Equinox Ground Balance

[Cal_Cobra]

If you read between the lines on the various search modes, Minelab seems to imply that the ground balancing is done automagically, regardless of whether you do a ground grab or not.

For example:

Park 1 Multi-IQ processes a lower frequency weighting of the multi-frequency signal, as well as using algorithms that maximise ground balancing for soil, to achieve the best signal to noise ratio.
Hence Park 1 is most suited for general detecting and coin hunting.

Park 2 Multi-IQ processes a higher frequency weighted multi-frequency signal while ground balancing for soil.

Field 1 Multi-IQ processes a lower frequency weighted multi-frequency signal, as well as using algorithms that maximise ground balancing for soil, to achieve the best signal to noise ratio. Hence being most suited for general detecting and coin hunting.

Field 2 Multi-IQ processes a higher frequency weighted multifrequency signal while ground balancing for soil.

Beach 1 Multi-IQ processes a low frequency weighted multi-frequency signal, and uses special algorithms to maximise ground balancing for salt.

Beach 2 MultiIQ processes a very low weighted multi-frequency combination, using the same algorithms as Beach 1 to maximise ground balancing for salt.

Not to be confused with auto tracking, but it appears that the Equinox is doing some ground balancing specific to each search mode irregardless of whether you do a ground grab (auto ground balance in EQ speak) or not. 

Not that I would skip doing a ground grab, but I find it interesting that Minelab seems to be implying they process the ground balancing differently for each search mode, even if you don't do it.

Steve do you see this as the case? 

[scoopjohnb]

It has a very forgiving ground balance point and at the same time a very correct ID. Compared to the single frequency VLFs that we are use to that is sort of a contradiction.

The CTX when ground balanced is spot on, deep. And slightly annoying when trying to do undulating terrain. It really likes everything just wright.

The Nox 800 with a ground grab feels like you are hitting it in the ball-park. And what those little Minelab electronic fairies are doing on the inside is super fast tuning it to the actual conditions. Funny thing is they don’t mistake faint signals for ground. Truly amazing!

Have had a few targets not ID-ing, but... it hit them with the wright tone. Something to think about.

[steveg]

The GB numbers are quite different, TN...the "2" modes balancing with a higher number in my dirt, than the "1" modes.  I had noticed this before, but didn't attribute anything significant to it, other than the balance is simply "different" at different frequencies/frequency combinations.  So?

Cal Cobra -- I've been thinking/pondering the very same things you are.  Look at Explorers; you simply do NOT "ground balance" them.  The "ground" is "handled" by the multi-frequency somehow -- my guess is using data from the different frequencies for "comparisons/differencing" in the algorithms.  Then along comes the CTX -- which is the same as the Explorers, EXCEPT, the manual tells us, in EXTREME ground, we "can" ground balance, "if we wish to."  But otherwise, "don't worry about it."  So now, along comes the Equinox, which seems to make the ability to ground balance a bit more "prominent,' (i.e. each mode has a ground balance setting), but curiously the balance number is set to default at "zero."  HMM.  And this, taken with the verbiage you pointed out above, which I've been pondering, all seems to suggest to me that the Equinox may not be all that much different than the CTX -- i.e. ground balance isn't all that important, except in more extreme ground.  ???  I don't know, but I sure would like an engineering answer to this.  A Minelab "treasure talk" blog on this from someone like Mark Lawrie would be AWESOME!

Steve

[Cal_Cobra]

Steve also consider that their quick start guide, and the Quick Start procedure on page 11 of the manual gives these four steps:

1) Turn On

2) Select a Detect Mode

3) Noise Cancel

4) Begin Detecting!

Thats Minelabs words, and exactly what I did my first couple of hunts, and the Equinox performed perfectly.  TID was solid, machine was quite, plenty of power as it detected deep targets.  So something is going on under the hood that almost makes me wonder if it's at all necessary to GB at all (not that I mind, mind you, it's very simple and I'm used to doing it on my other detectors).  Just something that struck my curiosity more than anything, especially as you noted their previous Multi frequency machines did it on the fly.  

[cladzilla]

I tried the turn on and detect the first few times out and it was a chattering messing. I detect in deep south gumbo clay and once I performed a GB it was very quiet, almost thought it was not working. My finds got much better and deeper after that. I think it is very dependent on soil etc...

 

[Chase Goldman]

The quick start guide also says in the inset note to GB if you are experiencing excessive ground noise.

I totally agree that despite the fact that each mode grabs a widely different number on the same patch of ground, possibly indicative of different GB algorithms, I also noticed that the ground grab appears to not be a critical component of detector performance, per se, (compared to other detectors I've used) as it really just incrementally improves signal to noise ratio (something Steve H has been emphasuzing a lot lately btw for this and other adjustments).  I also agree that another way to look at it is that Equinox is indeed more forgiving of a less than exact GB.

