MONOCOIL SUCCESS

[moodz]

Hey, what's up?

I couldn't verify the spice simulation results however (schematics from post 14 http://www.geotech1.com/forums/showp...1&postcount=14 ). I'm sure, I didn't make anything wrong. I have used the real spice models (IRF740) rather than the internal ones from the library.
???
Aziz

Aziz ... are you sure your library models are working OK ..... add on libraries sometimes dont work because they dont model underlying ideal device behaviour correctly.

Anyway here is the file ... make sure you run the sim till the voltage reachs 16 volts on C1 ( stable operating condition ).

Dont forget .. I ran this on the "real" simulator .... using real parts in a real circuit

moodz

MONODIFF_SIM.zip

[Aziz]

Hey, what's up?

I couldn't verify the spice simulation results however (schematics from post 14 http://www.geotech1.com/forums/showp...1&postcount=14 ). I'm sure, I didn't make anything wrong. I have used the real spice models (IRF740) rather than the internal ones from the library.
???
Aziz

[WM6]

Thanks Aziz ... I am wary of sims ... I did build and test this circuit before I simmed it and managed to get it up to 800 volts before I blew the mosfet up ( which was supposedly rated at 600 volts ).

moodz.

Hi moodz, do you need to change something in code or last version is applicable here too?

[moodz]

Thanks Aziz ... I am wary of sims ... I did build and test this circuit before I simmed it and managed to get it up to 800 volts before I blew the mosfet up ( which was supposedly rated at 600 volts ).

moodz.

[Aziz]

Hi moodz,

be careful, when using the (internal) semiconductors of LTspice library!!!!!!!
The diodes, mosfets (and transistors) didn't implement the break down behaviour (and very likely the other parameters as well).

Use the part's individual and specified spice model from the part's manufacturer in this case.

Cheers,
Aziz

[moodz]

..blocker for negative going flybacks.

... on the subject of blockers and if you have a negative flyback pulse for alternate circuit topologies there is the circuit below however P Mosfets are usually not available in high breakdown voltages.

moodz

[moodz]

MONODIFF2

.... here is another ( to be ) tested improvement ..... the differential head, damping and blocker circuit can be located in or very close to the search coil ... This will eliminate the flyback pulse having to travel up the coil cable to the control box ..

The blocker transistor Q2 should have a low "ON" impedance.

and again it is the most simple change ...

PS the TX switch device does not need a high voltage rating now ... the blocker handles the flyback volts

moodz

[moodz]

And the sims work as well ...

I ran the sims after I did the prototype ... and it works swell too.

note I ran a smaller tank cap on the blocker to speed up the sim runtime.

moodz

[moodz]

Here is the circuit ( as tested )

Thanks you guys ... well here is the circuit. ( told you it was simple )

Tepco dont read this because it does not work .. it will work for everyone else though.

For the coil there are two states ( TX and RX ) D1 and D2 provide current path through coil during TX and isolate it during RX. ( this is important )

R2 is the "real" damping resistor however R1 is provided to balance the coil during RX. ( balancing the coil is very important ) By keeping the coil balanced the differential amplifier will reject a very high degree of EMI and make it possible to operate with no coil sheilding on simple detectors.

The blocker is formed around Q2 ... it is very simple but provides diode speed blocking of voltage spikes reaching the input of the amplifier. It is not needed on both inputs as the flyback is only on one side of the coil.
Q2 is "on" at all times but goes open circuit or "OFF" when the voltage input exceeds the gate + threshold voltage being maintained by C1.
Q2 should ideally have a high breakdown at least equal to the TX switch Q1 and a minimal gate capacitance.... though I tried a "power" device here and it worked swell.

C1 is charged via D3 and R3 which impose a very minimal loading on the flyback pulse. This also forms a saftey delay after turn on as until C1 reachs sufficient charge Q2 remains OFF.

