Ground balance

[Carl-NC]

Given that EFE and ground/target signals are independently driven (assuming no major interplay here), from a purists perspective, I would have therefore thought that it's more straight forward to simply have a circuit that permanently makes A1(S3) - A3(S4) = 0 (with A3 being a new independent gain setting for the GB EFE sample), thereby avoiding any change to the ideal A1(S1)- A2(S2) (i.e. perfect world) equation. Otherwise, when you change A2 in A2(S2 - S4), the gain of the S4 sample will also be changed, and adverse consequences may arise.

EFE cancellation is achieved when S1-S3 = 0 (Target channel) and S2-S4 = 0 (Ground channel). Then when you apply whatever gain to each channel and subtract, EFE cancellation is maintained. That is to say, each channel is independently "EFE clean" before you ever apply gain or combine them. This is done at the integrators by simply making sure the EFE sample has the same pulse width as the main sample (S1=S3 and S2=S4).

What you are suggesting will work but it's the hard way to do it. You will need a separate integrator stage with a separate gain stage, and then you need to do a GB adjustment followed by a separate EFE adjustment. The way I suggest (which is the way all commercial non-bipolar ground balancing PI detectors do it) is simpler and automatic... EFE cancellation is always maintained throughout the full range of GB.

Separately, if one was to adopt the changing GB target sample width approach to ground balancing, wouldn't this enlarge the target "hole"?

Sorry, didn't understand this question.

[green]

No thoughts why the 2C integrator cancels only at certain target and EF delay times? Tried disabling target sample switch or GB sample switch with 1C and 2C integrator https://www.geotech1.com/forums/atta...8&d=1599836935 1C integrator same signal opposite polarity. 2C integrator: similar to 1C when +input switch disabled, about 1/10 the signal when -input switch disabled. Maybe could be the cause?

Another try. I read for a 2C integrator R3 and R11 should be matched, same for C2 and C5. My simulation cancels at 486us GB delay. Signal is reduced at other delay settings. Thinking I must be doing something wrong if matching is required. Another member is beating his head with a 2x4 because he missed something simple. Someone was kind enough to point out his error. I would like to beat my head with a 2x4 and get it over with. Maybe someone could point out my error.

Maybe someone could do a 2C integrator spice simulation that does cancel at different delay settings.

Tried test with R11=1.05meg(5% difference)mV readings changed but weren't any higher than 30mV when R11 was 1meg.

[Carl-NC]

No thoughts why the 2C integrator cancels only at certain target and EF delay times?

I looked at the sim setup and didn't notice anything wrong. Yes, a 2C integrator will cancel EFE, it is commonly found in Eric Foster's designs.

[waltr]

green,
Tried the 2C integrator sim you showed. Get similar results.
Then swapped the delay on SW1 with Sw2 and the offset flipped polarity.

What I think is going on is the second Sample is on a different part of the 1Hz Target 'sine' so is at a different Voltage.
If both sampling switches pulse at the same time then output is zero since there is NO Voltage difference between Sample1 and sample2.
I don't think using a Sine wave input is a valid 'ground' signal although is seems to make sense. Or does it???

A 'real' target has an exponential decay and is a straight line when plotted log/log.
'Ground' is Not exponential, it plots as a straight line when plotted Lin/log.

Can not explain how to make it work but this is my thoughts as to why this simulation is not working as expected.

Playing more with Sim.
Measure Current into the op-amp inputs and find, as expected, they are NOT this same.
In real GB circuits, there is a pot to adjust Gain of GB. This would compensate fro op-amp input current differences.

Also try setting Target generator, V3, to 0V DC. There will be offset at output due to op-amps.

Sorry this wasn't a 2x4 but hope it helps some.

[green]

green,
Tried the 2C integrator sim you showed. Get similar results.
Then swapped the delay on SW1 with Sw2 and the offset flipped polarity.

What I think is going on is the second Sample is on a different part of the 1Hz Target 'sine' so is at a different Voltage.
If both sampling switches pulse at the same time then output is zero since there is NO Voltage difference between Sample1 and sample2.
I don't think using a Sine wave input is a valid 'ground' signal although is seems to make sense. Or does it???

A 'real' target has an exponential decay and is a straight line when plotted log/log.
'Ground' is Not exponential, it plots as a straight line when plotted Lin/log.



Can not explain how to make it work but this is my thoughts as to why this simulation is not working as expected.

Playing more with Sim.
Measure Current into the op-amp inputs and find, as expected, they are NOT this same.
In real GB circuits, there is a pot to adjust Gain of GB. This would compensate fro op-amp input current differences.

Also try setting Target generator, V3, to 0V DA 'real' target has an exponential decay and is a straight line when plotted log/log.
'Ground' is Not exponential, it plots as a straight line when plotted Lin/log.
C. There will be offset at output due to op-amps.

Sorry this wasn't a 2x4 but hope it helps some.

I don't think using a Sine wave input is a valid 'ground' signal although is seems to make sense. Or does it???
The sine wave is for EFE, sweeping the target with the coil

A 'real' target has an exponential decay and is a straight line when plotted log/log.
'Ground' is Not exponential, it plots as a straight line when plotted Lin/log.
Think you have them reversed, should be Lin/log then log/log.

[waltr]

The sine wave is for EFE, sweeping the target with the coil

Then I am not clear on what you are trying to do in that Spice sim.

A 'real' target has an exponential decay and is a straight line when plotted log/log.
'Ground' is Not exponential, it plots as a straight line when plotted Lin/log.
Think you have them reversed, should be Lin/log then log/log.

Correct, I have them reversed.

Did you try setting the Target signal source to 0VDC??

