Minelab SD/GP and Garrett Infinium Dual Tones

[Anonymous]

Hi,
Just some rambling thoughts on a little discussed subject. I have considerable use both on Minelab SD/GP detectors and the Garrett Infinium. Both have "dual-tone" responses on various targets. Most gold items plus many other items like aluminum, foil, and wire-like steel have a high tone response. Most large steel or iron targets, plus some very large gold nuggets, and many coins, have a low tone response.
Bruce Candy explained it like this in a magazine article on the Minelab SD2200D...
"The Super-Detector series has two channels which contain "nulls" to specific target conductivities. Figure 1 shows a graph of the response of the SD2200D to target conductivity. In this graph, a positive response value means that when a target approaches the coil, the pitch of the audio tone will decrease (increase for the target moving away from the coil), and a negative value means that when a target approaches the coil the pitch of the tone will increase (decrease for targets moving away from the coil). The graph of channel 2 passes once through zero, marked as Ch2-zero. For targets with conductivities of this value, channel 2 will be very insensitive. However, in the vicinity of target conductivities near Ch2-zero, channel 1 exhibits substantial sensitivity. Because the SD series always relays the stronger signal to the audio in the "Both" mode, there is high sensitivity in at least one channel for all targets (except for the extremely high or low conductivity targets which lie outside the useful gold range)."
Unfortunately, the copy of the article that I have does not include this graph! I sure would like to get a copy.
This dual tone aspect of the SD/GP detectors has been little mentioned by Minelab. However, Garrett siezed on it as being a new "feature" in the Garrett Infinium, and tout it as "dual-tone discrimination". But as we can see it is not a new thing, and perhaps other PI units have also had this "feature".
Although large enough nuggets may give a low tone, I have found that these must be very rare in areas I hunt. Almost every nugget I have dug gave a high tone response. Low tones are almost always iron. I usually dig all targets, but I have employed this "feature" in high trash areas to ignore large iron signals. Not that I amy have missed a large nugget doing this, but from my experience in my locations this ahs been a good bet based on the items I've recovered. This seems to work as well with mono coils as with DD coils, unlike the Minelab iron discrimination feature.
The Infinium does much the same with it's system. I've used it for more than nugget detecting, and so know more about where the targets fall. If you use a normal VLF discrimination scale as a reference, items from zinc penny and below give the high tone response, and copper pennies, dimes, and quaters give the low tone response.
With both brands many trash items, particularly wire-like steel, you also get the high tone response. So while digging high tones does get the gold, it certainly does not get rid of all the trash, especially when jewelry detecting. But for nugget detecting it works pretty well.
Minelab has covered a possible "hole" in the target response by using two "channels" or two separate overlapping pulse trains to get better overall target response. There is no information to indicate that Garret does this, but I do not know that they do not. I suspect not, but I could be wrong. If not, then certain "good" targets may not respond well or at all on the Infinium.
Anyway, I do like the "feature" as it does give at least some indication of the nature of the target. When nugget detecting I sometimes get into areas with lots of cans or other large steel items, and I've found switching to digging just high tone responses in these areas beats just walking away in frustration. And in certain situations I have found I can do alright coin detecting by digging the low tones only and getting copper pennies, dimes, and quarters while eliminating most trash, zinc pennies, and nickels. This only works if large iron targets are not plentiful, however.
One thing I did not try with the Minelabs was to set Channel 1 only and check a target then set Channel 2 only and check it again. In theory, since there are two different "null" points you actually break targets down into three categories -
1. Targets that have a high tone on both Channels
2. Targets that have a low tone on both Channels
3. Targets that are high tone on one Channel and low tone on the other.
So you get low conductive, high conductive, and mid-conductive indications.
Garrett told me of something similar with the Infinium. Advancing the disc control also apparently shifts the point at which the tones divide to some degree -
"The audio tones are not independent of the discrimination setting. For most targets it appears as if the audio tones are independent of the discrimination setting (i.e. the tones do not change polarity, they only change in amplitude as you adjust discrim). However, for targets that
have conductivities in the transitional region (i.e. between poor conductors and good conductors), you'll notice that as you increase discrimination a
target that initially read low/high may invert to high/low; thereby, providing an additional clue as to the targets conductivity."
I need to play with this more to determine just how much "spread" there is in that mid-conductive range.
So we have the makings of a simple discrimination system with three target id "zones". Presumably adding another "Channel" with another "null" would allow for another comparison point whereby target conductivity could be further determined.
Anyway, I'm no pulse expert. These are just some thoughts derived from my end use of the units and the responses I've observed in the field. And maybe something for those working on pulse discrimination to chew on. Any thoughts?
Steve Herschbach

