Need some help Virtual GND circuit.

[mickstv]

In the last few days I've been working on my mono coil only bipolar frontend in the picture below. It uses a battery pack of 14.40 volts and I'm using a audio IC to create a "virtual ground" which connects to one side of the TX coil. The output from the TX coil goes into a standard NE5534 gain 100 and another inverting NE5534 running 0 gain both IC's outputs go into a 4066 switching IC and then a integrator "gain330". This cancels out the Earth Field, but adds the target signal.

Ok so the Virtual Ground point is my common ground which is also the coils shield point on the Minelab coil I'm testing it with.

The problem I'm having is the coil shows a response if I move my hand over it "nearly touching maybe 1/2 a cm away. The level of movement is about 1-3mv DC as measured directly on the output of the integrator using a multimeter. Does that sound ok, I was under the impression that the coil shield would totally stop any interaction with ground response "hand movement etc" I should also mention the sample point is about 10us after TX for both positive and negative pulses.

Any thoughts on this situation would be appreciated.


Sorry for the rubbish picture I was just after the quickest method to post the frontend circuit.

[kt315]

...

[Qiaozhi]

The problem I'm having is the coil shows a response if I move my hand over it "nearly touching maybe 1/2 a cm away. The level of movement is about 1-3mv DC as measured directly on the output of the integrator using a multimeter. Does that sound ok, I was under the impression that the coil shield would totally stop any interaction with ground response "hand movement etc" I should also mention the sample point is about 10us after TX for both positive and negative pulses.

How stable is your virtual ground?
Check with a scope connected between virtual ground and the -ve battery terminal

[mickstv]

How stable is your virtual ground?
Check with a scope connected between virtual ground and the -ve battery terminal

Just checked and the Virtual ground compared to -ve is smooth the cro was set to AC and 5mv per div.

I just removed the virtual ground circuit and tried with the batteries alone with a centre tap at 7.2 volts, this didn't make any difference to the outcome. I still have a very slight hand capacitance problem.



Mick

[mickstv]

Here's a picture of the Virtual Ground as compared to -ve. The scale is 50mv per div.

I should also mention what I'm trying to do is build a bipolar detector that doesn't revolve around the twisted pair coil design that I'm working on. Going to a mono coil configuration would then mean pretty much any off the shelf coil will work, with maybe just a change in the damping resistor.

[Altra]

Hi Mick,

Is it possible you have the coil wires reversed?

[mickstv]

Hi Altra, how are you going.

No I just checked again and it's correct. Also this minelab coil works ok on it's original detector.

Maybe it's my rough breadboard design thats causing the problems.

[Midas]

Hi Mick,

Your positive and negative pulses don't look entirely symmetrical to me. The negative seems to have a slightly higher amplitude as well as having a more linear rise. I've no Idea how you’re going to achieve it, but I would guess that symmetry is really important to achieve a complete balance.

Regards
Midas

[mickstv]

Hi Midas, the symmetry being wrong might be something to do with the P/N fets, I'm using a IRF9640 and a IRF630.

Also when monitoring either sample with the opposing disconnected, it shows a slight difference in level, I don't see this as a problem because the target samples add together. The Earth field still totally cancels out even with the slight difference between channels.



Mick

[mickstv]

Ok another test I have an unshielded coil if I monitor the output of the preamp I have the typical fuzz superimposed on the receive. With the minelab coil connected most of the fuzz goes away.



Mick

[Altra]

Maybe it's ok, and detecting the salt in your hand. Try the same test sampling at 12-15uS. That should tell you.

[mickstv]

Hi Altra, still the same problem.

I did notice another issue, at the output of the clamping diodes the waveform looks ok, but after the first preamp one the n fet side looks like its taking a 1us longer to decay, very strange.


I'll rebuild the circuit on the breadboard and shorten all the leads down to reduce any capacitance issues and change the frontend preamp and see what happens. I'll post up some more waveform pictures after I've finished.


Cheers
Mick

[moodz]

been there and done that ....

The P fets and N fets gate / source bias move in opposite directions this causes a fundamental unbalance ... you will only get symmetry with dual N or dual P switches.

moodz.

[moodz]

Here's a picture of the Virtual Ground as compared to -ve. The scale is 50mv per div.

I should also mention what I'm trying to do is build a bipolar detector that doesn't revolve around the twisted pair coil design that I'm working on. Going to a mono coil configuration would then mean pretty much any off the shelf coil will work, with maybe just a change in the damping resistor.

any old coil wont work with bipolar because the sheild is only attached to one side of the coil and this is a fundamental unbalance that you cant undo electronically.

moodz

[mickstv]

Hi Moodz, I just assumed that it would work because the shield is normally connected to VCC in a nchannel frontend and GND in a Pchannel frontend. I just expected Virtual Ground to be the common ground for both.







Cheers
Mick