Need some help Virtual GND circuit.

[mickstv]

I know this is dumb, but are you wearing a ring or a watch? Have you tried it with other people?

No metallic objects anywhere, Happens with more than one person, I also managed to get my cat to walk over the coil same problem.

I'm going to disconnect the Nchannel side of the circuit and see what happens.



Cheers
Mick

[mickstv]

Hi All, it looks like I might have fixed the problem. I changed the 2200uf caps for a pair of 4700uf caps in the same position but this didn't fix the problem. I then decided to move the 4700uf caps as close to the coil socket as possible. about 3cm away and all other leads were shortened.

This has fixed the problem with hand capacitance, It's quite possible that the 2200uf caps would work also in the same location.

I'll do some more testing tomorrow and post up some CRO shots etc.





Mick

[Qiaozhi]

Hi All, it looks like I might have fixed the problem. I changed the 2200uf caps for a pair of 4700uf caps in the same position but this didn't fix the problem. I then decided to move the 4700uf caps as close to the coil socket as possible. about 3cm away and all other leads were shortened.

This has fixed the problem with hand capacitance, It's quite possible that the 2200uf caps would work also in the same location.

I'll do some more testing tomorrow and post up some CRO shots etc.

Mick

That makes sense.
So it was a virtual ground problem after all.

[mickstv]

That makes sense.
So it was a virtual ground problem after all.

Yes and No. I actually think most of the problems related to my rough design. The working circuit is actually mounted on a unetched PCB and I've used the heaviest wires possible for all connections relating to the Virtual GND and P/N fets.

The current version is running a TX of about 100us and a sample delay of 8us, the delay between the two transmit cycles of about 300us.

The only other issue is a slight difference in target signal level between the P and N channels, but it's only very slight. I don't believe this will be an issue because the targets are added together.

The Earth Field always cancels out.




Mick

[Davor]

I'll be following your work on this monocoil solution. It fixes many issues by a simple arrangement.

[mickstv]

I'll be following your work on this monocoil solution. It fixes many issues by a simple arrangement.

Hi Davor, no worries. With any luck one of these designs might turn into a reasonable forum project. It just needs to be simple to build using standard analog processing and able to cope with mineralized ground.



Mick

[Davor]

IMHO the bipolar pulsing is a missing spice to the perfect dish. It solves problems like Earth Field by its very nature, hence it will work well with faster repetition and perhaps lower excitation. And all of it using a handy monocoil.
Perfect.

[Altra]

Hi Mick,

Here's a picture of what I did to avoid bread board problems. Its almost the same as the schematic I posted under the original bipolar pulsing thread. I added the dual fet driver TC4427A, with 2 pnp level shifters and jumpers so it can be switched highside or low side. The preamp is from the GS5 with a gain about 500. I found after experimenting with virtual ground circuits, it worked just as well without a supply splitter and just the two electrolytic caps. As long as you keep the n and p switching times symetrical the caps seem to maintain a good virtual ground. This is all on the pcb. Almost the same timmings as you 120uS pulse at about 3.1khz. The supply is 14 volts regulated to 10v, with the virtual ground it gives you a +/-5volts for your analog without additional regulators

I bread boarded the demodulator from the GS5 schematic and added a vco and some other things missing from the schematic. It was sensistive and could ground balance. Never had any problems with detecting the hand.

[mickstv]

Hi Altra, I'm impressed, thanks for posting the picture. I'll post up some pictures of my "Manhattan Style" board later.


I actually noticed whilst I was trying to solve the Virtual Ground problem that it would run with only the caps connected, I also noticed less whistling coming from the coil and lower current draw, also slightly less depth.


Cheers
Mick

[Altra]

Hey Mick,

The minelab coil (11inDD) caused my circuit to be unstable. Need to limit the current.
I am using a 3 ohm coil vs minelab 1/2 ohm. "Manhattan Style" thats funny we used to call it "Dead Bug"

Cheers

[mickstv]

Hey Mick,

The minelab coil (11inDD) caused my circuit to be unstable. Need to limit the current.
I am using a 3 ohm coil vs minelab 1/2 ohm. "Manhattan Style" thats funny we used to call it "Dead Bug"

Cheers

The minelab coil I'm testing on seemed to run ok, did you try adding some series resistance to the coil. With your Virtual Ground are you using a standard 78series reg ?


Mick