Iron Bias Fe Versus Fe2

[Steve Herschbach]

Jeff and I had a couple PM messages back and forth about Iron Bias, and what it does. Jeff made mention he thought it took items with mixed ferrous and non-ferrous responses and reduced or eliminated the non-ferrous response, leaving only the ferrous response.

I usually set Iron Bias to 0 unless I forget and leave it at some default setting. I hunt full tones and between high 39 spikes and audio "tells" while in full tones I generally do not have much issue with ferrous tricking me, and so it has not been anything I have investigated much. The bottom line is I was feeling vague on exactly how this control acts in practice.

As we all know bottle caps with the normal FE settings had a habit of reading in the mid-teens near 13. I have also run into issues with large flakes of decomposed steel cans in old mining camps that read in the teens. I'm not worried too much about bottle caps, but in some old camps that rotted tin can debris can really mess an area up.

Minelab introduced the new FE2 setting that allows for more range on the control. Jeff's message made me wonder - is Iron Bias reducing the non-ferrous false signals, or increasing the ferrous signal? This short video shows the problems and the solution, and answers this question.

I start with a piece of tin steel that gives a non-ferrous reading around 15 in normal FE mode, and show that from FE0 to FE9 the target gives a clean mid-tone with no ferrous report of consequence. The old FE setting was ineffective on this stuff.

Then I go to FE2 and show that increasing the setting rapidly shifts from a non-ferrous reading to a ferrous reading. At 0 it's clean non-ferrous. Just going up to 2 makes it a mixed ferrous and nonferrous reading, at 3 it's mostly ferrous with some high tone mixed in, and at higher settings it is a solid ferrous reading of -7.

To me it looks simple. The control is variable, and a particular item can be adjusted to read solid non-ferrous, or to give mixed tones, or to read solid ferrous. It is not a matter of just decreasing the false signals. More like converting what were non-ferrous signals to ferrous signals in a continuous range.

This all goes back to the ferrous versus non-ferrous range having a full overlap, and much like how you can use the tone break to decide what items near target id 0 read ferrous and which do not, the Iron Bias offers a secondary control that does the same with these problematic items.

The part I do not show? There are good items that will do the same, and I need to try and find one. In theory there is a good item that will give a false ferrous if this setting is too high. Otherwise we would not need the control. Just set it to max and forget. So the real question here that I think remains is not about the ferrous. The control will clearly help with making ferrous items read ferrous. But what about the reverse? What non-ferrous items will be revealed by leaving this control at a low setting? I take it as a matter of faith these items exist and so I leave the control low at or near zero. But how common are they? Am I suffering needlessly? What are the risks of setting the control too high? Minimal risk, or high risk? The best way to find out may be to simply go detect, but take the time to double check items at the full FE2 setting of 9 to see how many might read ferrous when they are not.

Be patient - it might take a minute for the video to start once you hit go.

If somebody could make a perfect ferrous/non-ferrous filter, then we would not need the control. Many casual detectorists don’t need it at all either. The reality is the setting is a trade off. Higher eliminates more ferrous, and also runs more risk of missing the desired non-ferrous. Conversely, lower settings reduce the risk of missing desired targets, but you dig more trash.

So the question is pretty simple. Are you willing to let the engineer pick the best setting for you? Or would you like the ability to make that decision yourself? People hate digging trash, so I find presets to be too aggressive for me. I get very worried if I’m digging no ferrous at all. So I tend to set where I’m digging some, but not so much as to bother me. It gets me “iffy” finds others miss.

This is a simplified example. Iron bias far right gets all gold, but you dig all the ferrous also. Far left eliminates all ferrous, but loses all gold. A typical factory setting as illustrated might eliminate most ferrous, but at what cost?

It is interesting that the control is on,y available in multifrequency on the Equinox. The issue predates multifrequency, with variable bottle cap reject controls going back to analog detector days. It’s an issue inherent in the physics of the targets being detected. 

If anything this is a problem that was accentuated with the switch from concentric to DD coils. Concentric better separate flat ferrous targets than DD coils, so this was less an issue when concentric reigned supreme. The newest bottle caps are even worse also, being made of a much larger range of materials than the simple caps of old. It is not just bottle caps though, but decomposed flat roofing steel, old rusted can pieces, and more.

