V3 Ground Filter Talk.

[Mike Hillis]

I haven't liked anything I've read about the V3 filters....The manual and the engineers have left this as a dark subject....So I have decided to turn the light on.

Ground filters are used to separate out and block the slower changing ground signal and pass the faster changing target signal. White’s engineers are using hertz to designate how fast a target signal needs to change in order to pass through the filter. They are measured in 'rate of change'.

In addition, they have provided two types of filters for each speed; a ‘Band Pass’ filter and a ‘High Pass’ filter.

A ‘Band Pass’ filter is exactly what it says it is. It will pass a certain signal “rate of change” that falls within or between the lower and upper ranges or boundaries that have been defined for that ‘band’. It will block all signals that fall below or above that Band’s range. It will pass ONLY the signals whose ‘rate of change’ falls inside the acceptance band. It is like using “Notch Accept” only instead of notching in a metal target’s phase response, you are notching in a particular signal “Rate of Change”.

A “High Pass” filter is different than the Band Pass filter the same way the notch feature is different from the discrimination feature. High Pass is like setting the Disc on your detector. Targets below the Disc Setting are blocked, targets above the Disc setting are reported. High Pass Filters block the signals whose ‘rate of change’ is less than the lower range defined for the filter and pass ALL signals whose ‘rate of change’ is above that range.

I would suggest that the upper end of all the available “Band Pass” filters are the same and that the filter speed selections allow you to raise or lower the lower limit of the Band Pass filter.

To say it another way, the Band Pass filter’s high limit is fixed. The filter speed selected moves the lower limit toward or away from that fixed upper limit.

I would suggest that a certain hertz High Pass filter and its related Band Pass filter share the same lower limit of the filter. The High Pass filter is only removing the upper limit of the Band Pass.

So…with the basics out of the way….let’s look at the filter selections…

5 kHz Band Pass = This filter is used for very low mineralized ground. Since there really isn’t a very big ground signal to block, this filter’s lower limit will allow very slow changing signals to pass. It has a high limit in place that a signal’s “rate of change” has to fall below. At this point…the mineralization effect of the ground does not significantly affect a target signal.

5 kHz High Pass = Same as the 5 kHz Band Pass except there is no upper limit.

7.5 kHz Band Pass = Now I’m starting to get some low to moderate ground response, more ground signal is received. At this point the ground signal is starting to have an effect on the target signal. This speed raises the lower limit of the Band Pass filter to compensate and requires a little faster “rate of change” from the target signal to report it as compared to the 5 kHz filter. It still has the same upper limit in place.

7.5 High Pass = Same as the 7.5 kHz Band Pass except there is no upper limit.

10 kHz Band Pass. = Now I’m into moderate to high ground mineralization. The ground response is significant. Surface irregularities can report as a metal target. The signal is degraded and weak signals are being masked by the ground signal. To compensate, this filter raises the lower limit even higher, requiring an even faster rate of change to pull the target signal out of the ground response.

10 kHz High Pass = Same as the 10 kHz Band Pass except there is no upper limit.

12 kHz Band Pass = High to very high ground mineralization. Weak signals are invisible and even moderaly deep signals will report as iron. Ground irregularities will report as metal targets. The target signal is highly degraded. This filter selection raises the lower limit to it’s maximum range to compensate. This is the narrowest Band Pass filter available.

12 kHz High Pass = Same as the 12 kHz Band Pass except there is no upper limit.

Next….How to select the right filter…….

HH
Mike

 

[Mike Hillis]

So after explainging how the filters work, all that is left is to explain how to use them.....

Yesterday I started off my testing in a woodchip play ground, then moved to a sand playground, then moved to turf. My normal progression for testing things metal detector. In the woodchip playground I had hardly any distortion of the Spectragraph signal as compared to bench testing results. I used a 5 kHz band pass filter because 1) I had no minerals to deal with, 2) it could handle both slow and fast coil sweeps, and 3) It was quieter than the High Pass option.

