Inductive Proximity Sensor Design

[Electronics_chaitanya]

Hello all friends

I am trying to make inductive proximity sensor.Sensing distance of 4-5mm is sufficient for me but i am not getting success in design.

As per theories provided on internet oscillations should be stopped or at least decrease when metal object in keep near coil of sensor but in my design when metal object is kept near sensor frequency of oscillation changes due to change in inductance of coil.but amplitude of oscillation is not decreasing.

I am using simple colpitt oscillator and trying to observe if amplitude decrease or not.

I have following confusion

1) Why oscillations are not stopping? at which point i am wrong?
2) Which frequency should i use for oscillations? i am at 1Khz right now.(I am talking about frequency at which coil resonates.)
3) In some industrial designs i have seen tapped coil having 3 wires coming out i have even seen 4 wire coils(2 coils).What is reason of that?
4) Should i use Hartly oscillator or colpitt oscillator ? which are best suitable?
5) I have attached a circuit diagram having two coils.Is it works for proximity sensor design.I understand how it oscillates but i cant get to know what is function of inductor connected to base of oscillator transistor.



I have also attached oscillator i am using for my experiment as well as photo of inductor i have used.I cant found ferrite pot core.So core i have shown will work ? at least for this experiment? or i need to go for pot core?





Regards

 

[BradtheRad]

Your project resembles a metal detector. As I understand it, the commercial type contains two air coils oscillating at the same frequency. When metal (especially ferrous type) enters the field, it alters the frequency of one coil, creating a beat frequency. It might be in the low kHz, which makes it easy for our ears to detect. The amplitude does not need to change.

The coils are air core, I believe. If they contain metal, it reduces sensitivity of your detector to other metallic objects.

 

[FvM]

How do you approach the metal? There's little magnetic coupling to the zone above (in axial direction) the bobbin core.

For considerable eddy current generation and oscillator damping, the main field must perpendicularly hit the metal surface.

Industry standard inductive proximity sensors often use P core halves (e.g. Feroxcube PH series). To get oscillation stop, the oscillator feedback must be adjusted respectively.

 

[CataM]

To get oscillation stop, the oscillator feedback must be adjusted respectively.

In Colpits oscillator C1/C2 gives the feedback factor and in Hartley oscillator L1/L2 gives the feedback factor.
I think it is more suittable a Hartley oscillator because you adjust the oscillator to start when no metal is around, and when is around the feedback factor is changed and stops oscillation.

 

[Warpspeed]

The first requirement will be for the magnetic field around the coil to extend well beyond the 4-5mm you require.
Those ferrite cored chokes are far from ideal for that, air cored coil will produce a much wider external field.

It is going to take a rather large change in inductance to "reliably" stop oscillation, and the circuit to do it will probably be pretty temperamental and unreliable.

A much more sensitive method would be to measure small changes in oscillation frequency when the magnetic field around the coil is altered.

 

[FvM]

A much more sensitive method would be to measure small changes in oscillation frequency when the magnetic field around the coil is altered.

That's true of course, however 99% of industrial sensors are working well with oscillator damping method. I would check the specifications of existing sensors in a first place to get an idea of their sensitivity for a specific coil size. If it's sufficient you have good chances to implement your sensor the same way.

 

[CataM]

Yes... FvM and Warpspeed are right. Here is how they do it.

- - - Updated - - -

Oscillator and damping on Page 19

 

[Electronics_chaitanya]

Thanks for your information every replay is very useful to me.

@FVM i am putting metal near top part of core please see attached image to get clear idea about it.

I also want to mention that if i use two coils in proximity sensor then do i need to use two core (each core for each sensor)?

I am thinking to go for oscillation damping method as it seems quite easy to design as multiple coils in sensor makes its much time consuming and also increase cost of sensor although i welcome your suggestions.


Another idea comes in my mind is : if i wound two coils on a single ferrite pot core in which one coil is a part of oscillator inductor(tank circuit) and other is output coil.i mean when metal comes near core it cause megnatic coupling between these coils so i can get out put on coil 2(coupled from oscillator coil) and then demodulate and futhure process the signal.


I am becoming more and more confuse as not getting clear idea about using oscillation damping or detecting frequency diffreance .

One more thing colpitt oscillator i am using some time not staring oscillating but if i try to give power multiple times then it start say 1 try out of 5 tends to success oscillations . i think i should increase gain of oscillator.

If you have any referance circuit diagram i want to request you to upload here

 

[CataM]

Another idea comes in my mind is : if i wound two coils on a single ferrite pot core in which one coil is a part of oscillator inductor(tank circuit) and other is output coil.i mean when metal comes near core it cause megnatic coupling between these coils so i can get out put on coil 2(coupled from oscillator coil) and then demodulate and futhure process the signal.

Even without any metal around you would still have "output" on your second coil and when comes close the metal, also. In my view, does the same thing with or without the second coil, just wasting other coil. With one coil is enough.

If you have any referance circuit diagram i want to request you to upload here

In the reference I just uploaded they give some information.

 

[Warpspeed]

That image shown on page 17 of what looks like an E core is what you need.
The open end creates magnetic fringing between the pole pieces.
Put a piece of metal into that magnetic field and the change in inductance will be considerable.

The ferrite bobbin core you are using only has an extended field around the sides, not at the end where you need it.
Try slipping a metal tube over your coil, it may then work.

But f you want it to be sensitive at the end, a different coil will need to be used to get that.

 

[chuckey]

With your cores, the magnetic field goes up the centre then radially out through the top hat then back down through the air into the bottom top hat and back to the centre.
If that is a picture of the coil you are using, unsolder the wire unwrap it. The remove the top hat or at the least damage it so some of the field can leak our of the top. Rewind the coil. The inductance will have fallen dramatically, so the frequency of oscillations will have increased and the Q of the coil will have fallen, so the oscillator might not start. If it starts, it will tend to be very sensitive.
One alternative is an ordinary solenoid coil, such as from a relay, these will radiate their field from their ends ,cos' that is how they attract the armature.
Frank
Frank

 

[Electronics_chaitanya]

Hello all friends
I finally found the problem. Problem is in core i tried with ferrite pot core and i finally able to damp oscillations ....


Thanks for your help