Ground balance add on board for Mini pulse .

[godigit1]

Thank you all for the answers, It appears I still have allot of reading to do.
Im really just barley starting to get grasp on how things work on a pi.
Im a carpenter by trade but have a distant backgroung in Pwbs.
This started out as a simple working GB solution started by Digger427. But when I started digging through the forum It got much more interesting.
So Let me Say Im sorry for my lack of true electrical experiance.
Ive built a HH surf Pi Cuda And 4 MPPs but have just been going through the motions tell now. I want to say thank you to all of you on the forum for helping to open my eyes lateley.
That doesnt mean I wont still ask stupid questions though so please bear with me and let me know when Im wrong. No one gets any better without constructive criticism.
That being said .

Teleno,
Ive been trying to wrap my head around your current feedback op amp design. Could you expand on it a little for me?
Im not positive how it works but Am I correct in saying this is a one sample amp and ef is cancled by the auto 0 function of the tl082.
So could then a second sample for GB be added right at C input or is the 0 function actual GB?
If so then the Mpp stock Sample pulse generator could be capable of providing all the samples I need without a second sample generator?

Like I said I have allot of reading to do so if this has been discussed in a thread I havent found yet please let me know.

Thanks again

[Davor]

EF sample turns every PI Rx into an auto-zero amplifier.

[Teleno]

EF sample turns every PI Rx into an auto-zero amplifier.

That's right. Auto-zeroing the amplifier is simpler than a differential integrator and less tolerance sensitive.

[Digger429]

Having been away from the forums for a few months, I come back to some questions being asked. Hopefully I can clarify here.

The on/off switch for the GB is indeed a two pole switch that just switches to ground. That is, the grey and white wires going to the FETs in this picture here...

...are switched to ground.

The two potentiometers (blue circle and pink circle) that geoscash1 asked about in post 6 no longer exist. Neither does the potentiometer at top left, which is now fixed timing by a resistor.

Here is a diagram of the timings I've setup, so I can run the machine at about 1000 pulses per second:


After two trips into the goldfields of Victoria, I've decided its about as good as its going to get. Anything from here on would be a complete redesign from the ground up. It is now able to completely cancel out the ground, but it comes at a great cost of reduced sensitivity and the "hole" where gold samples can completely vanish from detection. In practise I found myself having to turn the ground balance amount down to as low as possible without the ground interference becoming too annoying. Aside from the on/off switch, thats the only GB control - the potentiometer "GB sample width" that can provide 8 - 245 uS of sample.

What other questions were there?

[6666]

Digger I get invalid link, on both links.

[Davor]

That's right. Auto-zeroing the amplifier is simpler than a differential integrator and less tolerance sensitive.

Well, the servo you put to form an auto-zero is also an integrator. But running it separately from a main integrator makes duration of EF largely independent from a main sample duration. It should be possible to use more of the main sample energy for the purpose of ... detecting. Makes sense

[Digger429]

Second attempt at images. This time I'll just link externally to imgur.

The on/off switch for the GB is indeed a two pole switch that just switches to ground. That is, the grey and white wires going to the FETs in this picture here...

...are switched to ground.


Here is a diagram of the timings I've setup, so I can run the machine at about 1000 pulses per second:

[Davor]

A thing came to my mind... the timing may of course be produced by a PIC. But PIC may also double as a through zero VCO to fix the dual polarity issue. It may as well produce an effect in VCO sound to indicate positive vs. negative response, thus adding a sort of discrimination to it. The existing 555 may still act as an amplifier. A bit radical, but doable.

[godigit1]

Hi Digger,

Thanks for looking in on this thread. I didnt know if you were still around the forum or not but glad to see you here.
Yes my first question is Find any GOLD???
Second is have you tried different coil types like self shielding versus shielded coils? Does it make any difference in performance?
Thanks for sharing your circut Im just about set back up to finish it. I was going to play with a couple different balance methods if I can but for this add on board Im going to do it your way and just play with the timings so its a plug in add on for the stock MPP.
I may have more quetions along the way but you covered the one I had for now. I was not switching off at the fets instead at ground and -5.
Thanks again.

[waltr]

A thing came to my mind... the timing may of course be produced by a PIC. But PIC may also double as a through zero VCO to fix the dual polarity issue. It may as well produce an effect in VCO sound to indicate positive vs. negative response, thus adding a sort of discrimination to it. The existing 555 may still act as an amplifier. A bit radical, but doable.

yes, that is what I did with a PIC (could be done with an arduino as well) on my HH2.
The audio out does high verse low conductor discrimination. Ferrous however can go either way so no Fe discrimination.

[godigit1]

Hi Teleno,
Thanks for sharing your circut.
Of the two op amp configurations you have designed for the MPP which one would be faster?
Also interested in the DEboo configuration you have mentioned. Like I have said I have allot of studying to do.
Im going to order the 6181 and is the 2338 critical or will any NPN do ?

Also should I build the circut above or the one in the spice sim below?

Thanks again

[Teleno]

Of the two op amp configurations you have designed for the MPP which one would be faster?

This one is faster because the last stage is a current-feedback op amp.

Im going to order the 6181 and is the 2338 critical or will any NPN do ?

The 2N4338 is an N-channel JFET switch. I used it because it's available in LTspice.
The MPP uses J113, you can use it here too.

JFET switches with a lower VGS(off) have lower on-resistance but you need around -10V in order to shut them off properly. On-resistance is not critical in this design, so the optimal type depends on the minimum voltage available to be applied to the gate during tun-off. In the MPP is -12V, so you can go hapily with the lowest on-resistance types such as the J111. If you only have say -5V available then go for the J113.

The characteristics of these types are on page 3 of this data sheet https://www.mouser.com/ds/2/149/J113-889353.pdf

Also should I build the circut above or the one in the spice sim below?

The one in the spice sim below is better, but before using it on the MPP you have to configure U2 as NON INVERTING (it's inverting in the sim), otherwise the output of the preamp will be on the wrong side of GND. It has to be negative going for the MPP.

If you're detecting for nuggets then use the lowest coil current you can afford. it will reduce depth for general targets but since you'll be able to sample earlier, it may actually increase sensitivity to smaller gold. Or just use a DD coil and get the best of both worlds.

[SaltyDog]

Hi All,

I am new to this forum, and have just purchased the MPP Beach (RevE) kit, an IP68 instrument case and water proof pots/switches etc .. my intention is to have the complete detector fully waterproof ..

I have a question regarding ground balancing techniques, as I would like to add that too:
Why can you not just increase the sample time for the Earth field effect, so that it takes care of ground as well? Why the extra two samples ... (just trying to do the minimum possible, as long as it works )

Cheers

[6666]

bump

[Qiaozhi]

Hi All,

I am new to this forum, and have just purchased the MPP Beach (RevE) kit, an IP68 instrument case and water proof pots/switches etc .. my intention is to have the complete detector fully waterproof ..

I have a question regarding ground balancing techniques, as I would like to add that too:
Why can you not just increase the sample time for the Earth field effect, so that it takes care of ground as well? Why the extra two samples ... (just trying to do the minimum possible, as long as it works )

Cheers

Try this ->
https://www.geotech1.com/forums/show...769#post201769

or you could lose your mind by reading this ->
https://www.geotech1.com/forums/show...718#post208718
and the 5 posts after it.