[cladzilla]

1 hour ago, Chase Goldman said:

The quick start guide also says in the inset note to GB if you are experiencing excessive ground noise.

I totally agree that despite the fact that each mode grabs a widely different number on the same patch of ground, possibly indicative of different GB algorithms, I also noticed that the ground grab appears to not be a critical component of detector performance, per se, (compared to other detectors I've used) as it really just incrementally improves signal to noise ratio (something Steve H has been emphasuzing a lot lately btw for this and other adjustments).  I also agree that another way to look at it is that Equinox is indeed more forgiving of a less than exact GB.

Yes even as it was chirping and chipping, signals from actual targets still came through clearly. I just have a lot less fun if my ears are getting bombarded all the time :-)

[Mtwolf2270]

In my experience with the machine in multi here in moderately mineralized ground of east Tennessee as well as in super hot spots in Culpeper, va. That while sure it will work and find targets even not ground balanced what you will find by utlizing the auto ground balance and auto tracking enabled is that you can keep the machine insanely stable even at a higher  sensitivity if you choose. 

However, this is a super hot machine even at stock sensitivity.  By optimizing your ground balance not only will the machine be more stable but it will also have more accurate id and depth gauge will be much more accurate. 

You also eliminate most of the flase targets or can here deeper peeps in heavy iron. I also have found that reducing the sound of the iron to a background level and increasing the non ferrous tons running in all metal you can really cherry pick, even in  hot soil with heavy iron trash.

So, even though the manual may imply you don't have to ground balance which is true you will get a better overall experience if you do especially if your ground is hot. I truly feel that is site specific and you should treat the gb as you do the mode select the best settings for the location.

[bado1]

I agree with steveg's thoughts. Just like the other multi frequency units the GB isn't as critical as SF in most ground.

I've only run in to a couple of places where the machine got "out of balance" and had to change it from the factory presents ( or where it auto GBed) in that mode. Swinging a GPX a lot gives you a good ear for that slightly out of balance state. I've found that while being in multi if the machine becomes unbalanced, I go in to the GB mode and manually adjust the GB setting just a number or two until the machine is stable again. It doesn't take much. 

Kind of like the auto tune feature on the GPX. You let the machine find the most quiet frequency first and, after a while,  it may get a little noisy. Then usually all that is required is to go in and adjust up or down a coulple of numbers and she's quiet again.

Hope that made sense.

Dean

[Steve Herschbach]

On 3/28/2018 at 11:43 AM, Cal_Cobra said:

If you read between the lines on the various search modes, Minelab seems to imply that the ground balancing is done automagically, regardless of whether you do a ground grab or not.

..........

Not that I would skip doing a ground grab, but I find it interesting that Minelab seems to be implying they process the ground balancing differently for each search mode, even if you don't do it.

Steve do you see this as the case? 

Lots of great commentary here, and I apologize for being late to the game since I was addressed specifically.

What we have at work here is something Minelab is never going to discuss in great detail. There really is some magic at work in a Minelab multifrequency detector, and for a very long time almost everyone has been missing what is really going on. The focus is always on frequencies, and more specifically transmitted frequencies. Minelab has catered to all this with their marketing touting the number of frequencies. Good marketing but also misdirection to protect proprietary information.

The fact is transmitted frequencies is only a small part of the big picture, and some competitors have called Minelab out on this by pointing out that detectors can transmit all kinds of frequencies. In a poke at Minelab Garrett was quoting the Garrett Infinium as transmitting 96 frequencies, which is sort of true as a PI detector is like a transmitting shotgun.

Anyone can scope detectors to determine what they transmit, and if that is all there was to it the Chinese etc. would have ripped Minelab off ages ago. As it is they can only make detectors that look like a Minelab. The real secret is in what frequencies the detector is receiving and even more importantly yet, what sort of signal comparison and algorithms are being applied to the received signals. All the real magic occurs on the receiving end and very specifically in the signal processing.

Now for those that don't know it Minelab founder Bruce Candy* actually is an expert in audio processing and holds patents in that field as well as having founded another company, Halcro, that specializes in distortion free audio amplifiers.

And that my friends is the Minelab secret. Distortion free audio amplification. Using signal processing technology to eliminate as much unwanted signal (distortion) as possible and then amplifying the desired signal. It is very telling that the Minelab BBS patent expired long ago and yet we have not yet seen a third party BBS machine. Why? Because the patent does not reveal the exact processing methodology and so it is not a matter of just reading the patent and making a BBS detector.

So what's going on in there?

Single frequency detectors really do treat ground balance as another sort of discrimination method. Ground signals are part of the full target id spread or phase chart (see below). Basically the ground signal is determined by some method, like bouncing the coil over the ground (ground grab), and then this ground signal simply subtracted from all other signals. The best explanation and diagram I have ever seen of this was done by Carl Moreland on pages 1-3 and 1-4 of the V3i Advanced User's Guide. Pretty simple really.

