Perhaps I should have applied for a patent?

[Idefix]

Hi,
I have done this sim, in order to found an approaching model,
The way is to test the real coils with ... network analysor, for example.


I can make runs varying the transfo coupling.
And stepping some values too.

I have ommited the coax, witch adds some pF .



to admin : what the way to paste a picture instead of an Attach with link?

[Carl-NC]

to admin : what the way to paste a picture instead of an Attach with link?

Click "Go Advanced"... there are then additional options, including image uploads.

[Teleno]

If the difference relative to the prior art is just the inductance of the Tx coil, then it falls within the design options of the person skilled in the art and is not inventive.
If the difference lies in using a shield as a Tx coil, the question is whether this has an unexpected technical effect relative to a normal coil (like cancelling the Tx field ), otherwise it's just an equivalent embodiment, which is not inventive.

[Idefix]

More parameters need to be known about the wiring and the coil, to have a better model.
Is the RX coil damped? and what is the value ?
What was the SRF before shielding, if measured, and after shielding ?

The runs hereafter, are with different value of the coupling of the transform. All with 800 ohm damping the RX coil.

[green]

Used a model I had for an IB coil. Probably needs added coupling between Tx and Rx (K3 L1 L3 .5?). Can't sample at 1usec when added. Maybe something wrong with model.

Been playing with spice. Would like to understand spice better and predict what Eric gets in hardware. Not adding amplifier to spice model, just amplifies Rx signal and could slow response. Adding the series diode does make a difference with Tx around 300uH, doesn't seem to help with Tx .5uH(does effect oscillation trace with Rd=100k). Adding coupling between Tx and Rx looks like it increases the delay time past 1usec. Don't know what the coupling constants should be. Not sure any of the spice model makes sense. Any thoughts on what should be changed? Including spice zip without coupling between Tx and Rx(some component values different than Eric2.png) .

Changed target to 1.6usec TC to simulate 1x1inch aluminum foil.

[Ferric Toes]

Hi Green and Idefix,

I have a 4K7 preset for damping across the RX coil and get the best damping at about 3K.

Eric.

[green]

Hi Green and Idefix,

I have a 4K7 preset for damping across the RX coil and get the best damping at about 3K.

Eric.

Thanks Eric. Do you have leads to the Rx coil? I calculate a SRF about 3MHz for Rd=3k.

[Ferric Toes]

All so far with the coil and shield just wired straight to the board with about 6in of wire.

Eric.

[green]

All so far with the coil and shield just wired straight to the board with about 6in of wire.

Eric.

Changing Rx capacitance to 10pf(about 3MHz resonance) does make the spice model look closer to what you are getting even with Tx Rx coupling(can sample at 1usec). Tx still oscillates with no Rd.


Idefix reply#19 is correct

More parameters need to be known about the wiring and the coil, to have a better model.
Is the RX coil damped? and what is the value ?
What was the SRF before shielding, if measured, and after shielding ?

[Idefix]

Thanks Eric. Do you have leads to the Rx coil? I calculate a SRF about 3MHz for Rd=3k.

Hi,
May be I don't take everything in account,
RX is 325µH, the ideal damping is 3000ohms : So it gives 145pF as value of C, and Freq of 733 KHz.

[KingJL]

Green, I would expect the parallel capacitance of a 0.5u single winding TX coil to be extremely low (mainly the capacitance of the TX coil leads and gap capacitance). The model you have does not model the 80pf (as measured by Eric) coupling between the TX and RX coil (which is probably the only TX RX energy transfer in the real circuit).

Kind regards,
J. L. King

[green]

Hi,
May be I don't take everything in account,
RX is 325µH, the ideal damping is 3000ohms : So it gives 145pF as value of C, and Freq of 733 KHz.

I use Rd=pi*L*SRF

[green]

Green, I would expect the parallel capacitance of a 0.5u single winding TX coil to be extremely low (mainly the capacitance of the TX coil leads and gap capacitance). The model you have does not model the 80pf (as measured by Eric) coupling between the TX and RX coil (which is probably the only TX RX energy transfer in the real circuit).

Kind regards,
J. L. King

The model you have does not model the 80pf (as measured by Eric) coupling between the TX and RX coil (which is probably the only TX RX energy transfer in the real circuit).

reply#? Where would I add the 80pf to the spice model? Tx couples to target and target couples to Rx, Why doesn't Tx couple to Rx?

I had forgotten Eric stated Rx SRF=650kHz reply#11 instead of 3MHz I calculated for Rd=3k. I'm getting more confused.


I have compared the spice model with what I get on the bench. Most of the time makes sense. My bench circuit coil not similar to what Eric is trying so I'm missing something, looks like I'm missing a lot.

[KingJL]

The model you have does not model the 80pf (as measured by Eric) coupling between the TX and RX coil (which is probably the only TX RX energy transfer in the real circuit).

reply#? Where would I add the 80pf to the spice model? Tx couples to target and target couples to Rx, Why doesn't Tx couple to Rx?

I had forgotten Eric stated Rx SRF=650kHz reply#11 instead of 3MHz I calculated for Rd=3k. I'm getting more confused.


I have compared the spice model with what I get on the bench. Most of the time makes sense. My bench circuit coil not similar to what Eric is trying so I'm missing something, looks like I'm missing a lot.

I have taken the liberty to modify your model and timings to reflect what I understand of Eric's problem/circuit statement and a minor change to what I use for a target. The 80p coupling between TX and RX is modeled by C1. In the real world this is not quite correct as the 80p capacitance is distributed down the whole coil not just at one end. To show this distribution you would have to break up the RX and TX coils into multiple coil segments connected in series and the TX/RX capacitance divided equally across to the individual RX/TX coil segments.

[green]

I have taken the liberty to modify your model and timings to reflect what I understand of Eric's problem/circuit statement and a minor change to what I use for a target. The 80p coupling between TX and RX is modeled by C1. In the real world this is not quite correct as the 80p capacitance is distributed down the whole coil not just at one end. To show this distribution you would have to break up the RX and TX coils into multiple coil segments connected in series and the TX/RX capacitance divided equally across to the individual RX/TX coil segments.

Thanks, that is what I was hoping someone would do. Tried making the changes. Doesn't look like I could sample at 1usec. Maybe missed a change or something hidden(like coil capacitance). Could you post your spice model with changes?