complete" understanding.
I shall have to apply myself in the new year to such a task, keeping in mind not to blow a fuse!

[green]

green,

Please define what you mean by coil TC? Here is what a coil TC could mean.
1. The delay time between the TX turn off and the RX turn on.
2. The discharge TC of the coil as defined as the coil inductance divided by the effective damping resistance value. This is critical to fully stimulate small, low TC targets at five times less than the target TC. Since damping resistor values are related to coil current levels and coil, coax, MOSFET and circuit capacitance as seen by the coil, reducing all these capacitance levels will collectively allow a higher damping resistor value and cause the coil discharge slope to become more vertical (faster) and better able to fully stimulate smaller target TCs.
3. During the delay time between TX turn off and RX turn on any eddy currents induced into a target are decaying. If you are sampling at a 10uS delay and your target TC is 2uS, after five target TCs or 10uS your target currents will have decreased to about 99.5 percent of peak with little left to detect. That is why with 2uS targets you want to have a sample delay of about 7uS or 8uS to pick up the tail end of this targets signal before it drops to zero or into the noise level.
4. Now, when you get the delays down below or about 8 to 10uS you need to consider if the coil wire itself is retaining any eddy currents and acting like a very small target.
5. Preamp saturation time requires a careful design to minimize it's saturation time so the RX signal can be turned on as soon as possible.

There is an interaction of variables but as delays get lower these interactions become greater and need to be considered in any analysis of tradeoffs related to:
1. Coil size (diameter)
2. Coil type (mono, DD, other)
3. TX power
4. RX gain
5. Damping resistor value
6. Coil design techniques
7. Full target stimulation
8. Rx signal integration and sweep speed
9. Optimum detection depth or strength relative to target size,, target TC, coil size, coil type, soil type and sweep speed.

I hope this helps.

Joseph J. Rogowski

Discharge TC of the coil, time for coil volts to decrease to 37% of start value. TC=time to change 1 decade/2.3. Circuit decay, includes external capacitance. The simulation shows coil decay TC=L/2Rd not L/Rd.

[Carl-NC]

The simulation shows coil decay TC=L/2Rd not L/Rd.

The math also shows this. It is an RLC circuit, not an RL circuit.

[bbsailor]

The math also shows this. It is an RLC circuit, not an RL circuit.

Carl,

My question is based on helping me to better understand the difference between simulation measurements versus real circuit measurements. When I stated "effective damping resistance" value I was inferring that the clamping diodes place the input resistance value (R12 of 1000 ohms of Hammerhead) effectively in parallel with the damping resistance (Rd) value in a mono coil. In your Hammerhead design article, Rd (R11) is 680 ohms but the input resistor to the input (opamp stage) is R12 of 1000 Ohms. While the fly-back voltage is above the conductive voltage of these diodes (typically about 0.6V), these diodes put the 1000 ohm resistor effectively in parallel with Rd. This makes the 680 ohm (Rd) resistor now about 400 ohms while the diodes are conducting and extending the discharge TC until the voltage drops to below 0.6V so that only the Rd value represents the coil discharge TC.

My question is: Do simulation values represent the actual mono PI circuit performance measured values?

Thanks for making a great metal detecting forum.

Joseph J. Rogowski

[Carl-NC]

I'm not quite sure of the question, but for a critically damped RLC circuit (which is the case for a PI) the decay TC is L/2R, where R is the total effective damping resistance. That may include the clamp R. Obviously, the decay can be critically damped while clamping, or while not clamping, but not both, so it's a little erroneous to consider it critically damped all the way. But I usually consider it "close enough."

[dbanner]

There is going to be a transient response and you can bet that spice modelling is going to notice this.

[green]

This is critical to fully stimulate small, low TC targets at five times less than the target TC.