Why ferrous items can read non-ferrous

From Whites XLT manual (6.59 kHz single frequency):

BOTTLECAP REJECT

Adjusts how strongly the instrument rejects or breaks up on iron. Most starting programs use the minimum setting. As larger numbers are selected, more bias rejection against iron occurs.

The advantage of higher BOTTLECAP REJECT settings is that in high-trash areas more decisive iron rejection occurs. Trash becomes easier to identify by the broken sounds they produce.

The disadvantage of a high Bottlecap Rejection setting is if an iron target is close to a good metal, the high degree of bias against iron may cause the detector to cancel both responses. Another disadvantage is that all targets, iron and non-iron, tend to start sounding more broken at high levels of BOTTLECAP REJECT.

The operator needs to fine tune BOTTLECAP REJECT according to their preferences and the conditions being searched.

Note that according to White’s, setting this control too high can contribute to masking issues. It makes sense, as all we are doing is adjusting the tipping point on mixed target responses, and a ferrous target next to a non-ferrous target presents just that issue.

[Steve Herschbach]

Note that the Iron Bias function is only available when the Equinox is operating in multifrequency. The setting will not be available in single frequency modes. Use the Accept/Reject button to toggle between FE and FE2 once you have accessed the function.

Equinox Manual, page 52 (emphasis added):

All ferrous targets produce a combination of a ferrous and nonferrous response. Large ferrous targets can even present a stronger non-ferrous response. Also, a ferrous target adjacent to a nonferrous target can produce a similar response.

The Iron Bias Setting provides some control over the Target ID response. A lower Iron Bias setting will allow the natural response to dominate which means that the target is more likely to be classified as a non-ferrous target. A higher setting will increase the likelihood that the target is classified as iron.

In environments with dense iron trash, a higher Iron Bias is recommended in order to mask them. In areas where you do not want to miss any non-ferrous targets amongst iron trash, a lower setting is recommended. This will cause more ferrous targets to be detected and identified as desirable non-ferrous targets.

[Jeff McClendon]

Where iron bias gets iffy is on really deep iron/not iron targets and on really big iron targets (especially deeper ones) that will break through any attempt to discriminate them out with some major falsing being the result. I can get an on the surface railroad spike or large square nail to stop falsing on a setting of F2=7 to 9. However, any target relatively close to that spike will be almost completely masked.  Also, the deeper those big iron targets are the harder it is to get them to stop falsing which is why I raised the question about how deep should I expect the Equinox iron bias function to work effectively and on what sized targets.

[Steve Herschbach]

14 minutes ago, Jeff McClendon said:

how deep should I expect the Equinox iron bias function to work effectively and on what sized targets

Like most things I would expect that to be totally dependent on ground mineralization, and therefore highly variable. No matter what detector I use, my expectation is all functions break down with depth and any increase in mineralization. The question for me is most applicable to gold nuggets in mineralized ground, where my inherent distrust of ferrous filtering always comes to the fore.

I have seen countless ferrous signals turn out to be gold nuggets in high mineral ground. My fear is increasing the iron bias at all in such ground conditions all but insures this will occur. I will dig literally thousands of nails rather than risk seeing a large nugget at depth passed up as ferrous. A 10 ounce nugget will read ferrous right at the edge of detection depth, so size does not mean as much as signal strength. A small nugget in real bad ground can easily read ferrous, and adding enough depth to turn that big nugget signal into a faint reading risks having all that intervening soil create a ferrous masking effect.

In general I consider Iron Bias to be a last ditch function, left at zero unless I am forced to use it. By that I mean ready to walk away from a site due to false signals causing me to dig endless ferrous. In such situations it’s better to just crank the control up, and accept the losses for whatever they are. Better that than abandoning the location, which means no finds at all.

[midalake]

26 minutes ago, Steve Herschbach said:

In general I consider Iron Bias to be a last ditch function, left at zero unless I am forced to use it. By that I mean ready to walk away from a site due to false signals causing me to dig endless ferrous. In such situations it’s better to just crank the control up, and accept the losses for whatever they are. Better that than abandoning the location, which means no finds at all.

I am posting this from a different discussion, my words.

 

I can elaborate on IB. Because the 1,2 and 5 peso coin rings in ferrous from -3 to -6 I run Iron bias of 4-5. This keeps these coins in check for falsing. 

