TINKERERS GROUND BALANCE

[Tinkerer]

A good ground balance system is very important.
For Version 1 we want to keep everything as simple as possible.
One ground sample.
Single tracking speed, for a sweep speed of about 1meter second max.
Static ground balance for pinpointing.

For V2 and V3 we will improve things a bit.

Anybody knows of Ground Balance methods? Please help here.

Tinkerer

[Tinkerer]

TINKERERS GROUND BALANCE

Attached is an idea for ground balancing. It works on the breadboard, but has not been tested in the field yet.
The GROUND BALANCE is auto tracking, but when a target is located, it automatically switches to pinpoint mode.
A manual override will be added later.
There is no loss of sensitivity in the tracking mode.

Although the idea works, I am sure that there are improvements that can be made in the implementation of it.

I hope that people better versed in electronics than I will offer improvements, before I start on the PCB.

Tinkerer

[Mechanic]

Hi Tinkerer,

Any chance of posting a better quality pic or a spice sim? It's a bit(extremely)difficult to read.

Cheers

[Tinkerer]

Hi Tinkerer,

Any chance of posting a better quality pic or a spice sim? It's a bit(extremely)difficult to read.

Cheers

Hi Mechanic,

At TINKERERES PREAMPLIFIER, there is the .pdf version.
http://www.geotech1.com/forums/showp...7&postcount=25,

I am still working on it and have made some minor changes. There is a problem at switch ON, the automatic change from tracking mode to pinpoint mode latches up.
This could be fixed on the software level, with a soft start sequence.

I would appreciate any help.

Tinkerer

[Kev]

Hi Tinkerer,
By using an inverted input signal are you not subtracting too much target response along with the ground signal? Would it not be more accurate to use a combination of the Fe sample combined with a late time decay curve sample?

Cheers
Kev.

[Tinkerer]

Hi Tinkerer,
By using an inverted input signal are you not subtracting too much target response along with the ground signal? Would it not be more accurate to use a combination of the Fe sample combined with a late time decay curve sample?

Cheers
Kev.

Hi Kev,

I seem to have been missing all your posts.

In this circuit, 3 different samples are taken, then they are scaled and summed. This, then is the sum of all signals produced by the ground, the coil itself and whatever other signals that make up the NO TARGET signal or basic CONSTANT.

The WITH TARGET signal is composed of all of the above plus the target, therefore if we subtract the NO TARGET signal, which we can call THE VARIABLE, we are left with only the target.

The spots in time, where the 3 different signals are taken, are chosen for the specific information they can give.
One spot gives best information about magnetic targets.
Another spot gives best response of the smallest targets.
The third spot gives a good overall colored metals response.
By scaling the responses, certain features can be attenuated and other features can be enhanced.
The sum of the scaled responses can therefore enhance the desired information and attenuate the unwanted responses.

All the best

Tinkerer
Tinkerer

[Kev]

Hi Tinkerer,
I had in mind a sample timing diagram you posted elsewhere, but did not appreciate the number of iterations and succeeding improvements that were implemented over the development, and now I see that with the uC sample combinations, variable and constant updates are relative to present feedback. So what I first thought was hardwired is clearly not so.

Thankyou for your clarification.

Kind regards
Kev.