ABANDONED

[green]

I started simulating this patent yesterday in LTSpice.

Wasn't getting good results and needed to adjust both the Subber RC values and the Diode Bias.
It is the Diode biasing that is crucial to getting this to work. As with most patents I have read they only write up the concept and never the critical details. I'm having best results when Vbias is +20 to + 30 Volts with 10ma through the diodes when conducting.

I am getting OUTPUT matching the Coil decay at less than 3usec. This should be great for low TC targets.

There is also the claim that gain is not needed but unsure why. In the 'simple' PI detector we run a gain of 500 to 1000 before the integrator. This is equivalent to only 'seeing' the last 4-8mV of the coil decay. In this patent is seems we can 'see' the last 4 Volts of the decay. Again unsure without building the circuits.

Anyway, very interesting.

Maybe you could post your simulation. I tried again and still have nothing. Maybe someone can see where my error is.

[Sean_Goddard]

Lack of disclosure is one of the reasons to refuse a patent application. The USPTO didn't do its job.

Does this come as a surprise ?

Green, I didn't simulate it, I built the circuit (minus the snubber for which I used a 2W 150V Zener) then I found no matter how hard I tried, I couldn't get the diodes to go into conduction. I sat and looked at what I had built, and what was in the patent then the answer hit me. I have yet to try my solution which is a diode with the anode connected to the coil, this then sits across an RC network (the R is variable). The flyback pulse charges the C and the R sets the discharge time, by carefully setting the time constants to match the coil I am using I think this should work to prove the theory. I have not yet worked out how to set the capacitor Vmax so that I can create a set point for the "trigger" point, again, I might resort to the crude my effective method of using a Zener.

Im my last post I mentioned Moodz PI, I meant Tinkerer.

[waltr]

Maybe you could post your simulation. I tried again and still have nothing. Maybe someone can see where my error is.

Make V4 30-40Volt. This is the trick.
Then R5 & R6 are about 2k.

Just ran your sim and it is working. Compare trace at a to d in the 1 to 5usec range and -9V to 20V range.
Note how trace at d and a match at about 2.6us.
However, the decay rate get much slower. Change V4 to 30V and R5 & R6 to 2k and look again.

My sim:

pulsingCoil-720.zip

Set in Edit Simulation Command:
Stop time = 0.00106
Time to start saving = 0.001045
Run then put Voltage probe on Coil (L1-D3), Between D3 & D4, at D4-R4.
Manually scale Voltage plot to: Top = 20V, Tick = 5V, Bottom = -10V
Now you should see the Coil Voltage decal and about 1.32us after TX off the Output (D4-R4) matching the coil decay.

[green]

Make V4 30-40Volt. This is the trick.
Then R5 & R6 are about 2k.

Just ran your sim and it is working. Compare trace at a to d in the 1 to 5usec range and -9V to 20V range.
Note how trace at d and a match at about 2.6us.
However, the decay rate get much slower. Change V4 to 30V and R5 & R6 to 2k and look again.

My sim:
[ATTACH]45186[/ATTACH]
[ATTACH]45185[/ATTACH]

Set in Edit Simulation Command:
Stop time = 0.00106
Time to start saving = 0.001045
Run then put Voltage probe on Coil (L1-D3), Between D3 & D4, at D4-R4.
Manually scale Voltage plot to: Top = 20V, Tick = 5V, Bottom = -10V
Now you should see the Coil Voltage decal and about 1.32us after TX off the Output (D4-R4) matching the coil decay.

Still think it's not working. Changed target TC to 4.35usec, decay 10usec/decade. Added simulation without added circuit to show what I think decay should look like.

[green]

Hi waltr

I've been simulating targets with an inductor and resister in parallel. In your simulation, pulsing coil-720.zip. you have more. What did you use to determine components needed? Including picture with your targets.

[waltr]

Hi waltr

I've been simulating targets with an inductor and resister in parallel. In your simulation, pulsing coil-720.zip. you have more. What did you use to determine components needed? Including picture with your targets.

I should have stated these Targets have been posted by others in this forum. I just did a forum search on "target simulation" and found a thread about doing these. There are a couple of LTSpice files posted in that thread with targets.
Not really sure if these are accurate but they do change the decay curve when used.

