Find's Treasure Forums

Welcome to Find's Treasure Forums, Guests!

You are viewing this forums as a guest which limits you to read only status.

Only registered members may post stories, questions, classifieds, reply to other posts, contact other members using built in messaging and use many other features found on these forums.

Why not register and join us today? It's free! (We don't share your email addresses with anyone.) We keep email addresses of our users to protect them and others from bad people posting things they shouldn't.

Click here to register!



Need Support Help?

Cannot log in?, click here to have new password emailed to you

Need help with methodology/logic

A

Anonymous

Guest
Hello to all,
I recently posted some suggestions (on the Finder's Forum) relating to how one might compare the relative depth performance of two detectors, specifically Minelab's GP Extreme and SD2200D. I intended the suggestion to relate to those people who had one of each unit available to test at the same time, but rereading my posts perhaps I didn't convey that very well. Anyway, I was wondering if anyone would care to look over my posts (detailed below) and tell me where, and if, my thinking is flawed. I try to keep as many variables as posible constant in the proposed comparison test, so that (hopefully) only each detector's electronics is a variable. I also suggested that if the results are presented in relative terms, the average of several independent comparisons should provide an indicator of overall relative performance of the detector types.
I am becoming concerned that I am missing something very obvious; didn't explain things very well, or have gone mad <IMG SRC="/forums/images/smile.gif" BORDER=0 ALT=":)">. I figure this forum is probably the most appropriate place to figure out which it is.
Any comment would be greatly appreciated, as I'm keen to learn.
Thanks in advance to all who respond. The posts are detailed below,
Cheers.
Lachlan.
Post No. 1
Hello to all,
There is still alot of debate about the relative depth performance of GPs and Sds. Most of us do our own tests and invariably we come up with different results. Unfortunately, I believe that some of the differences that are arising in testing are due to the methodology being employed.
I have worked in scientific research for the last 20 odd years and would like to share some of my knowledge of research methodology to help anyone that is interested undertake useful comparisons of GP and SD performance. I also hope that perhaps some of the time-consuming testing that many of the forum's contributors willingly undertake, and share, will end up in a more helpful and consistent form.
The problem with many of the depth comparison tests is that they are not very well constrained. This means that many of the tests are examining more than just depth performance. For example, suppose two people, one with a GP and the other with a 2200, individually attempt to detect the same test target (presume the detectors are fitted with the same size coil and battery). Scientifically, this is not a well designed test because there are several varibles being tested at the same time. In this case, differences in relative performance may arise from variability in the performance of each detector's coil, headphones, electrical connectors, and battery. In addition the acuity of each indvidiuals hearing is also a variable.
In a well designed test only the detector electronics (i.e., the control box) should be tested. Everything else should be kept as constant as possible. This is most easily achieved by using one operator and detector (stem, coil, battery etc) and exchanging just the control box between tests. That way the ears, headphones, battery, cables and coil are exactly the same for each test. The only variable is whether the SD or GP control box is attached. (Obviously, a battery change would be required for a test of the 12v system, as it is incompatible with the GP.)
Another factor that I believe may be of consideration in GP vs SD tests is the ground balance drift that occurs in the GP after the unit is switched on (Minelab warns that the ground balance will likely drift for a couple of minutes after switch-on). In many tests, GPs will have been tested over the target within seconds of being switched on. This isn't an ideal time to test, as this is during the drift period. In one of my own tests I had my GP turned on for some time and could easily hear a test signal. I switched the unit off. Then immediately after turning it back on couldn't hear the signal. I believe that occurrance may have been due to ground balance drift. So, there are lots of things to consider in comparing SDs and GPs.
I tested my own GP (buried target test) against a 2200 on a 10 gram nugget in quiet ground. I used the method I recommended above. I found the GP had about a 20% depth advantage on that nugget. The two controls that seemed to affect sensitivity the most are the threshold and signal controls. I run the threshold at 12 to 1 o'clock and signal at 80 percent. Overall, I am happy with my GP's performance and have confirmed it has a distinct advantge over my 2200 on nuggets up to 10 grams in size.
Cheers
Lachlan
Post No. 2
Hi Eric,
Would you care to comment, criticise, or make suggestions about what I have written below.
In an earlier post I proposed a test that members of the forum might use to examine the relative performance of SD and GP detectors. I used the word "relative" carefully in that post, as absolute tests are useless for the purposes of comparison given the varibles such as target cross-sectional area, and ground-type (and gold content of the target too, it seems).
Given that most will be interested in the relative performance of these detectors, and in a variety of ground types, would you agree that a number of independent tests of the type I described should build up a picture of the overall relative performance difference between the two?
I believe that as long as the results are always expressed in relative terms they should be useful. In each test, the ground type and target composition/size will be largely irrelevant because these conditions will be the same for each detector and the results will be expressed in relative terms. Thus with enough tests it should be possible to obtain a broad indication of the relative performance difference between the two detectors across a range of ground-types and target "sizes". Obviously, with such an approach any specific differences between ground-types and target "sizes" will be averaged out. Still, an idea of the average relative advantage that might be expected across all types of ground and target "sizes" would be of interest to many.
To all reading this, an important omission (and I apologise) from my previous post was that the target was not disturbed between the testing of each control box. This meant that the target maintained a constant orientation, and thus presented the same cross-sectional area to the detection field in each test. The testing process first involved burying the target to a depth where it was just detectable with the 2200. Then with the GP control box attached the target was located and the coil sweep raised above the ground until the the target again became just detectable.
Cheers
Lachlan
 