 

[Anonymous]

Hi Steve,
I have written a little about what can happen when ground balance is introduced on a PI, which could account for the different signals you have mentioned.
Typically, the ground signal extends out to somewhere less than 30 usec delay. By this, I mean that if you could turn the delay up on your detector to 30 usec or so, then most if not all of the ground signal would be gone.
Now, if the delay were set that late, then most small nuggets will be all but gone also, since smaller nuggets have a very short delay. So, that really isn't an option for a serious gold hunter.
Now, what does this mean and how does it apply to your discussion of the two channels and the different signals? I will get to that in a minumte.
One more thing to note is the ground signal, iron signals, and gold signals do not decay at the same rate. So, if a detector has the ground balance feature and is set to eliminate the ground signal, then different things will occur to iron, gold or any other metal signals that have a decay time longer than the ground response.
What happens is this, at a typical ground balance setting, the ground will produce no response, larger iron objects will typically produce a positive response and most or at least some larger gold will produce a negative response. (I don't have any really large nuggets to use for testing). So, at this point, one can tell many iron objects from larger gold objects. Now, one can take these signals, and change or invert the signals or do whatever they want to produce the desired audio output. One could even just light different colored LED's and produce a positive response regardless.
This situation will hold true for any ground balance technique where a later sample is taken, amplified, and subtracted from an earlier one to produce the ground elimination AND the target signal delays extend past the ground balance setting.
Now, most small nugget signals and small iron objects, such as thin wire, small pieces of old rusty cans, etc, will produce a response that will disappear before the ground balance (GB) setting. In other words, their decay time is less than that of the ground. So, they will only produce a positive response, regardless of the GB setting.
If two different channels are employed using different delays for their main sample and subtract signals, the ratios or amount of subtract (amount of amplification necessary for the subtraction) signal required for ground balance changes. This will alter the point at which certain objects begin to produce negative responses.
Now, one more thing, at the point of ground balance, then any gold signal that has the same effective response time as the ground will be eliminated right along with the ground signal. In other words, some gold nuggets will have an equal subtract signal so they will not produce any signal. Fortunately, since the two ground balance signals have different "null" points, different nuggets are eliminated. Stated a little differently, nuggets that are eliminated on one channel, will be detected on the other channel.
The problem with what was just stated about positive and negative signals from targets having a decay time longer than the ground balance is what happens when there is very little ground signal to eliminate. When this happens the subtract process is different and there may not be any negative responses simply because the subtract process didn't produce them Thus, both large iron and large gold may also produce positive responses.
If all this sounds confusing, it is. It is one reason that would explain just why certain nuggets produce a very poor signal at one area and a strong signal at another. Simply changing the ground balance can make a big difference in the signal response of certain nuggets.
Now, to use the two channels a little differently, one could check a target with one channel and then the other. If there was a significant difference i in signal between the two channels, then the probability of the target is gold or other non ferrous item is high.
I hope you do some experimenting with the null points of the two channels. If you can adjust the ground balance of each channel, it would be interesting to see just how this affects some gold targets. What I would expect is this; at some specific ground balance, a nugget, maybe in the 1/4 oz range will be "nulled" right along with the ground. Several nuggets both larger and smaller than the 1/4 oz should be tried. The reason is, the null point can vary due to the size, shape, or composition of the nugget.
Now, this "null" of the gold signal will not occur like one might expect. The gold might produce a strong response as the coil is passed very close over it and then as the coil is raised and passed over the nugget again, you may see the nugget suddenly just disappear rather than having the signal fade as the coil is raised. So, it may take playing with GB and making multiple passes at different heights above the gold to see just how the GB can affect the signal from the gold.
The point I am trying to make is, this nulling of gold is quite deceiving and somewhat difficult to find.
Once the null point is found, then adjusting the ground balance, may cause this nugget to disapper at a different height or continue to produce a positive response regardless of the coil height. Adjust the ground balance the other way, and the nugget may begin to produce a negative response. The logical channel to try is the one most sensitive to small gold. However, both channels should display somewhat similar results.
Let me know if you try this. I would appreciate hearing of the results. It will give me a little better idea of just what ML is doing.
If you experiment with both the Infinium and the ML, you might see some differences. Any differences might just tell what Garrett is doing also. I would be interested in the Infinium's tests also.
As for a zoned disc mode, the technique works fine for the larger gold. However, the challenge is the smaller stuff which is far more common, and trying to separate the small gold from the small iron objects becomes more difficult. There are some tricks being tried to address this situation also.
Reg

 

[Anonymous]