The spice directives for the inductor coupling can be 'commented' to remove the targets.

[waltr]

Still think it's not working. Changed target TC to 4.35usec, decay 10usec/decade. Added simulation without added circuit to show what I think decay should look like.

Yes, we are all unsure if this circuit works except that Sean_Goddard say he has built it and it does work.
I was able to get a better match in the decay of coil verse output (nets a vs c) by careful tweaking of Vbias.
I originally just plot lin/lin and it looked good. Then tried log/lin (thanks for showing this) and the mismatch is easier to see.
It is strange that the decay curve is not straight in the log/lin plot but this may be a clue as to how to do discrimination.

Here is image of sim: Note Bias Voltage.

[moodz]

Yes, we are all unsure if this circuit works except that Sean_Goddard say he has built it and it does work.
I was able to get a better match in the decay of coil verse output (nets a vs c) by careful tweaking of Vbias.
I originally just plot lin/lin and it looked good. Then tried log/lin (thanks for showing this) and the mismatch is easier to see.
It is strange that the decay curve is not straight in the log/lin plot but this may be a clue as to how to do discrimination.

Here is image of sim: Note Bias Voltage.

Hey Waltr .... Sean's last post says it did not work .... and hes going to try a zener.

Anyway looking at the sim ... you still have several mA of current in the coil even out to 10 microseconds. If you've got current flowing in the coil during the RX then your detector has a problem.

[moodz]

If we just revise the circuit to generate the "bias" off the output off the opamp then we end up with a workable solution that is simpler. ( works too LOL ;-) )

[KingJL]

If we just revise the circuit to generate the "bias" off the output off the opamp then we end up with a workable solution that is simpler. ( works too LOL ;-) )

[ATTACH]45194[/ATTACH]

Without the disadvantage of a ton of noise!

[waltr]

If we just revise the circuit to generate the "bias" off the output off the opamp then we end up with a workable solution that is simpler. ( works too LOL ;-) )

[ATTACH]45194[/ATTACH]

Interesting. Opamp output is inverted output of coil after about 4us.
Also seems to have amplified the Coil V a good bit.
This is whats great about this forum.

[Teleno]

If we just revise the circuit to generate the "bias" off the output off the opamp then we end up with a workable solution that is simpler. ( works too LOL ;-) )

[ATTACH]45194[/ATTACH]

Well, this is a perfect attenuator (1/2000 gain) and noise generator (2mV input noise).

The trick of the patents' diode switch is to bias the diodes in such a way that the signal's attenuation is minimum. The conditions are:

- High forward voltage (higher current). The transition from reverse to direct bias (capacitance change) takes place earler in the decay curve, with less attenuation of the useful signal .
- High dynamic resistance (constant-current bias). The signal sees an infinite impedance, minimizing attenuation.

A theoretical circuit would then look like this:

[moodz]

Interesting. Opamp output is inverted output of coil after about 4us.
Also seems to have amplified the Coil V a good bit.
This is whats great about this forum.

There is a quirky bit ... the opamp must be run single ended ( ie +volts and ground supply ) if a bipolar supply is used the circuit latches up ... well in the sim anyway ... havent analysed it but the probably the numerous diodes. :-) .... but yes it does appear to work.

[green]

Added a battery to output of Teleno's zip for offset adjust. Makes it easier to see if simulation makes sense charting linear-log. Straight line decay. Tried on other suggested sims and didn't see straight line decay.

[Teleno]

A well configured diode switch beats the hell out of the traditional back-to-back diodes.

In the back-to-back configuration, the resistor in series with the diodes and the damping resistor are connected in parallel during the whole decay except for the last 500mV, when the diodes stop conducting. This causes a change of the time constant at the end of the decay, which becomes faster (green, this is why you never got straight lines before). In order for the decay to be stable, the damping resistor has to be made lower than optimum. Instead, the diode switch allows to use the optimal damping resistor throughout the whole transient, which translates in the ability to sample of 1us - 2us earlier compared to the diode switch.

Depending of the series resistor of the b-t-b configuration, the input noise can be 10 times lower in the diode switch!

The circuit of the patent cannot deliver any of these benefits, the starting point is the conceptual switch I posted above. I've come to a simpler practical design that I won't share for the time being.