Silverdog PI 1.2 kit Ground balancing

[Reg]

The problem is two samples would never work, to remove mineralization. This is because the second sample sees less response than the first sample, to compensate for this the second sample would need to be amplified such that it outputs the same level as the primary sample then they get subtracted leaving only target response.

The next problem is if you amplify the second sample then Earth Field cancel would no longer work because it's normally a constant level during the entire receive period.

This is why ground balancing detectors use 3-4 or more samples.


PS I suspect Kt315 is getting mixed up between Earth Field and mineralization.

Mickstv,

When I wrote the Understanding the PI article back in 2003, I tried to keep the article simple and keep it short so I didn't discuss the fact secondary samples to eliminate the field effect were needed. Also, back in 2003 few people knew how ground balancing was done so adding a lot of secondary information would really confuse the issue.

Reg

[Reg]

Goldscan IV does 4 samples. does anyone know about it can be good ground balanced at your heavily mineralized areas?

The Goldscan IV (4) works fine in really bad ground and does a very decent job of ground balancing out the ground signal. In fact the Goldscan series was built to be used in OZ and used for gold hunting. The weakness and probably the reason it never caught on was the fact, the design was totally new and little to nothing was written about all the controls and what they did. Worse yet, SEL1 and SEL2 modes would cause a person to totally ignore large gold because of the intentional design of these two modes to eliminate signals from iron while detecting certain English coins. This ability to ignore iron was great, but it also ignored all high conductor signals including large gold nuggets. Had that feature been fully explained as to what would happen, I strongly suspect the detector would have become much more popular.

Since there was no noted discussion this ignoring some objects would happen, people felt the detector was failing when it ignored a large gold nugget and, in my opinion was one of the big reasons the detector never caught on.

I purchased a GS 4 years ago and it took some time for me to better understand just what was going on. In fact, when the detector ignored a 1/4 oz nugget using the SEL1 mode while trying to ignore a nail, I was shocked.

It wasn't until I was experimenting with the GS 5, did some of the GS 4 features begin to make sense.

It is too bad someone hasn't simplified the design a little and created the pc board for sale for the GS 4. It would make a great project for a ground balancing PI.

Reg

[Reg]

Bill512,

You can use the second sample to create a form of ground balance. You will have to run the general sensitivity much lower to minimize the earth field effects, though.

Now, to use the second sample you will have to change the timing such that the second sample begins about 10 usec after the main sample ends. Next, you will have to make the gain of this sample or the main sample to be variable because you need to amplify the GB sample signal a lot with respect to the main sample. This isn't exactly easy to do and keep all things in balance but it might work well enough for for simple testing to get an idea of how it might work.

Again, it is too bad the GS 4 isn't built in a kit form.

Reg

[Bill512]

Bill512,

You can use the second sample to create a form of ground balance. You will have to run the general sensitivity much lower to minimize the earth field effects, though.

Now, to use the second sample you will have to change the timing such that the second sample begins about 10 usec after the main sample ends. Next, you will have to make the gain of this sample or the main sample to be variable because you need to amplify the GB sample signal a lot with respect to the main sample. This isn't exactly easy to do and keep all things in balance but it might work well enough for for simple testing to get an idea of how it might work.

Again, it is too bad the GS 4 isn't built in a kit form.

Reg

Hi Reg, I've done this effort a year ago.
As I remember, it had variable position, width and gain on the second sample.
Not really a super detector but a good start for ground balance learning.
I agree 100% also, about GoldScan. It is time to start a new era in DIY PI detectors...

[Wessex warrior]

Hi there Bill 512,

I am impressed and interested in your ground balancing on the Surf PI. How did you do it exactly? Did you use another op-amp to do the subtraction/integration? Maybe another 4066 as well.

Thanks

[Orbit]

combination with the integrator and gate 4066 and you have ground balance !

[Bill512]

Hi there Bill 512,

I am impressed and interested in your ground balancing on the Surf PI. How did you do it exactly? Did you use another op-amp to do the subtraction/integration? Maybe another 4066 as well.

Thanks

Hi,
this project was just to prove a concept : ground balance using only the second sample.
As I wrote before, you need to change the second sample in terms of position, sampling width and gain.
It worked reasonably well for the simplicity, but not good as a full working detector.
After all, it was for learning and experimentation on GB.
The next step was to use an extra 4093 and 4066 ,plus an extra differential integrator, etc.
Also major changes was needed on the sound part in order to get responce from both of the sides of the ground ballance hole.
Then, I realized that all of these mods, will be better to apply directly on a new detector ,rather to modify further the SurfPi.

[6666]

you need to change the second sample in terms of position, sampling width and gain.

How much variance in the gain were you experimenting with ?
thanks

[Wessex warrior]

Thanks for the reply Bill512,

So using just the two samples for ground balance, was there any reduction in detection depth?

Thanks again.

[Bill512]

Thanks for the reply Bill512,

So using just the two samples for ground balance, was there any reduction in detection depth?

Thanks again.

yes, depending from the type of the ground. Roughly from 20 to 40%.
As far as I remember, it was capable of ground balancing on mild red clay, but not enough for really hot red clay (terra rossa).
The photo above, is from a field with very "easy" ground balancing soil.

[Bill512]

How much variance in the gain were you experimenting with ?
thanks

If I remember correctly, from normal gain, up to "all that you can take" from the integrator- (200K+500K potentiometer).
But while adding 1M pot, there wasn't any significant additional gain, so I was forced to reduce the 1K resistor.
I must say that didn't take detailed notes and after a few trials I took off the GB modes.

[6666]

Thanks Bill512.

[6666]

Well it can be done using just 2 samples..... But you need to use 2 different tx lengths.

Now the easy part!!!! How to make it happen. There is discrete logic, use a micro controller of some sort to generate the tx and sample timings, or use a pre-programmed eeprom

In reality though, if you wanted to use a detector for searching for gold you would be better off starting with a better platform than the surf pi.

Cheers Mick

Mechanic are you still on the forum
could you give an indication as to the 2 tx pulse lengths thanks.