Ground balance

[Carl-NC]

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.
[ATTACH]53196[/ATTACH]

The GB gain required depends on the spacing between the Target and Ground pulses. The more spacing, the more gain. This will also alter the location of the target hole. For, say, 10us spacing VR1 should never need to be as high as 10k, probably not 5k, maybe 2k. Also, if you run the Ground pulses wider than the Target pulses then that will reduce the Ground channel gain requirement.

The trimmer is not needed if you match the input resistors to, say, 1% or better. Also the resistors & caps in the integrator need to be well-matched.

[green]

The trimmer is not needed if you match the input resistors to, say, 1% or better. Also the resistors & caps in the integrator need to be well-matched.

With spice EF simulation. 1C integrator_ 5% mismatch of input resistors causes output to be about 5% of the input. Zero output if input resistors are matched. 2C integrator_ cancels at one second delay time only. 5% mismatch moves the second delay time that cancels. Signal out less than 5% of input at other second delay times?

Could someone explain EF. Connected a 8inch mono coil to Rx input to see what we are trying to cancel. Don't see a change at amplifier out when moving coil back and forth. Rocking coil causes a signal. What amplitude should I expect for EF signal at coil out(amplifier out/amplifier gain). Should I expect to see EF if coil remains flat while sweeping?

[Carl-NC]

With spice EF simulation. 1C integrator_ 5% mismatch of input resistors causes output to be about 5% of the input. Zero output if input resistors are matched. 2C integrator_ cancels at one second delay time only. 5% mismatch moves the second delay time that cancels. Signal out less than 5% of input at other second delay times?

Don't understand why, sorry.

Could someone explain EF. Connected a 8inch mono coil to Rx input to see what we are trying to cancel. Don't see a change at amplifier out when moving coil back and forth. Rocking coil causes a signal. What amplitude should I expect for EF signal at coil out(amplifier out/amplifier gain). Should I expect to see EF if coil remains flat while sweeping?

Any dB/dt cutting through the coil will produce a signal. If the Earth field is perfectly uniform and you swing perfectly level then there should be no EFE effect. Rocking/twisting the coil will maximize the effect. Amplitude depends on coil turns & area, gain of preamp, etc. A good bench test method is to use a magnet, but it must have no eddy (metal) response. Many ceramic magnets will work.

[Qiaozhi]

As far as I can ascertain there is nothing wrong with the simulation. If you set the switches to permanently on and set the target signal to 10mV dc, then the circuit outputs a signal that is "near as dammit" zero volts.

Going back to a sine wave input and sampling the switches with main and EF pulses, if you move the 10k input resistors to the output side of the switches (instead of the input) then you can probe the switch outputs to see what is happening. Comparing the two outputs shows clearly that the differential opamp cannot possibly subtract the non-inverting input from the inverting and achieve a zero volt output as the two signals are in constant fluctuation.

The strange thing is that Hammerhead uses a similar configuration and I know for a fact that it can cancel the Earth field. The circuit used in the Minipulse is different in that the feedback resistors are connected to the switch inputs.
We will all need to think about this problem a bit more.

[green]

Changed the input to pulse. The +input has a sawtooth waveform(R11, C5). If the second sample is taken at the right time of the sawtooth the output is zero. Change the 589us sample time to 800us and see what happens. Think I see the problem now?

The strange thing is that Hammerhead uses a similar configuration and I know for a fact that it can cancel the Earth field. 2C integrator does reduce signal and cancels if sample delays are correct. Maybe good enough or just lucky.

[Qiaozhi]

Changed the input to pulse. The +input has a sawtooth waveform(R11, C5). If the second sample is taken at the right time of the sawtooth the output is zero. Change the 589us sample time to 800us and see what happens. Think I see the problem now?

The strange thing is that Hammerhead uses a similar configuration and I know for a fact that it can cancel the Earth field. 2C integrator does reduce signal and cancels if sample delays are correct. Maybe good enough or just lucky.

If you change the input signal to 10mV dc, and do the sampling, it doesn't completely cancel.
There is a dc level at B of about 2mV. Since the inverting gain of the opamp is 100x there is clearly some cancellation going on. It's just not that good.
I suspect you may have the EF amplitude set too high in the simulation anyway. I've never measured the EF signal level in practice but it's probably much lower than 10mV. Which would explain why some simple designs like PI-4 do not appear to have an issue with the Earth field due to their much lower gain.

[green]

If you change the input signal to 10mV dc, and do the sampling, it doesn't completely cancel.
There is a dc level at B of about 2mV. Since the inverting gain of the opamp is 100x there is clearly some cancellation going on. It's just not that good.
I suspect you may have the EF amplitude set too high in the simulation anyway. I've never measured the EF signal level in practice but it's probably much lower than 10mV. Which would explain why some simple designs like PI-4 do not appear to have an issue with the Earth field due to their much lower gain.

