But, I think it does matter. Otherwise, why would Minelab give you the option to change the assigned channel under the Expert Menu?

First, some background on why I think it matters (and that comes down to three incidents I've experienced), followed by the noise cancel procedure I use to select a channel.

Big deal? Certainly not if you are interested in digging clad quarters; everyone is gonna dig both of these. I like clad quarters as much as the next guy, but I saw this as a clue into the inner workings of the machine. Remember, TID is a function of both the transmit frequency, and the target metal. The software will take the transmit frequency, the response, crunch some numbers, and normalize it to a consistent response to make it easy for the user, so that when the transmit frequency is either X or Y, the TID will be the same, if the target is the same.

But, there is bound to be some rounding error in this process. For example, lets say everything normalizes to channel 9, and channel 9 transmits at 2KHz. When using channel 9, no computation is necessary, blast out 2KHz, and report to the user what you see.

However, maybe channel 2 transmits at 1.5KHz. If so, to normalize to the channel 9 experience, some math needs to be done, and its possible that there is rounding error in that math. If so, it may report 11-46 instead of 11-47 on the same target.

The fact that I was seeing this allowed me to speculate that each channel was causing the machine to send out different frequencies, and that perhaps the difference was material.

Note: I understand that the FBS technology sends out multiple frequencies. The above example is simplified. What matters to anyone looking for silver is the low frequency; the lower the better. If you know one channel is sending out a lower frequency than another, you want that channel if you are looking for deep silver.

So, with this discovery and the inference that each channel may be sending out a different low frequency, that led me to my second experience.

Eventually, I came to find a very deep silver in the wild, just on the edge of detectability. You know the sort, quick, coin-sized signal with that FE bounce, in the high zone. I was running channel 9 when I found it.

I pinpointed exactly where the target was, placed a plastic marker, and then went thru each channel (Expert Menu/Channel), to see if it made a difference.

It did: I could only hear the target when running channels 2, 6, 9, or 10. While other channels may have given a threshold null or perhaps an undiggable chirp or non-repeating signal, there was no mistaking that channels 2, 6, 9, and 10 gave the best signal. I repeated this test for quite a long time, and from different directions, to smooth out any randomness. The results were consistent. Some channels didn't even give a threshold null.

I had at least confirmed my hypothesis from above -- that, at least for this target in this dirt, channel mattered. I don't know if it is because of the transmit frequency, the software, or the phase of the moon, it matters.

I dug the target (it was a merc), and vowed that I would run on channels 2, 6, 9, or 10 going forward, no matter what.

Happily hunting on channel 9, I was pulling silver after silver from a massive honeyhole that was giving them up at 1 per half hour. Deep wheaties and clad on top of it meant the high tone (low frequency) targets were frequent, and there were very few shallow targets, due to, perhaps, the site being hunted by weaker machines/weaker detectorists.

Then suddenly it stopped. I kept gridding out of the hot zone, and couldn't pull more than a couple targets to save my life. Not even clad. There is no reason that this section of the site should have fewer targets than the section 10 feet over; I remembered from the aerial photos that this entire section contained an old baseball diamond, and that some dirt I was digging here was orange, consistent with dirt often found in baseball diamond infield baselines.

I got an iffy signal consistent with deep silver, but it turned out to be a clad dime at only 4 inches. Ouch. Something was terribly wrong.

I got another iffy signal, and decided to do a test. I tried every channel on this iffy signal, and all channels saw it loud and clear, except channel 9, my preferred channel, and the one I was running on earlier, and pulling silver after silver.

What was different? What was different was the orange dirt. Channel 9 was having a problem with the orange dirt, for whatever reason.

As an aside, I always run manual, and cranked, but noticed that the auto rec was very low in the orange dirt section (what I mean by "auto rec": what it shows you auto would run if you hit the right arrow key to go into auto mode or look at the auto bar). I noted this.

So, I hypothesized that the orange dirt had different mineralization characteristics than the normal dirt 10 feet over, and that target response was also a function of the mineralization of the dirt and the channel operating on that dirt.

I regridded that section on a different channel, and pulled targets that channel 9 missed. Evidence that this stuff matters, at least in this dirt.

However, in abnormal conditions, such as strange dirt or variable mineralization, one or more of the these channels may not be the best; some channels may not like particular dirt or mineralization, for whatever reason (I mean, heck, I'm an economist, not a physicist, so who knows the reason?, but my tests and the silver don't lie).

Note: the ostensible purpose for the noise cancel/channel selection function is to mitigate EMI. Therefore, it is possible local EMI will render the above calculus useless. This is most likely to happen in an urban setting -- you've got it all figured out that channel 9 may not like the dirt, so use channel 2, but that powerline may neuter channel 2. Ouch. In my experience, this is a very rare problem, and to use the noise cancel function to worry about EMI noise should be the lowest of priorities; do so only if there is an obvious EMI problem.

Finally, watch the auto rec and a sudden dropoff in deep targets. A sudden drop in either is a red flag that you may need to change the channel. I look at anything below 19 on the auto rec as a cause for concern (tho some dirt is just bad, and auto rec runs low all the time -- I'm talking about a sweet site where all the sudden drops for no reason).

First, I follow the manual's procedure, and do a noise cancel holding the coil 1 foot above the ground. This gives me a read of which channel is most favorable to the local EMI conditions. I note the channel, and hope it is 2, 6, 9, or 10.

Next, I do a noise cancel holding the coil flat to the ground. This gives me a read of which channel is most favorable for this dirt. If it is 2, 6, 9, or 10, I am happy and start detecting. If not, and the channel from the above step was 2, 6, 9, or 10, I change it to that channel, and start detecting. If neither were 2, 6, 9, or 10, I use what it gave me on the ground test, and do the above 2 steps in 10 minutes or so (under the logic that at this site, it may not matter, and if I get a different answer next time, use the best one for silver, and under the logic that the mineralization test is more important than the EMI test).

Then, throughout the detecting session, I watch the auto rec, and watch for a dropoff on deep targets. If I see problems with either (and by that I mean auto rec below 19; if running manual, hit the right arrow to see what the auto rec is, then remember to switch back to manual), I repeat the above to see if I can get a better channel. Note that some dirt is just bad, and auto rec will be bad regardless of channel, and I believe that this is the case most of the time, but my tests suggest that it matters some of the time. In some cases, the auto rec is giving me a clue that this channel doesn't like this dirt.

Finally, in my log, I note the channel that seems to work best at a given site, for next time, just so I know. Also, as an aside, I prefer 2 and 9 to 6 ans 10 when I have a choice (sometimes if the air likes 9, and the dirt likes 10, I'll choose 9), but all of that is further refinements of the model that go beyond the scope of this article (and should rightly be ignored if you are a newbie to all of this). Just notes for those who want to do more tests. That said, there is one site that just loved channel 6, and that is the channel I used to clean it out.

Will it help you, who knows? I think the dirt matters alot. Where I live, the mineralization is highly variable, and can be extreme. Moreover, most of the silver is at urban sites (duh, population density), meaning more EMI. So, it matters where I detect.

In other parts of the country, with cleaner dirt (and I've detected in these places; you get more depth in this cleaner dirt without thinking about this stuff), I think it matters less. As always, YMMV, and FWIW. (and HTH too, this is just my experience, and presented as food for thought).

Finally, I am not a physicist. I have no clue how metal detectors work. All of the above is speculation and observation on my part. Again, YMMV.

Its just food for thought and a recollection of my experiences. Make of it what you will, and do with it what you wish.