Ground balance

[6666]

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.

Interesting experiment, when you rock the coil do you see much of a signal at preamp out ?

[green]

Interesting experiment, when you rock the coil do you see much of a signal at preamp out ?

Preamp out is what I was looking at. Didn't see the change at integrator out until I adjusted EF trim pot.

[ZED]

The integrator circuit i use is the same as the one shown in the 2C sim,except with different value components,frequency's of interest to me are from 0.1 to 5 Hz,in the past i used single ended integrators,i feel my current setup does a better job of tuning out the EF.
I use a ferrite speaker magnet when trimming the EF and like to get the magnet touching the coil without sounding off,some sample timings will lend themselves to that effect,others not so much.
Its not uncommon to bump into earth field effects when detecting out in the field,particularly with large coils,its most notable when raising the coil from ground level up to the 12 o'clock position,many retail detectors will exhibit this effect as well
I also wonder whether the high mineralization in this region of the globe would increase the density of the earths magnetic field and make EF noise more pronounced.

[liekwang]

I've seen the ground balance 3 sample schema schema in handwriting in this forum too

[green]

The integrator circuit i use is the same as the one shown in the 2C sim,except with different value components,frequency's of interest to me are from 0.1 to 5 Hz,in the past i used single ended integrators,i feel my current setup does a better job of tuning out the EF.
I use a ferrite speaker magnet when trimming the EF and like to get the magnet touching the coil without sounding off,some sample timings will lend themselves to that effect,others not so much.
Its not uncommon to bump into earth field effects when detecting out in the field,particularly with large coils,its most notable when raising the coil from ground level up to the 12 o'clock position,many retail detectors will exhibit this effect as well
I also wonder whether the high mineralization in this region of the globe would increase the density of the earths magnetic field and make EF noise more pronounced.

Interesting that a 2C does a better job for you than a 1C. Do you have a EFE trim pot? Either 1C or 2C should cancel with a trim pot but the simulation suggests the 2C doesn't cancel if the sample timings are changed without retrimming. Including a schematic of the integrator I'm using. 1 turn of the 100 ohm trim pot adds 5R to one input and subtracts 5R from the other. 0.1% change, can see effect at integrator out on the scope when bouncing a magnet over the coil. Not suggesting integrator frequency response is ideal but it's what I'm using at the moment.

[ZED]

I use a micro-controller to adjust sample widths plus i store upto 16 individual timings,so using a trim pot doesn't help because its adjustment wont carry over to the next timing.
In the beginning i used a 3.9k input resistor on the diff integrator and adjusting EF was a finicky affair,i now use 5.1k or more and now EF adjustment is easier,also i use 0.1% resistors,this
helps when i move onto the next revision,no EF adjustments are necessary on the later revision.
i would like to edit my previous post,freq of interest is more along the lines of 0.5 to 5 Hz.

[Elliot]

Interesting that a 2C does a better job for you than a 1C. Do you have a EFE trim pot? Either 1C or 2C should cancel with a trim pot but the simulation suggests the 2C doesn't cancel if the sample timings are changed without retrimming. Including a schematic of the integrator I'm using. 1 turn of the 100 ohm trim pot adds 5R to one input and subtracts 5R from the other. 0.1% change, can see effect at integrator out on the scope when bouncing a magnet over the coil. Not suggesting integrator frequency response is ideal but it's what I'm using at the moment.

Hi Green, What are the advantages of having a 3 stage integrator circuit as you have ?

[6666]

If the high pass filter on the front end of the PI4 eliminates EFE, why are we messing around with EFE pulses ?

[Elliot]

If the high pass filter on the front end of the PI4 eliminates EFE, why are we messing around with EFE pulses ?

Good question! Trouble is that they are not brick wall filters and have a first order roll off.

[green]

Hi Green, What are the advantages of having a 3 stage integrator circuit as you have ?

Good question, only reason I can think of. If I want more or less total gain, can change input resistor(100k) to second stage and not effect frequency response.

[green]

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

Hi Elliot
Looked at Excel analysis again, looks like A2 gain=2 is a little narrower than A2 gain=1. I have been using A2 gain=1. Adjust first delay to desired delay time and adjust target sample to balance ground signal with second delay=4us and ground sample=100us. Did an Excel analysis, adjust first delay to desired delay time and adjust target sample to balance ground signal with second delay=20us and ground sample=50us. A2 gain=2, a little narrower, maybe a little more noise?

Wondering what sample times others adjust and leave fixed when ground balancing.

[liekwang]

I found this schema is anyone already using it,

[green]

Tried a spice simulation using analog switches. What am I doing wrong?

[ZED]

Hello green

Looking at your sims mate,now i am no where near as savvy as you when it comes to LTspice but as far as i can tell sim "A2=1_sw" appears to be ok,i get similar wave forms on my project in noisy environments,as for sim "A2=1_ADG1612" well it appears to be fine as well,its doing exactly what its supposed to be doing,except the circuit has a problem...Charge injection ! Charge injection is a characteristic of analog switches,each time the switch turns on or off it adds a small charge to the signal,and this charge always has one polarity,this results in a DC offset appearing on the output of the integrator,the more samples you have within a cycle the bigger the offset gets,some analog switches are nasty when it comes to charge injection,but there are some good ones as well,unfortunately the integrator in your sim has a DC offset on the output of about 540 mv,that's enough to drive the following stage to the rails,to continue with your sim just reduce the gain in the last stage.

[KingJL]

First, are you sure you want to use the ADG1612 (vs. maybe the 1611)?

on edit: corrected ADG vs ADC.