LT spice

[green]

Davor, I see what you did now, simple I was trying to make it difficult.

[Davor]

Coins under a PI tend to have high dependency on their thickness, due to skin effect obviously. How to simulate skin effect? I have no idea yet.

[green]

Coins under a PI tend to have high dependency on their thickness, due to skin effect obviously. How to simulate skin effect? I have no idea yet.

An attempt at the 1oz copper coin. Could be better. I had thought skin effect was causing straight line decay log-log in the beginning but was told it's something else.

The coin decays straight line linear time log amplitude 700us after Tx off with a 500us TC.

[Davor]

Nassim Nicholas Taleb propagates the idea of the Wittgenstein's ruler (Ludwig Wittgenstein, 20th-century philosopher):
"Unless you have confidence in the ruler's reliability, if you use a ruler to measure a table you may also be using the table to measure the ruler."
(In other words, if reality continuously delivers something that doesn't "make sense" to you, it is *you* that you need to worry about.)

There are a few things at play, so until we are able to rule out the contributions of the measuring equipment, say Tx charging period, we may only speculate.

[green]

Nassim Nicholas Taleb propagates the idea of the Wittgenstein's ruler (Ludwig Wittgenstein, 20th-century philosopher):
"Unless you have confidence in the ruler's reliability, if you use a ruler to measure a table you may also be using the table to measure the ruler."
(In other words, if reality continuously delivers something that doesn't "make sense" to you, it is *you* that you need to worry about.)

There are a few things at play, so until we are able to rule out the contributions of the measuring equipment, say Tx charging period, we may only speculate.

I've been hoping someone else would chart some target decays to compare with mine. If they got something similar maybe data might be correct.

Copied some simulation targets from another thread. Probably changed the driver to a current sink. Added the three on the left(L parallel with R). All targets decay straight line linear time log amplitude except for the foil target, it oscillates. L parallel with R makes a good foil target. Most if not all targets I've charted with a TC 10us or less chart straight line linear-log. Targets with a TC over 10us aren't straight line linear-log in the beginning unless they are thin. The reason I thought skin effect was causing it. Seems like skin effect has to cause some effect, to chart straight line linear-log, R has to be constant. Is there a way to calculate time when R is constant for thicker targets? Wanting some spice targets that chart what my tester does assuming it is correct.

How to prove or get some confidence in my tester? First it does make sense to me that thicker targets shouldn't chart straight line linear-log in the beginning. Don't know if it should be straight line log-log in the beginning. I'm trying to learn so maybe making sense isn't necessary other than trying to prove what is correct.

[Davor]

Very few people will have the equipment like yours at hand.
As for skin effect, perhaps the inductively coupled compound targets would do the trick.

[green]

Very few people will have the equipment like yours at hand.
As for skin effect, perhaps the inductively coupled compound targets would do the trick.

Before I built the target tester I charted a lot of target decays. Recorded amplifier out, no target and a target recording. Subtracted the no target recording from the target recording in Excel and then charted the result linear-linear, linear-log or log-log. Subtracting no target from target eliminates amplifier offset if recordings are done close enough together. Didn't get as clean a chart but good enough to see decay. Chart I posted awhile back using the subtraction method.

[green]

Did a test with different thickness copper plates(.05mm protective film each side of plates) and stacked(.1mm protective film between stacked plates) to see if anything makes sense. Still trying to make sense of the data. All plates decayed straight line log-log in the beginning. I have some copper foil(.0178mm and .0343mm thick)to cut in 102mm squares to test yet. Maybe someone sees something that doesn't make sense or does?

[green]

Did a test with different thickness copper plates(.05mm protective film each side of plates) and stacked(.1mm protective film between stacked plates) to see if anything makes sense. Still trying to make sense of the data. All plates decayed straight line log-log in the beginning. I have some copper foil(.0178mm and .0343mm thick)to cut in 102mm squares to test yet. Maybe someone sees something that doesn't make sense or does?

I see I forgot to label Y axis charts reply #23. log scale .4V/decade, 25mV in=0V out

[Davor]

Guess faster decaying materials would lead to more useful data. This way the end of curve picks a lot of noise. I think you should put both lin and log time diagrams. Point is that you may put a tangent line that will fit just about anything. The fact these fit fine in a long span only confirms you may make a new patent on old technology by merely using some odd curve that you implement a math module for. The company-whose-name-must-not-be-said recycles the GB mechanism at early samples to attenuate water response. As we know GB obeys 1/t law, but water may respond as ever it likes, and yet at a limited stretch of time the same mechanism will do.

[green]

Guess faster decaying materials would lead to more useful data. This way the end of curve picks a lot of noise. I think you should put both lin and log time diagrams. Point is that you may put a tangent line that will fit just about anything. The fact these fit fine in a long span only confirms you may make a new patent on old technology by merely using some odd curve that you implement a math module for. The company-whose-name-must-not-be-said recycles the GB mechanism at early samples to attenuate water response. As we know GB obeys 1/t law, but water may respond as ever it likes, and yet at a limited stretch of time the same mechanism will do.

Thanks for the replies. Best way to record and chart the data? What targets to record? Constant current or constant rate Tx? Tx time, 100us, 5000us or something between? Tx time needs to be long for ground 1/t decay. Interested in your or anyone thoughts.

Stacked quarters I did awhile back charted linear-log and log-log for discussion. Record time needs to be greater than 2times the target TC to get straight line decay linear log(maybe longer).

Charted the copper plates because I had different thickness plates. Could cut into smaller plates for faster decay. Think different thickness might not be same as stacked.

If we could make some spice targets that acted the same as real targets it would be easy to see what effect different Tx parameters have.


Your ground simulation looks good, 1uH parallel with 1ohm simulates a piece of regular strength aluminum foil 16mm square and 1uH parallel with .1ohm simulates a US nickel. Could use a few more.

Another chart I forgot to label Y axis, log scale .4V/decade 25mV in=0V out

[green]

I have a piece of the .533mm copper plate cut to 45mm square(TC about 180us). TC is proportional to width and thickness. If I cut the 1.02mm plate 23.5mm square it should have the same 180us TC when it becomes straight line linear-log. Wondering how much different the decay will be at the start of the decay, any thoughts? Would be easier to make simulation if there wasn't any difference.

[Davor]

When you see a ski-like bend at the beginning or the end, you know you are in a wrong scale. You may record the no target response, and subtract it from responses with targets. That way you'll have one variable less in the mix.

[green]

I have a piece of the .533mm copper plate cut to 45mm square(TC about 180us). TC is proportional to width and thickness. If I cut the 1.02mm plate 23.5mm square it should have the same 180us TC when it becomes straight line linear-log. Wondering how much different the decay will be at the start of the decay, any thoughts? Would be easier to make simulation if there wasn't any difference.

Tried the test. Less signal with the smaller plate, used the smaller coil to get enough signal strength. Spaced each target for near full scale signal at start of decay. Recorded twice to test for repeatability. Don't know what to conclude other than thicker plate decays faster at start and decay rates are similar near the end.

[green]

An attempt at a US quarter. Ball park, at least it's not straight line decay lin-log. Charted ground, couple ferrous targets and couple non ferrous targets to try to simulate if anyone wants to try. Davor has a ground simulation reply#3.