My other saved program that I use for checking "iffy" signals has an Iron bias of zero. 

Anytime I go to my other program I ask myself, am I digging this target or not?  Out of all the times I been undecided and then checking with the other program. NOT ONCE have I found a target of worth.  

Now there is not a huge amount of targets that change my mind when I look with the second program, but still an accurate summary of actual events. 

The Equinox is a very powerful discriminator. One needs to trust the machine! 

 

[Steve Herschbach]

The thing is, the vast majority of detectors I have ever used like to call those rotted can sections non-ferrous. It is a rare detector that can tune the stuff out halfway well, and none that have a specialized variable control for the function besides the standard disc notches. I’m kind of anxious now to give the Equinox a try at a highly polluted site and see how it goes. Never a free lunch though. Cranking the setting up will help shut up lots of the flat ferrous, but also increase the masking. Then you are right back to square one, digging the ferrous to reveal the hidden treasures. 

[midalake]

11 minutes ago, Steve Herschbach said:

Never a free lunch though. Cranking the setting up will help shut up lots of the flat ferrous, but also increase the masking. Then you are right back to square one, digging the ferrous to reveal the hidden treasures. 

Well for all reading, I only hunt salt beaches. So my results are geared toward this venue. I also run a recovery of "6" in my go to program. When I go to my double check program not only is the IB "0" but the recovery is "4".  I hunt a unique situation with ferrous coins being a very high number of daily targets. But yet my yield of gold seems to be unaffected with my settings. 

If I ever do find a target of value with the second program it will be a significant finding. 

On average I would estimate I do 20 or so secondary program looks a day.  

[GB_Amateur]

6 hours ago, Steve Herschbach said:

The part I do not show? There are good items that will do the same, and I need to try and find one. In theory there is a good item that will give a false ferrous if this setting is too high. Otherwise we would not need the control. Just set it to max and forget. So the real question here that I think remains is not about the ferrous. The control will clearly help with making ferrous items read ferrous. But what about the reverse? What non-ferrous items will be revealed by leaving this control at a low setting?

 

44 minutes ago, Steve Herschbach said:

Like most things I would expect that to be totally dependent on ground mineralization, and therefore highly variable. No matter what detector I use, my expectation is all functions break down with depth and any increase in mineralization.

IMO you can combine these two thoughts.  Here's my hyphothesis.  Iron bias doesn't affect surface non-ferrous targets, but as that item goes deeper the falsing (non-ferrous item sounding ferrous) starts to occur.  The higher the iron bias setting the shallower that crossover occurs (for a given ground mineralization). 

There may be competition between the target 'disappearing' completely due to being too far from the coil (i.e. too deep) and the target (not so deep as to be lost but) sounding completely ferrous.

Thus for deep non-ferrous you want the iron bias setting minimized.  That's where it matters.

I emphasize this is merely an hypothesis.  I've not proven it (yet).

[Chase Goldman]

Steve's test results are very interesting.  I never knew that F2 could be that effective because I have been afraid to crank it that high out of fear of masking.

As far as a no free lunch is concerned, my main concern has always been masking.  IB tends to counter the target separation effects of higher recovery speeds effectively mitigating the setting.  Put another way, you might have to crank recovery speed higher to overcome the target masking effects of IB, but there is probably a point of diminishing where depth becomes limited and exacerbated by the other downside of IB, where deeper non-ferrous on the edge of detection look ferrous.  This is similar to the masking effect but a little different.  As with most of these settings, the key is finding the sweet spot intermediate settings of IB and recovery speec that provides the best balance of these tradeoffs. That, of course, is dependent on search mode, ground conditions (e.g., mineralization), ferrous and non-ferrous junk density, and the primary targets of interest (gold vs. jewelry vs. coins vs. relics, mid-conductors vs. high conductors...) and their depth.  I guess that takes me back to my preferred middle-of-the-road IB setting of F2 = 4 to 6 as a good starting point.

 

[Steve Herschbach]

25 minutes ago, Chase Goldman said:

I never knew that F2 could be that effective because I have been afraid to crank it that high out of fear of masking.

For me I was surprised FE2 kicked in so quickly. FE has no effect run to the extremes. FE2 has an effect almost immediately, with the ferrous breaking up significantly at only FE2-2 The control is more aggressive than I anticipated.