PI: On-time sampling versus Off-time.

[Teleno]

By "on-time" I mean sampling the target's signal while the Tx coil is carrying a constant current.
Off-time is the classic method of sampling after the Tx current has dropped to zero.

Advantages of On-time sampling:

1. A short pulse (30us) follow by a long off-time (1ms) maximizes the response of both short and long time constants.
2. Ground response is full, allowing cancelling by a 1/t approximation.
3. Tx transient is faster because damping is unnecessary (Tx coil is always connected to a low impedance).

Disadvantages:

1. Separate Rx induction-balance coil is required.
2. Additional control circuit to maintain a constant current in the Tx coil.
3. Additional circuitry to quickly charge-up the coil (high voltage).
4. Variation in inductance by magnetic permeable and ferrous objects cause false signals.
5. Sampling takes place while the magnetic field in the T coil is moving, causing false signals (true on false????)

What are your thoughts, people?

[dfbowers]

I see this subject appear from time to time.
Here is one from back in 2007.
http://www.geotech1.com/forums/showt...+discriminator

Don.

[Teleno]

I see this subject appear from time to time.
Here is one from back in 2007.
http://www.geotech1.com/forums/showt...+discriminator

Don.

That's an IB PI (which is an option for the MPP, for example), but there's no mention of on-time sampling which is the issue at hand.

[green]

That's an IB PI (which is an option for the MPP, for example), but there's no mention of on-time sampling which is the issue at hand.

Looks interesting. Some thoughts, might be all wrong. Including spice simulation to compare signal at CC and after turn off. Don't see a big difference with 2 and 10usec TC targets. Depending on sample time 30usec on time might not be enough to do GEB. Have you tried the circuit on the bench to see how much it gains in signal strength?

[Teleno]

Looks interesting. Some thoughts, might be all wrong. Including spice simulation to compare signal at CC and after turn off. Don't see a big difference with 2 and 10usec TC targets. Depending on sample time 30usec on time might not be enough to do GEB. Have you tried the circuit on the bench to see how much it gains in signal strength?

No bench yet.

Simulation for 10us and 100usec target, Tx pulse is 20us with a 2ms period. Polarities are opposite for on-time and off-time signals. Left (+): on-time, Right(-): off-time.
There's a 16.25dB gain on the 100us signal (6,5x)

[green]

No bench yet.

Simulation for 10us and 100usec target, Tx pulse is 20us with a 2ms period. Polarities are opposite for on-time and off-time signals. Left (+): on-time, Right(-): off-time.
There's a 10.5dB gain on the 100us signal (3,3x)

Your simulation for the 10usec TC target is similar to mine. I didn't include higher TC targets because the most of the higher TC targets I've charted don't chart a straight line on a linear log chart. Closer to a straight line on a log log chart with slope varying with coil on time, the reason for the question about bench testing.

[Teleno]

the most of the higher TC targets I've charted don't chart a straight line on a linear log chart. Closer to a straight line on a log log chart with slope varying with coil on time

Yes, that's expected when you smple off-time. When you sample on-time is the opposite, the slope is proportional to off-time The longer the off-time period, the larger the on-time signal.

..the reason for the question about bench testing.

Bench testing is fine once you've designed a working CC driver for your coil. It's not straighforward.

[green]

[Bench testing is fine once you've designed a working CC driver for your coil. It's not straighforward.]


The reason I haven't tried it. Hope you succeed, I would like to see the result.

Was playing a little more. Had missed having to measure target response on top of the IR drop across the coil. Wondering how much of a problem that might be.

[Teleno]

Was playing a little more. Had missed having to measure target response on top of the IR drop across the coil. Wondering how much of a problem that might be.

Targets during on-time can only be measured in a separate Rx coil. The Tx coil is as good as shorted.

[green]

Targets during on-time can only be measured in a separate Rx coil. The Tx coil is as good as shorted.

My error, thanks