What do you get at b when 800us is changed to 589us?

Rocked about 1/2inch, my 300uH, 200mm diameter coil connected to my amplifier(gain=450). Easy to get 50mV peak, 100mV harder. Wondering what others get.

Try moving R3 and R11 to the input side of the switches. Cancels but I think other problems.

[green]

[QUOTE=Qiaozhi;280665
I suspect you may have the EF amplitude set too high in the simulation anyway. I've never measured the EF signal level in practice but it's probably much lower than 10mV. Which would explain why some simple designs like PI-4 do not appear to have an issue with the Earth field due to their much lower gain.[/QUOTE]

Tried to calculate the minimum value I can detect with my detector. Maybe correct maybe not. Think 200uV maybe 100uV at integrator in if coil pickup noise is low enough(.2uV to .4uV at amplifier Rx in). 2C integrator reduces input by at least 30 so I would need at least 3mV EF to detect. Not sure I did it correctly. Interested in what other members get.

[6666]

Which would explain why some simple designs like PI-4 do not appear to have an issue with the Earth field due to their much lower gain

No EFE for my PI-4

[Carl-NC]

PI-4 also has a HPF at the input of the preamp.

[6666]

PI-4 also has a HPF at the input of the preamp.

Is that C2+R4

[Qiaozhi]

Like waltr, my initial thought was that the circuit was sampling on different parts of the sine wave and this was the cause of the problem, but I was sidetracked into thinking there may be something wrong with either the simulation or the circuit itself.

The whole premise for EFE is that the early and late samples will contain the same amount of EF signal, and therefore a simple subtraction process can be used to remove the EF. You can prove that the cancellation process works in a simulation by using a dc voltage as the EF signal. However, in the real world the EF signal is not exactly the same during the early and late samples, and the EFE can become compromised. I suspect some circuits are better than others at providing good EFE, and green's use of a 1Hz sine wave seems good at exposing these limitations.

The question is: "How good does it have to be?".

[green]

Like waltr, my initial thought was that the circuit was sampling on different parts of the sine wave and this was the cause of the problem, but I was sidetracked into thinking there may be something wrong with either the simulation or the circuit itself.

The whole premise for EFE is that the early and late samples will contain the same amount of EF signal, and therefore a simple subtraction process can be used to remove the EF. You can prove that the cancellation process works in a simulation by using a dc voltage as the EF signal. However, in the real world the EF signal is not exactly the same during the early and late samples, and the EFE can become compromised. I suspect some circuits are better than others at providing good EFE, and green's use of a 1Hz sine wave seems good at exposing these limitations.

The question is: "How good does it have to be?".

1C integrator cancels 1Hz sine or dc voltage at different delay times(it samples the sine wave at different parts of the sine wave). 2C integrator doesn't cancel 1Hz sine or dc voltage at different delay times. The capacitors, feedback and +in to common effect error amplitude. The question is: "How good does it have to be?". Good question, how do we find the answer? ZED reply #55 seems to have a problem with EF.

[green]

Tried to calculate the minimum value I can detect with my detector. Maybe correct maybe not. Think 200uV maybe 100uV at integrator in if coil pickup noise is low enough(.2uV to .4uV at amplifier Rx in). 2C integrator reduces input by at least 30 so I would need at least 3mV EF to detect. Not sure I did it correctly. Interested in what other members get.

https://www.geotech1.com/forums/atta...6&d=1599429210 A scope picture of detecting a US quarter. Modified one of my spice simulations to match my detector circuit used to detect the quarter. 170uV in gives same signal as quarter. Should be able to detect 100mV signal at output, 85uV in. Similar to the 100uV in suggested with my first try.

[green]

Any dB/dt cutting through the coil will produce a signal. If the Earth field is perfectly uniform and you swing perfectly level then there should be no EFE effect. Rocking/twisting the coil will maximize the effect. Amplitude depends on coil turns & area, gain of preamp, etc. A good bench test method is to use a magnet, but it must have no eddy (metal) response. Many ceramic magnets will work.

Connected the 8inch 300uH mono coil to Rx in. Tx disabled. Holding coil N and S sides stationary and rocking, minimum signal change. Holding E and W sides stationary and rocking, maximum signal change. Bouncing a 1/2 x 3/16 ceramic craft magnet over the coil, a little more signal than rocking. Bouncing a 12 x 3 mm neodymium magnet over the coil, a little more than the craft magnet. Lot stronger magnet, expected to see a higher signal. I have a trim pot to adjust circuit for EF cancel. A .1% change on either input resistor is easily visible looking at integrator out with a scope while bouncing the magnet. Still don't know how signal amplitude sweeping the coil compares with rocking the coil or using magnets.