bipolar pulsing to remove Earth Field

[6666]

Good work Mick

[mickstv]

Here's the improved version using a single powersupply but still using p/n mosfets. this version uses Moodz twisted pair coil, but not in the center tapped configuration. The cro picture is of the output of a NE5534 preamp running at about 300 gain and you can see it's settling at about 10us, using two stages I can get it to settle at about 8us with a gain of about 600.


The CRO shot is at 50us per div

[WM6]

Good work Mick

Yes, again! Electronically solved 8-shape coil.

Do we need perfect symmetry and what cause if symmetry less than perfect?

[mickstv]

Yes, again! Electronically solved 8-shape coil.

Do we need perfect symmetry and what cause if symmetry less than perfect?

Thanks WM6


At this stage I don't think symmetry is an issue. I've done some further testing by quickly adding the second preamp and inverting amp then into the 4066 and combining the first sample from each pulse via a 200ohm multiturn then pot into a integrator. The 200ohm pot is nearly in the middle for EF cancel and if you feed either input from the 4066 to the integrator directly I only get half as much level on a given target.

So what I'm trying to say is I'm getting pretty much double the signal whilst cancelling out EF completely.


Also you don't need a damper resistor on the Pchannel, If you did an an extra 390 ohm resistor to that side all it does is extend the time taken before first sample can occur.



Mick

[6666]

Mick what configuration are your amps ?
thanks
6666

[mickstv]

Mick what configuration are your amps ?
thanks
6666

Standard op amp at moment NE5534, so pin 2 goes to the input resistor and pin 3 goes to the B+ just like in the HH circuit.

The output of the first amp goes into a non inverting amp they both have 33k resistors on them then the output of the second amp goes into a third amp which is inverting and unity gain, I used a pair of 10k resistors for that. The second and third op amp outputs goto a 4066 and the output of the 4066 is connected to a 200ohm multiturn pot which then runs of to an integrator. These first two samples one from each pulse are for target sample - EF only. The detector also needs a second pair of samples which will be GB-EF. The the output of both (target-EF and GB-EF) will get diff integrated and hopefully work to cancel out mineralized noise.


The spacing between TX pulses at the moment is about 150 but it can be less just so long as they don't happen at the same time.

The Pch mosfet is only used to generate the alternate pulse and the signal is inductive coupled to the Nch side.



Mick

[mickstv]

Here's a rough idea of what I'm doing, Once I sort out a couple of small issues I'll post up a proper circuit for the preamp and pulse integration.

Basically any EF generates a response that's always in the same direction, whereas a target will have a positive and negative response depending on the polarity of the TX. So if you invert the signal and cancel between the two transmit pulses the EF becomes EF-EF which will equal 0, whereas the Target doubles.


Diagram isn't 100% correct but it gives you the basic idea





Uploaded with ImageShack.us

[Altra]

Hello Mick
Have you field tested your design yet?

Got busy and left mine alone for a bit. Waiting for some parts and I'll finish up the front end. Then I can move on, try some different experiments with adc's and demodulators.
Will post when I get something worth while.

I still want to see your original circuit. If you were using 2 n-ch fets and a center tapped Tx coil to generate your bipolar signal, it would interesting to see how a differential
amp could improve the Rx.

Regards

[moodz]

...good work mick ...see from 2009 vintage post
http://www.geotech1.com/forums/showp...&postcount=175

Moodz

[Altra]

Moodz thats what I was thinking. "Vintage", we need an index of all your ideas.

[mickstv]

...good work mick ...see from 2009 vintage post
http://www.geotech1.com/forums/showp...&postcount=175

Moodz

Hi Moodz, I actually did try that method using the standard config and using two nch mosfets, but like always I couldn't get the circuit to sample in less than 10-15us (with no amp direct off the diodes).

With the configuration shown in the hand drawn circuit earlier I can get the sampling to start in less than 8us from TX off (seen directly of the diodes with no amp).


So this circuit isn't really a moodz coil configuration



Mick

[mickstv]

Hello Mick
Have you field tested your design yet?

Got busy and left mine alone for a bit. Waiting for some parts and I'll finish up the front end. Then I can move on, try some different experiments with adc's and demodulators.
Will post when I get something worth while.

I still want to see your original circuit. If you were using 2 n-ch fets and a center tapped Tx coil to generate your bipolar signal, it would interesting to see how a differential
amp could improve the Rx.

Regards

Hi Altra, no the N-ch only version isn't using a centre tapped coil it is a true mono coil single winding. It's not going to be posted until I sort out a couple of self destruct issues.



Mick

[Davor]

You mean features, right?

[mickstv]

You mean features, right?

What are you talking about Davor

[Davor]

Just kidding. Didn't you notice the LOL smiley next to my post?
In too many films with superheros etc. there were gadgets with self destruct feature so i just couldn't stop laughing.