SRF for circuit reply #1=318kHz (.5us TC). Not a fast coil. Changed capacitance and Rd for SRF=1.59MHz (.1us TC). Did a simulation for both to see the difference in target stimulation. Been thinking target decay should cross coil decay to detect target but all that is necessary to detect a target is for the coil voltage to change more than the noise. Wondering if JL could detect the .5us target with a circuit SRF of 318kHz with his 20bit A-D. Wasn't thinking about targets with a TC less than the coil decay TC decaying at the coil decay TC. Makes sense because the are coupled?

[bbsailor]

I'm not quite sure of the question, but for a critically damped RLC circuit (which is the case for a PI) the decay TC is L/2R, where R is the total effective damping resistance. That may include the clamp R. Obviously, the decay can be critically damped while clamping, or while not clamping, but not both, so it's a little erroneous to consider it critically damped all the way. But I usually consider it "close enough."

Carl,

Thanks for clarifying my observation that the effective Rd (damping resistance) value includes the input R (Rin of the first operational amplifier stage) that is effectively in parallel with Rd during the time that the voltage is above 0.6V (the typical clamping voltage of these diodes). This would mean that the discharge slope would include Rd in parallel with Rin until the slope reaches 0.6V when the clamping diodes opened. Clarifying your equation for coil discharge TC should include this subtle consequence.

My tinkering with coils made me observe this.

Thanks

Joseph J. Rogowski

[Skippy]

[non-PI guy here]
Do you think there's any technical advantage to maintaining a constant damping resistance for the entire decay time ? I was thinking along the lines of switching in a damping resistor, once the transient voltage had fallen below the approx 0.6 V where the protection diodes conduct.

Apologies if it's a well-used technique....

[dbanner]

Carl,

Thanks for clarifying my observation that the effective Rd (damping resistance) value includes the input R (Rin of the first operational amplifier stage) that is effectively in parallel with Rd during the time that the voltage is above 0.6V (the typical clamping voltage of these diodes). This would mean that the discharge slope would include Rd in parallel with Rin until the slope reaches 0.6V when the clamping diodes opened. Clarifying your equation for coil discharge TC should include this subtle consequence.

My tinkering with coils made me observe this.

Thanks

Joseph J. Rogowski

How exactly does one calculate a single TC for two different values of effective Rd. That would seem to be a rather complex equation. Or maybe not. The time taken to reach the clamp voltage (TC1) and then the time taken to reach zero (TC2)
(TC1xTC2)/(TC1+TC2): (L/2Rd1)x(L/2Rd2)/(L/2Rd1+L/2Rd2) and you get the average TC for the decay.

Which such fuzzy math, I might blow something up!

[green]

How exactly does one calculate a single TC for two different values of effective Rd. That would seem to be a rather complex equation. Or maybe not. The time taken to reach the clamp voltage (TC1) and then the time taken to reach zero (TC2)
(TC1xTC2)/(TC1+TC2): (L/2Rd1)x(L/2Rd2)/(L/2Rd1+L/2Rd2) and you get the average TC for the decay.

Which such fuzzy math, I might blow something up!

Rin going to +input is the only time there are two slopes. Coils with separate Tx and Rx and mono coil with Rin going to -input should have one slope. For Rin going to +input calculate parallel resistors above .6V. Don't know formula to calculate decay TC.

Added 1k input resistor and diode to simulation reply #1. Input resistor doubled TC for the simulation.

Do you think there's any technical advantage to maintaining a constant damping resistance for the entire decay time ? Yes if looking for short TC targets.

[KingJL]

[non-PI guy here]... I was thinking along the lines of switching in a damping resistor, once the transient voltage had fallen below the approx 0.6 V where the protection diodes conduct... Apologies if it's a well-used technique....

Well used... NO... but you are on the right track. The theoretical optimum would be to switch in the damp at the instant that the coil current reaches zero. You can approach this in practice, but it is not just a matter of switching in the damp... some additional things must happen at the same time... but it is doable.

[Skippy]

If Mr. Greens latest simulation were changed to R_d = 430 R, and input protection resistor R = 1k0 , then he would have critical damping ( 300 R ) until the protection diode stopped conduction, and a modestly under-damped response thereafter.