Menu
What's new
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

GQ ground elimination

Hi Eric,
In my early post, I was trying to say that you can't use anything that equalises the current in both pulse lenghts or it won't work. You might find fudging the properties of each pulse benificial after getting the initial result close though.
I took some pics of the pre-amp output last night only to find that I can't post them here. I will send them to you and you can post them if you wish. The effect is dramatic as you will see. I had to use minerals that aren't quite universally cancelled by a ML as magnetite isn't as dramatic. If I have misunderstood you or got something wrong then let me know.
Your second proposal is novel and I expect that the FRDs will be crunched up tightly together in the short pulse and will vary in the long one and you can chase them with a change in voltage to the long pulse. I hope I'm wrong but I can picture the FRDs ramping back and forth in the long pulse while remaining stationary in the short pulse? All I know is that matching the effects in two pulses is not easy.
Like I said, I might be wrong as I have been consistantly wrong since turning sixty and if I was honest I would make that twenty one or much earlier.
<img src="/metal/html/smile.gif" border=0 width=15 height=15 alt=":)">
Rob.
 
After reading my own message again, I have noticed that I presented the matter a bit in a simplistic way. I would like to answer your questions one by one.
Q1: Yes, there is definitely a consistent difference in response. Except for some difficult cases, for the same metal of target, the same curve shape is observed within reasonable distance range.
Q2: Given that the most interesting part of the decay is located in the first 50
 
Hi Rob,
Thanks for the email and pictures. I find that I can duplicate your waveforms with my setup and box of rocks when testing on the Minelab 11in coil. Also I varied the pulse width from 70uS to 280uS as you did. However, with the low resistance Minelab coil, there are two things changing at the same time. Firstly, we have the change in pulse width, and secondly a change in pulse amplitude. When running at 70uS, the current just before switch off is 2A (I set the supply voltage to give this convenient starting figure). When you wind the pulse width out to 280uS, the corresponding peak current rises to 6A, due to working further along the coil time constant. This threefold increase in current will result in 3 x the signal at all points on the FRD. This makes it appear to shift to the right on the scope. If you now apply the constant current circuit, you then take out the variation in signal due to the change in current amplitude and you are left with the change due to pulse width, which in my testing, is very little.
The SD2200, which I had a while back, had three short pulses and a long one as does the GP Extreme, which I am now looking at. In the SD, the short pulses had a similar 1 : 3 ratio in peak current, which you would expect. Assuming that the long pulse is there to generate the ground signal for subtraction, this would give a greater amplitude to play with on a late sample. On the other hand, the GP Extreme applies a high voltage to the short pulses so that the peak currents of both the long and short, end up the same. The difference in voltage driving the two pulses, no doubt results in the
 
Hi Eric,
Is it possible to share on the forum the interesting pictures sent to you by Robby?
Willy
 
First is the preamp output with no signal. Second is the signal from iron mineralised rocks with a 70uS TX pulse, and third, the same but with a 280uS TX pulse. You can see that the signal stretches out to the right with the longer pulse. I think the horizontal scale is 5uS per div. Vertical scale 2V/div.
Couldn't figure out how to upload the three images on one post <img src="/metal/html/frown.gif" border=0 width=15 height=15 alt=":(">
Eric.
 
Hi Eric,
I gave up a while back trying to emulate what happens in nature re the exitation and decay of ferrites in a non-saturating linear field and how this can be even compared with the abrupt field from a coil that is like an over-inflated wheel barrow tyre. The claims made seem to apply to very long duration pulses that are useless with a moving coil so you are forced to get the results any way you can. Using the circuit that gave the pics I sent results in no item having a constant in either pulse so constant fudging is needed and the point was that any change in the DC resistance can have a marked effect on the results. I see the high voltage short pulse in the GP as fudging some properties in an attempt to bring universal cancelling closer to the ideal.
A patent by another guy here claims to have obtained the ideal and that is, minimul or no intervention is needed by auto or manual GB to obtain full universal cancellation of all offending minerals.
There are a stack of effects that you can find when just playing with pi.
I have noticed one that appears to rely on a few things that should have no effect along with a bit of novel fiddling around the damping circuit and isn't tied to pulse length. I can't explain this effect at all. It allows the decay from a long pulse to seemingly reflect the pulse history or the spikes history or anywhere in between making it a thin skin surface area detector or a mass, eddy current detector. It is only of value if it can be switched in and out on the fly and I've never been able to do this as I can't figure out how. I keep running into bias problems caused by the back emf amongst other things.
I keep thinking about your proposal and it all comes down to your findings in the end but I see the FRDs that differ from that of your base sample as having a milder effect in your short pulse than in your longer GB pulse, possibly driving it crazy as the ground varys but if this is right then you are in a good position to correct this by providing adjustments to the series resistance in the short one to get the ratios right, I think???
Tis all mind boggling this pi stuff as you can see a million effects and wonder how they can be made to work for you.
Rob.
 