Hello Lachlan,
I saw your post on the Finders Forum and was taking a little time to consider it before replying. Metal detectors fall into that category of electronics where so much is subjective. Much like Hi-Fi. This is the reason why there are so many models on the market; what suits one person is totally wrong for another. Certainly, performance figures are useful in making a choice, but in a detector, ease of use, comfort, reliability etc are equally important. Who will find most gold? The person with a detector that has great range but is a dog to use, and whose arm is at breaking point after an hour or two. Or, the person with 20% less range, but has a machine that is comfortable enough to swing all day. Many detectorists have told me that a positive frame of mind helps enormously in the number of finds made. Again, a comfortable detector encourages this frame of mind. Experience with a detector also enhances results. I have been out with a chap, using an identical detector, who can find a hundred coins to my ten on the same patch of beach. Why? Because he has become so atuned to his detector that he can hear those little whispers that are inaudible to me. I consider that my hearing is good, as Hi-Fi is my second interest after detecting, but it has become obvious that in the hands of a competent and experienced user, the detector becomes an extension of his senses; part of his being, almost. This in itself is equivalent to a few more inches.
Personally, I don
 
Eric,
Thank you very much for your insightful and philisophical response. There are indeed many factors that affect how much gold an operator finds and how their detector performs.
My original post to the Finder Forum was inspired by my reading of frequent posts in that forum that compared the depth performance of each detector type (SD series and GP Extreme). Some posts were scathing of the manufacturer's performance claims, others supportive. What I was trying to offer was a standardised method of testing that removed as many variables as possible, in the hope of seeing the presentation of more consistent results. My proposed test does overlook the influence of the "machine familiarity factor" which, as you rightly point out, can be large. I guess I overlooked it because in the test I conducted the performance difference was immediately apparent, even though I hadn't previously used the GP Extreme.
It looks like the debate will continue to rage in Finder's Forum. Thanks for your input.
Lachlan
 
Hi Lachlan,
Using coins as test objects does remove the variable of the different sizes, shapes and compositions of nuggets. They are also available countrywide for tests and comparisons. I would use a 5c coin to simulate low conductivity items, a 20c coin for medium and 2c copper coin for higher. Detectors can vary in their performance over this range i.e. one that gives best detection on a 5c may not be as good on a 2c. I don't know if you have any larger copper coins, or silver coins in Australia? Here in England we have to be careful over using the more recent "copper" coins. They are in fact copper plated steel, and do not give the correct response.
Eric.
 
Hi Eric,
Thanks for the suggestions. Only our pre-decimal coinage (pre 1966) contained large copper and silver coins; (the penny (about the diameter of our current 20 cent piece, and the "two-bob", respectively). Although a silver 50 cent coin did continue for a couple of years after 1966. I don't think we have any copper-coated steel coins in circulation.
Cheers
Lachlan
 
Top