My secret tip to ignore bottle cap
Yes — some bottle caps can produce a clean-looking Target ID. A second sweep from another angle can reveal a change that one direction missed, giving you another data point before you dig. It does not identify the target by itself, but it is a quick check worth adding to borderline signals.
Why a bottle cap can fool the number
Target ID numbers are built from how a coil's signal reflects off a piece of metal — conductivity, size, shape, and orientation all shape that reflection. A crimped or corroded bottle cap, especially one lying flat, can present a conductive footprint that lands right in the same range as a coin or a small piece of jewelry. The ID isn't lying to you; it's reporting the average of everything the coil sees on that one pass. The problem is a single sweep only samples the target from one direction, and metal doesn't always answer the same way twice.
The cross-sweep check
This is the low-effort move that pays for itself every outing: instead of accepting the number from one straight-line swing, pivot and sweep the same spot again from a different angle — roughly perpendicular to your first pass. Do this before you commit to digging.
- First sweep: note the ID and tone as usual.
- Second sweep, different angle: cross over the same spot from the side or at an angle instead of repeating the same line.
- Compare: some compact targets may hold a similar ID and tone across both directions, while an irregular cap or crimped foil may wobble, drop, or break up from the second angle.
A wobble is a clue, not a verdict. Consistency can raise confidence, while instability is a reason to slow down and consider that the hole may contain an irregular object or more than one target.
Comparing discrimination and notch behavior on borderline signals
Discrimination and notch settings both act as filters, but they answer slightly different questions when a signal is borderline like a shallow bottle cap.
| Approach | What it does | Risk on borderline signals |
|---|---|---|
| Discrimination (broad accept/reject) | Rejects a wide swath below your line | Can silently reject real targets that share a range with junk |
| Notch (narrow reject) | Cuts out one specific slice, accepts around it | Narrower gaps can still swallow a genuine target that happens to fall in that same slice |
| Cross-sweep confirmation | Doesn't change what's accepted or rejected — just re-checks the same physical target | Doesn't reject anything; it only adds information before you decide to dig |
Notice the cross-sweep isn't a machine setting at all — it's a technique that sits on top of whatever discrimination or notch approach you're running. That's exactly why it works as a universal habit regardless of your setup: it doesn't ask you to trust the number less, it asks you to gather one more data point before the number becomes your only input.
The habit that actually pays off
The peak takeaway is simple: a Target ID is one observation, and a cross-sweep is a second observation from another angle. Irregular trash may change its response, while some compact targets may stay steadier, but neither behavior is guaranteed. Use the check on borderline signals so one clean-looking number never becomes your only evidence.
Numbers are a starting point, not a verdict. Let the cross-sweep inform the decision; do not ask it to make the decision alone.
TL;DR: A clean Target ID does not guarantee a clean target. Before digging a borderline signal, cross-sweep it from another angle. A similar response can raise confidence; a wobble or drop calls for caution. Neither result proves what is in the ground.