Bandito, logically

[porkluvr]

Hi, I have a problem understanding the Bandito. From the schematics posted earlier in this forum, I can't logically see how a single target that rings the bell in "All-Metal" mode can also ring in "Non-Ferrous" mode. Here's why:

I'll try to make this simple: Non-Ferrous channel has two legs. Let's call Ground Balance leg GB, and Discriminate leg is D.
All-Metal channel is ummh, All-Metal. There's a bell downstream.

It looks like the "GB" and "D" legs combine to form the "Non-Ferrous" channel. Ignoring polarity, for now: Non-ferrous "high" is logically [GB AND D], through comparators IC8a AND IC8b. Now, remember inverting opamps IC10a and IC10b.

Except for about four(?) degrees offset, the all-metal clock is the same as the ground channel clock. But, GB and D have an inverter, and the all-metal channel signal path has no inverter. When a target causes GB high at IC10A-1, then All-Metal should be low at IC6B-7.

So, if a target causes (G AND D) high, and rings the "Non-Ferrous" bell, how can the same target still ring the "All-Metal" bell, since GB and All-Metal are essentially opposite each other?

I have built up the circuit in LT-Spice and run similations, but the only way I can make this thing work in a virtual panorama is to fudge, - bigtime. I may as well live in my car for all I can do to breadboard this thing, and I don't have the grasp of reflection theory to "fix" things in my computer.

Please somebody illuminate me!

[Qiaozhi]

Hi, I have a problem understanding the Bandito. From the schematics posted earlier in this forum, I can't logically see how a single target that rings the bell in "All-Metal" mode can also ring in "Non-Ferrous" mode. Here's why:

I'll try to make this simple: Non-Ferrous channel has two legs. Let's call Ground Balance leg GB, and Discriminate leg is D.
All-Metal channel is ummh, All-Metal. There's a bell downstream.

It looks like the "GB" and "D" legs combine to form the "Non-Ferrous" channel. Ignoring polarity, for now: Non-ferrous "high" is logically [GB AND D], through comparators IC8a AND IC8b. Now, remember inverting opamps IC10a and IC10b.

Except for about four(?) degrees offset, the all-metal clock is the same as the ground channel clock. But, GB and D have an inverter, and the all-metal channel signal path has no inverter. When a target causes GB high at IC10A-1, then All-Metal should be low at IC6B-7.

So, if a target causes (G AND D) high, and rings the "Non-Ferrous" bell, how can the same target still ring the "All-Metal" bell, since GB and All-Metal are essentially opposite each other?

I have built up the circuit in LT-Spice and run similations, but the only way I can make this thing work in a virtual panorama is to fudge, - bigtime. I may as well live in my car for all I can do to breadboard this thing, and I don't have the grasp of reflection theory to "fix" things in my computer.

Please somebody illuminate me!

OK - this is a little subtle, but I'll try to explain.
With this type of I.B. detector the GEB channel actually gives a positive response for all targets, whether they are ferrous or non-ferrous. Whereas the DISC channel gives a high signal for non-ferrous and a low for ferrous. The cross coupling of the GEB and DISC channels allows the low (ferrous) signal to cancel the high signal from the GEB channel that corresponds to a ferrous target. The ALL METAL channel (like the GEB) also gives a positive response for all targets whether they are ferrous or non-ferrous, so the inversion in the signal path doesn't matter. The ALL METAL channel is really a non-motion pinpoint mode of operation.

You are most likely having a problem with your SPICE simulation because the sample points for the GEB and DISC channels are incorrectly defined. The GEB channel must be set to sample at the zero-crossing point of the RX signal, and the DISC must be set to sample at the peak of the RX signal. In order to see what happens in the presence of a ferrous or non-ferrous target you then need to adjust the phase of the RX signal during the simulation.
Hope this make sense.

[porkluvr]

Qiaozhi,
I think I understand what you mean.

I believe I can set my phase relationships correctly, but what I cannot do is
similate the effect of search coil sweep. That might have something to do with why I have been inverting the GB clock to get my Non-Ferrous channel output! (Gee, do ya think?)

Since all of Tesoro's designs that I have come across use the same basic topology exhibited by the Bandito, I will take it on faith that the topology has some merit, and umh, I just can't fake it with LTSpice.

And thanks everybody who has contributed their reverse engineered (or otherwise) schematic, or even a cool web site. You know who you are.
Brilliant!

[Qiaozhi]

Qiaozhi,
I think I understand what you mean.

I believe I can set my phase relationships correctly, but what I cannot do is
similate the effect of search coil sweep. That might have something to do with why I have been inverting the GB clock to get my Non-Ferrous channel output! (Gee, do ya think?)

Since all of Tesoro's designs that I have come across use the same basic topology exhibited by the Bandito, I will take it on faith that the topology has some merit, and umh, I just can't fake it with LTSpice.

And thanks everybody who has contributed their reverse engineered (or otherwise) schematic, or even a cool web site. You know who you are.
Brilliant!

There are a couple of ways you could do this:

1. Add a third inductor to the search head to allow injection of the target signal. Then you could use a VCVS with a polynomial definition to create a multiplier. The multiplier could then be used to gate the target signal.

2. If you haven't actually modelled the TX and RX coils using coupled inductors, an alternative method would be to sum the multiplier output with the RX signal to create a composite signal.

In either case the result is that the RX signal changes in amplitude and phase at some point during the simulation, thus representing the case where a target is present under the coil.

I don't know how good you are at using SPICE, but if you're having difficulty then please post the netlist here and I'll try making the modifications.