When I moved to the sand playground, the sand is much higher in mineralization than the woodchips. Even though I ground balanced to the sand, the higher surface irregularities on the sand gave off a target response. The 5kHz filter "rate of change" was slow low enough to allow the mineralization "rate of change" of these surface irregularities to report as if they were a metal target. There are two ways to deal with this. One way is to do a ground balance offset. A ground balance offset can remove the false repsonse of the surface irregularites. But the same mineralization that was causing the response at 5 kHz filtering was also degrading my spectragraph responses. They no longer matched the ideal response I had determined from my bench testing. So while a ground balance offset would remove the false target response it did nothing to improve the signal clarity. That required a filter change to 10 kHz and some coil sweep trials to see what was required to restore clarity to the spectragraph. The filter change also allowed a regular ground balance with no offsets.

That is the purpose of the ground filters.

Regarding sweep speed, the filter in use determines the slowest sweep speed you can use without degregating the spectragraph signal clarity.

When you have no to low iron mineraliztion, you use the 5 kHz filters. Which one to select depends on the EMI and perhaps your fastest coil sweep speed. As the minerals increase, you select the filter that helps you to maintain the best spectragraph signal clarity with the slowest sweep speed. If you tend to sweep fast to cover large areas, and slow down when you find a target I'd recommend that you use the High Pass filter option if EMI levels allow you to use this filter option.

HH
Mike

 

[ArizFlash]

Thanks Mike. Your message is better than the manual and more applicable than Carl Moreland's Advanced User Guide.

I've always found the 5 Hz band pass get its best depth when swept slow. I haven't paid a lot of attention to the Spectrograph display as it relates to ground minerals. So the minerals will also skew the sizing graph as well. I've often wondered why that happens as well as the polar plot.

 

[Ytcoinshooter]

Mike that is an excellent post and users can apply that information easily. This post should be a "sticky".
I've understood as the minerals / ground matrix response goes up higher filters need to be used. What you made clear for me is the high pass filters having the same "ceiling" or upper limit. The filter selection for my local area I worked on in my test plot patiently and methodically. I've had my V3i 4 1/2 years and getting the best filter selection for areas hunted was the first hurdle in obtaining deeper good targets.
Nice going !
HH - Bruce

 

[Mike Hillis]

Thank guys,
HH
Mike

 

[PostalTwo]

thats a great post mike! but i will say that im allways using the 5 filter due to the fact its the deepest by alot . my ground is very low minerailzed at about 1%. two weeks ago i took a trip with some friends to the mountains there i was dealing with 13% -93 iron mineralization. even in that medium minerailzed ground the 5 filter still yeilded the best results for me. in my test bed at home i found that the 10 filter just kills your depth. im sure u find the same in your testing . it would have to be some very hot ground for me to bump the filter up to 10 . thanks again for your informative write up it was a joy to read cheers

 

[gabbyhayes]

Thanks for the great info!
Well written, concise, and to the point.
I think your post is worth printing off and adding it to White's advanced posts book.
Good Job!

 

[liondogs]

Mike, great post. You have a knack for taking a complex idea and breaking it down in to simple terms that the average person can understand. Your post on filters is one of the best ones I have read on the subject. As a former F5 user I read all your posts on that detector and learned a great deal. I am glad you now have a V3i and look forward to many future posts about this detector as well.

 

[texjim]

Awesome thread Mike.Thanks..

 

[cypearl]

Great post Mike, it really clears my mind now on why I was using 5 band on a great program I found for use and when I used the coin and jewelry preset was deeper. Thanks again. V3i always found good targets though hardest to find for other detectors I have used.

 

[PostalTwo]

Great post Mike, it really clears my mind now on why I was using 5 band on a great program I found for use and when I used the coin and jewelry preset was deeper. Thanks again. V3i always found good targets though hardest to find for other detectors I have used.

interesting . Please share with us . What filter works best for u in your ground?

 

[cypearl]

Hi postal two, I was using 5 band filter most of the time cause I thought was deeper after reading a few posts, but I had to cover large areas so I was sweeping fast and I was wondering why only surface targets sounded good and changing to the coin and jewelry preset I have found deeper and smaller targets!! With V3i you need to be concentrated and do the work to tweak the settings, I am really happy with this detector cause it has a fast response to targets and close to each other others can't deferentiate. With 5 band I noticed works deep with slow speed but the interesting was smaller targets disappeared ! , at least sounding good and robust was absent. 10 band on coin jewellery gets tiny targets on surface and large deeper! My ground is medium mineralised with iron and copper bits at most.