Eric.
I can see that I have kept up the tradition of being consistently wrong. I suppose my taking claims literally is somewhat dangerous.
So here is another attempt at explaining the claims.
My measurement jig is in pieces at the moment so I can't say for sure whether the below is right or not but from some crude measurements made last night it seems a nugget gives proportionately the same response in different pulse lengths (not the same response) when using the front end design and coil that ML uses. You would be in a better position than I am to test this at the moment.
When a small nugget is brought near the coil there is an amplified response from a sample taken in the long pulse compared with a short-pulse sample when using the same nugget as a target so a gain needs to be applied to the short pulse sample to give the same response for a (small) nugget in each pulse length and a large nugget could then be tested to see if the response was also equal in both.
If so then the FRDs should hopefully give a different response in each pulse length and the claimed relationships may then come somewhat closer to becoming a reality.
If the results were found to fit the claims then I suppose it would be somehow related to the rapid build up of current at switch off keeping eddy currents at a fixed ratio in both pulse lengths but at the same time resulting in ferrites being affected differently.
The reason for the increase in voltage supplied to the short pulse in the GP then becomes a bit more evident if the above is true as the pulse length relationship is still maintained but the applied gain could then be lowered or perhaps even made equal. Maybe accounts somewhat for the disappearance of the long off period that was there in previous models to subtract the EFE??
Rob.
 
I just had another look at patent US 6,586,938 B1. Paltoglou makes much the same statement as Candy and states,
 
How is your PI project coming along?
What part number Shottky Diodes are you using? Do they work better than the traditional 1N41458 diodes?
How soon before you have boards and plans available?
This would make a good forum project if the parts are easily available.
bbsailor
 
Q1:We have now a working prototype #1 giving us NO-MOTION sensitive ALL-METAL search mode and a specific Discrimination mode with EEPROM table-driven target signature interpretation. We also have an automatic ground control.
We prepare the re-packaging of the modules to make a separate, intelligent human-interface PCB (16F876A, digital VCO-like Audio generator, 2x16 LCD and joystick) on top of the already existing analog front-end and the post-processing modules. We shall then build our prototype #2 in 3 copies by three (quite) separate persons of our team to test the construction manual and the quality of the PCB's.
Q2:Shottky 1N5819.
Q3:probably early this summer, we shall have tested our proto#2 with the new PCB packaging.
Q4:It has always been our intention to openly discuss about this project on a few serious technical forums but only after we have built more confidence by more field tests. Currently, all the tests have been done in lab.
We have been very careful to use only standard (but up-to-date) components generally available and economic (the most specific ones like the fast ADC, fast OpAmp and PIC can even be got free of charge from their suppliers like analog devices, microchip and maxim-ic). The analog parts are as few as possible and kept on a separate PCB. All the signal generation, timing, filtering, integration and other processing are currently executed by one PIC16F877A programmed in CCS C.
In the proto#2, all the human-interface will be remotely executed by a separate PIC16F876A and the main PIC will also be replaced by a (cheaper) PIC16F876A. We shall then have much more freedom to keep the main electronics and power supplies on a big blind box located under the elbow and a small box in front of the operator containing all the human-interfaces (lCD,Audio,Joystick) and connected through I2C to the main board.
Willy
 
You must log in or register to reply here.
Top