The problem is that a single frequency can only set one ground balance point at a time. It can balance to minerals for instance, or it can balance to saltwater. As can be seen in the chart above from the link above, soil and salt normally read in completely different parts of the phase chart.

Fisher and Minelab both solved this problem at nearly the same time so I will pass on who was first. The Fisher CZ-6 and Minelab Sovereign both came out in 1991 and both employ multifrequency to solve the ground versus salt problem that foiled single frequency detectors on saltwater beaches up until then. Multiple frequencies can have multiple ground balance points and this is one of the major advantages of multifrequency, and why multifrequency rules in saltwater detecting. For a detailed explanation of how this is done see the three videos below where White's engineer Mark Rowan explains all this - really fabulous stuff. The real meat starts in video 8 at the 4 minute mark but video 7 does have some parts worth seeing. FYI the DFX employed multifrequency technology licensed by White’s from Minelab. My belief is Minelab had already come up with better stuff in BBS and so had no problem licensing a more basic version to White’s. Therefore studying the DFX does have bearing on Minelab’s thought processes regarding multifrequency.

What all this means is that Minelab is using frequency comparison and signal analysis to do most of the heavy lifting of ground rejection/ground balance. It just happens automatically as part of the signal processing. This is why when the Minelab Explorer came out there was no ground balance control. In most ground there is no need to ground balance because the processing already removes the ground signal.

However, it is not enough for the worst ground. This great article by Nenad Lonic explains the differences of the BBS/FBS detectors on bad ground. Later models added the ability to ground balance for extreme ground conditions.

After all that you all should be seeing the picture. Minelab is not discouraging people from ground balancing per se. It's just that in mild to moderate ground there is no need - the signal processing already does the trick. Yet in more mineralized ground you absolutely should be doing at least a ground grab. I always ground balance my Equinox because I am almost always on bad ground it seems. No matter what always ground balance anytime you go to a single frequency mode because you no longer have the advantage of the multifrequency processing.

I have made mention several times that every search profile has its own ground balance and so it is critical in bad ground to ground balance each search profile separately. Again, in milder ground just leave it be. How to know when to ground balance? In any mode, hit the horseshoe button and detect over metal free ground. If you get lots of puttering in the -9 and -8 and possibly -7 area that is the ground signal as is shown in the phase chart above. This is another reason why running full tones (no items rejected) can be beneficial as you do hear the ground signal. You might eliminate that signal by dropping the sensitivity a notch or two. But if it is persistent it does indicate you probably should be ground balancing the detector. Even when you notch out this region the ground signal is still there and so trying to get that negative range to settle down is the best bet for most people to help eliminate ground masking effects. For shallower targets you might also ground balance, then jack the sensitivity up and block out the negative numbers. This will make shallow targets like perhaps small gold nuggets pop but does risk deeper items going missed due to ground masking.

Ground balance numbers in single frequency detectors are normally related directly to the phase chart above but it is critical that people know that a ground balance number does not tell you how mineralized the ground is. I have written a very detailed article on this subject. The reality there is that the phase chart and ground balance numbers only really equate for a single frequency detector because it should be obvious that if a multifrequency detector is balancing to the ground and to saltwater there are actually two ground balance points. How do you represent that in a meaningful way with one number? This goes back to the fact that Minelab is not really working directly with each frequency but instead with post processed "channels" that represent the information streams from multiple frequencies. The latest Treasure Talk blog on Multi-IQ makes this quite clear.

This means that the various search profiles and even the various frequency selections all can produce quite different ground balance numbers on the same square foot of dirt. The numbers are tied more to the underlying signal processing and there is no attempt being made to "normalize" the ground balance numbers. This is why in bad ground you must ground balance each profile separately. The numbers to some degree are an arbitrary construct by Minelab and people trying to compare them as an indication of ground conditions are at a minimum going to have to relate the exact modes/settings in use for any such comparison to have any validity at all. And frankly I don't think even then there is any real point in comparing ground balance numbers. Again, just checking ground feedback with all target id numbers set to accept (all metal horseshoe in effect) will tell you more about ground or salt mineralization than ground balance numbers.

For more on detector frequency use see my article on Selectable Frequency And Multiple Frequency

* by Bruce Candy:
• Co-founder of Minelab.
• Pre-Minelab: designed advanced communication electronics (linear HF transmitters, VHF radar transmitters and receivers, ultra fast-frequency hopping etc), ultrasonic cleaners, fast photon counters, light detection.
• Designed concepts, analogue electronics and discriminator algorithms of Minelab detector (e.g. GS15000, GT/FT/XT. Eureka Gold series, Musketeer, Sovereign, PI units, Explorer series, Excalibur).
• Designed Halcro audio amplifiers.
• Holds patents in metal detecting and audio fields.