Hi Reg,
So what you are saying is that this "dual-tone discrimination" is directly related to PI ground balancing? That makes perfect sense. I was wondering if ground conditions would affect to nature of the tones produced, and I see now that it does, at least in theory.
I do plan on playing with this more. I've been an avid detectorist since 1972. I've taken a particular interest in PI detectors recently as VLF technology seems about maxed out. I'm looking to PI for new frontiers in detecting, with my particular interest being ground balancing PI units. But for PI to really get popular with the masses more needs to be done with PI discrimination. I put up with digging junk as the price I pay for superior PI performance, but even I shy from certain uses just because of the large amount of trash I will have to excavate with a PI detector. Then it's back to a VLF detector.
But it looks like I need to get a Minelab SD2100 to really be able to explore this side-effect of PI ground balancing more. My GP 3000 and Infinium both can being automatically ground balanced or set at a "fixed" setting, but both lack manual ground balance controls. I've kind of wanted an SD2100 as a backup unit anyway, so this may be my excuse to get one.
Many areas simply do not need PI ground balancing. That is why so many common PI units work just fine without it. If that is the case, it seems I could use the two-channel manual ground balance controls on an SD2100 more as discrimination controls than ground balance controls. I'll round up an S2100 and do some experimenting with it. I'll let you know any results I come up with. But with winter nearly here in Alaska I'm going to be pretty much shut down for awhile.
Thanks for your lucid explanations of how this stuff works. I'm really only interested in results, but I have found knowing how my detectors work aids in my having successful results!
Steve Herschbach

 

[Anonymous]

Hi Steve,
Sorry about the long explanation but I really didn't know any other way to point out what CAN happen when a certain technique is used.
Dual tone disc could be tied directly with ground balance, or it could have its own circuitry basically perfoming a similar ratio technique. I just don't know what ML or Garrett are doing.
I just know that when the subtract process is used to produce a ground balance and the ratio is such to ignore certain strong ground conditions, the signal reversal of some targets occurs. This method could be used for disc purposes also.
The main point I was trying to make besides the disc feature was the fact that the ground balance technique could have a direct bearing on the depth of detection capabilities of certain nuggets. I say "could" because I know it does using the specific technique I use for GB but I don't know the exact technique used by ML or Garrett. I suspect they are using something similar, but don't know that for a fact.
What I have found is I can vary the maximum depth of detection of certain nuggets considerably by just adjusting the ground balance on my PI.
Now, if manufacturers are using something similar, then it would explain why some nuggets are missed by one detector and found by another. It could also explain why, at times, one coil might seem to perform better than another.
Most people do not want to know or need to know the technical aspects involved. But, I sure want to know if such aspects could have a bearing on what I might be able to detect, or more importantly what I might miss.
Reg

 

[Anonymous]

Hi Steve,
I got a chance to play around with an SD 2100 for a while and tried to figure out their ground balance. The tests I made were quite simple just using a piece of basalt and a few nuggets in the 1/4 oz range. All testing was done using a mono coil and a simple air test where I just passed the nugget by the coil.
Now, this is what I have determined based upon what I was able to notice during the testing.
ML's ground balance controls are very restrictive, meaning they are quite limited in their range. This makes sense considering the primary reason for them is to eliminate ground/basalt signals. So, there is no need to have an expanded range.
Now, I mainly concentrated on testing just using the above size nuggets since they were the most likely to be affected. Some would produce only a positive response, some only a negative response, but I was lucky enough to find one nugget that almost disappeared when I would adjust the balance control through its entire range. I could increase or decrease the detection range more on this nugget than any other by adjusting the balance. In fact, on one channel, I could almost make the nugget disappear, but not quite. I could alter the detection range by a couple of inches though.
My conclusion is, the subtract method for ground balance is the technique used. Since the GB adjustment range is so restrictive, it will be tricky to find just the right nugget to display what can happen.
Now, I would be interested to see what happens with a Garrett.
Reg

 

[Anonymous]

Hi Reg,
Thanks for all the info. I will let you know if I observe anything of interest with my units.
Steve Herschbach

 

[Anonymous]

Dear Steve,
I was reading this post you placed on the PI forum.
In this posting, you stated that minelab uses two channels in the SD series detectors,
but didn't know if Garrett did this with the Infinium. You also said that if the Infinium
didn't have a dual channel it would possibly have a hole in the target detection.
I can honestly say that the Infinium definetely has more than one channel of
operation as I was involved with the development of the unit.
If you do an air test with different conductivity targets in the gold range, you will also
find that the Infinium has a flatter response than the SD,'s.
It was found that on certain conductivity targets, there was a drop of on depth with
the Sd's and this was compensated for with the Infinium.
Hopefully this information will be of some use.
regards,
Phil. Gold Mining Centre. mail@goldminingcentre.com.au