[Carl-NC]

What I think is going on is the second Sample is on a different part of the 1Hz Target 'sine' so is at a different Voltage.
If both sampling switches pulse at the same time then output is zero since there is NO Voltage difference between Sample1 and sample2.
I don't think using a Sine wave input is a valid 'ground' signal although is seems to make sense. Or does it???

He's simulating a slow (1Hz) signal to emulate EFE. This is valid.

[Elliot]

EFE cancellation is achieved when S1-S3 = 0 (Target channel) and S2-S4 = 0 (Ground channel). Then when you apply whatever gain to each channel and subtract, EFE cancellation is maintained. That is to say, each channel is independently "EFE clean" before you ever apply gain or combine them. This is done at the integrators by simply making sure the EFE sample has the same pulse width as the main sample (S1=S3 and S2=S4).

What you are suggesting will work but it's the hard way to do it. You will need a separate integrator stage with a separate gain stage, and then you need to do a GB adjustment followed by a separate EFE adjustment. The way I suggest (which is the way all commercial non-bipolar ground balancing PI detectors do it) is simpler and automatic... EFE cancellation is always maintained throughout the full range of GB.



Sorry, didn't understand this question.

Thanks Carl. I think I've got it now. I appreciate the patience and time you've taken to explain it. What sort of maximum relative gain might be required for the Ground channel in highly mineralised fields compared with the Target channel (in other words, should VR1 (refer attached) be say 4k7 or 10k)? Also, some say that you a should insert a trimmer (VR2) between the switches S4 and S2 to replace the link. But if your explanation is correct, I don't see the need for it. Am I missing something here? With respect to the "hole width" question at #39, Green has answered that for me at #44. My question was basically, if you adopt a GB solution that changes the width of the GB sample (as opposed to changing the gain (A2) approach), will the TC hole widen? It appears the answer is NO.

[Qiaozhi]

There is a mystery here to be solved.

Have a look at the attached image. Green's simulation is now working!!!
See if you can spot the difference.

Ignore the fact that I swapped out the universal opamp model for an LT1007, and changed the pulse rise and fall times to 10ns as that's not the fix.
Also, I set the input resistors to the switches to 10k. That's not it either.

[ZED]

The gain of A3 would be independently adjusted to fully cancel out EFE from the equation, regardless of the GB gain setting (A2).
I'm thinking if first delay is changed, GB(A2)needs to be readjusted. Think if A2 is changed A3 would also need to be adjusted. Maybe I'm wrong.

Your thread started me wondering again if changing gain or sample time to GB is better. Thinking it might depend on the target. Been using a 1C integrator and adjusting sample time to GB. Tried a 2C integrator awhile back, didn't work as good. Tried 2C circuit in spice and can't see why it wouldn't work. Thinking of making another integrator. Make delay times #1 and #2 adjustable. Make target sample time adjustable. Make GB sample time and gain adjustable. Test circuit to see if I can tell if adjusting GB sample time or gain is better. Any suggestions appreciated.

G'day green

I like to GB by changing the phase of a sample,changing the phase of the target sample (sample delay) works but it makes the detector sound off so its usually the second or third sample that i adjust,i dont like adjusting the gain cause it affects the earth field
balance,i aim for a very tight EF balance and i dont want anything upsetting it,the ground here with its remnant magnetism has me getting very fanatical about EF balance.Also tried and still experiment with adjusting the width of samples to GB,now this does work but curiously when the EF is balanced and the sample width is adjusted i can notice the samples going out of EF balance despite the sample widths remaining the same ,even down to an accuracy of 25 nano sec.

I will explain that a bit better,i have a very late sample (before the next TX period) that is my compensating sample (EF sample),so when i adjust the width of the GB sample it also adjusts the width of the late compensating sample,the resolution is
25 nano sec so its very accurate, but as the GB sample is increased or decreased and the compensating sample is adjusted accordingly it still throws out the EF adjustment,admittedly this has me scratching my head with puzzlement.

[green]

I was sampling 1000 times/second for 10 seconds. Looks like you are sampling once in 10 seconds? reply #54

[Qiaozhi]

I was sampling 1000 times/second for 10 seconds. Looks like you are sampling once in 10 seconds? reply #54

Oh bugger!
I thought it was an issue between using the "signal" symbol in the misc library and the independent voltage source, even though the netlist looked identical.
Unfortunately I then introduced a typo in the source parameters.
Back to the drawing board for another look.

[Qiaozhi]

By the way, you don't need to explicitly put in the number of cycles. You only need to do that if you want to set a limit, otherwise the cycle just keeps repeating until the end of the simulation.

[ZED]

The hole as mentioned by some members (a point where the target TC is the same as the ground balancing null point) is not related to the TC of the ground signal,but related to the interplay between how the samples are adapted and arranged.I made a number of targets of different time constants,and each time constant progressively getting longer,this allows me to detect target holes and to map them as i use different sampling coefficients,i noticed as other members have that the target hole remains relatively constant as the ground balance is adjusted, but when a completely different set of sampling coefficients are used,that is to say dramatically different,then the target hole shifts to a different position altogether.

[waltr]

I was sampling 1000 times/second for 10 seconds. Looks like you are sampling once in 10 seconds? reply #54

Saw that also.

Oh bugger!
I thought it was an issue between using the "signal" symbol in the misc library and the independent voltage source, even though the netlist looked identical.
Unfortunately I then introduced a typo in the source parameters.
Back to the drawing board for another look.

It happens.
Still think my answer points to part of the issue, Op-amp input bias currents and Voffsets.

By the way, you don't need to explicitly put in the number of cycles. You only need to do that if you want to set a limit, otherwise the cycle just keeps repeating until the end of the simulation.

Saw that and learned something new. Thanks.