The blocker circuit can be used in any situation where you want to protect a sensitive input such as an amplifier. It is not specific to this diff circuit ... use a diode clamp if you want but make sure the circuit is balanced ( coils sees same circuit loads on both sides )

the coil is your standard monocoil ... I would recommend that if the coil is sheilded that the sheild is not connected to one side of the coil. Should be connected to ground.

moodz

[Aziz]

Hi all,

there are two major improvements with the blocking fets:
- Lowering the source impedance -> lowering the thermal noise (very important on bipolar input)
- Not affecting the damping resistor (time constant of the coil remains constant)

Hey, forgot the other major improvement:
Blocking the high voltage beast. One can let increase the beast.


Proper ground balancing/discrimination is a whole different ball game. It hasn't do to with the capability of ealier sampling.
Good job Paul. I appreciate your effort. Keep up the good work.

Cheers,
Aziz

[mickstv]

Good stuff Moodz.


I guess this mono front end is similar to the diff frontend in that we no longer need input resistors, which would normally add to noise to the frontend.




Mick

[moodz]

Very nice. Especially good for detecting metal targets suspended or levitating in mid air. However, in real ground will fail miserably at 3uS sampling.This pic is from one ordinary PI and coil. Speed limit applies, i blasted off myself more than once this way. Question is, what is the minimal sampling time useful in reality, not what can be achieved technically, and seems to be maybe 8uS or somewhat faster, this can be easily achieved whit regular detector and proper coil design
PS what is the weight of 18" coil made from 2mm wire, looks like nice gym equipment?

I am not sure what your point is, Tepco ...

Last things first .... 16 turns of 2 mm 2 wire ( AWG 14 ) is about 75 feet of wire that comes to about 1 lb or 454 grams ... The roll I used said 500 grams ... The whole coil weighs 690 grams with the mounting hardware and bash plate ??? This is a largish coil sure ... but "gym equipment" ... give me a break.

Secondly the whole point of this exercise is to determine if an improvement can be made over existing techniques ... you cant use exisiting ideas to prove that new ideas dont work If the new ideas improve the performance then ergo they must be better.

Early sampling is probably not going to improve target detection per se however the descrimination information is not the same as a overall target response .. there are bulk magnetic effects ( inductance / diamagnetic etc ) AND there are eddy current effects .. NOT the same thing.

Being able to sample early was not the point of this exercise .. it was only an excercise to apply some learnings from my differential coil experiments and try to improve the state of the art for monocoil front ends.

moodz

[Midas]

Very nice. Especially good for detecting metal targets suspended or levitating in mid air. However, in real ground will fail miserably at 3uS sampling.This pic is from one ordinary PI and coil. Speed limit applies, i blasted off myself more than once this way. Question is, what is the minimal sampling time useful in reality, not what can be achieved technically, and seems to be maybe 8uS or somewhat faster, this can be easily achieved whit regular detector and proper coil design
PS what is the weight of 18" coil made from 2mm wire, looks like nice gym equipment?

Although I'm sure you've got some valid points there Tepco the sarcasm seems a little unnecessary, this isn't the LRL forum. In any case low capacitance coils have always seemed to me to be a bit of a compromise (since moving the windings apart has to be reducing the intensity of the magnetic field created), if you can reduce your sampling time (to whatever is reasonable in real ground) without having to use them there may still be advantages in sensitivity. Or even if that's not true, just being able to use cheaper, easier to build coils and getting the same speed and sensitivity would be an advantage.

Anyway, I'm still interested.

[Tepco]

Very nice. Especially good for detecting metal targets suspended or levitating in mid air. However, in real ground will fail miserably at 3uS sampling.This pic is from one ordinary PI and coil. Speed limit applies, i blasted off myself more than once this way. Question is, what is the minimal sampling time useful in reality, not what can be achieved technically, and seems to be maybe 8uS or somewhat faster, this can be easily achieved whit regular detector and proper coil design
PS what is the weight of 18" coil made from 2mm wire, looks like nice gym equipment?

[moodz]

...flyback blocking

Below is a pick of flyback blocking with input and output of blocker shown ... obviously there is a difference in voltage scaling ... the input is 400 volts peak.

( no resistors were harmed in the making of this picture )

moodz