Interference problem of the circuit

[kt315]

single TX splitted in three sections. all sections connected serially.

[Carl-NC]

You can either synchronize the 3 circuits exactly so they all transmit and receive at exactly the same times;
Or you can synchronize them so the 3 circuits alternate.
Either way, you gotta synchronize them, and none of the RX samples can happen during TX. IMO, the first option is the easiest.
Shielding doesn't matter much. Mostly helps for external EMI.

In most walk-thru designs, there is one big TX coil and multiple RX coils fed to multiple RX channels, all of them sampling at the same time. The TX coil is on one side of the gate, the RX coils on the other side. Actually, most have TX & RX on both sides, what I call a "cross-fired" arrangement.

[DJSUOL]

ok i will try to use one 555 to drive three circuits

[DJSUOL]

I use one 555 timer circuit for all three circuits but only one circuit can work.


Here is my circuit. The PCB connected to the middle coil has 555 timer, However, only in the top one the speaker has sound. The other two circuits does not have sound.
What is the problem of this?

[waltr]

You must connect O'scope and Troubleshoot.......

[dbanner]

I use one 555 timer circuit for all three circuits but only one circuit can work.
[ATTACH]45773[/ATTACH]
[ATTACH]45774[/ATTACH]
Here is my circuit. The PCB connected to the middle coil has 555 timer, However, only in the top one the speaker has sound. The other two circuits does not have sound.
What is the problem of this?

Very cool, you have made serious effort.
Remember not to disconnect the pulses to the 4093's. Hope you didn't pull the other 555's
I would have thought that cross receiving between adjacent coils so closely placed together would have been an issue. But you clearly have more experience than me with walk through.

[dbanner]

When you finally get the thing working, and I am sure you will, consider square coils instead of circular, so there can be no blind spots.
Good luck.

[DJSUOL]

Actually I think the adjacent coils have interference. How to solve this?I have made Faraday shield on my coils

[dbanner]

When I speak of cross interference, what I mean is the middle Rx picking up from top and bottom TX when there is object in either the top or bottom plane.
This might happen if the tx pulses spread out beyond their intended horizontal zones. But I only speculate. You are the one who is doing the work, you will find out for yourself.
For me it is much better for single tx above, single Rx at bottom below the feet. Where synchronized, Rx becomes tx and vice versa, switching back and forth.
Just an idea.

[DJSUOL]

Could you please tell me what is the best way to eliminate the interference between two adjacent coils? The middle pair of coils always make the lowest output and it is sensitive when the metal passes the other pair of coils.

[dbanner]

If someone passes through with big knife in the waste, this may get a response from both low and middle. But so what? What does it matter if one or all are triggered, point is that it's done it's job.

[Qiaozhi]

If you are certain all three detector circuits are functioning correctly, if used independently, then try using one big coil as Carl suggested.
In other words, Using one TX circuit to drive a big coil on one side of the gate, with three RX coils on the other side. Obviously all three RX circuits will need to be synchronized to the single TX.

[Ferric Toes]

Could you please tell me what is the best way to eliminate the interference between two adjacent coils? The middle pair of coils always make the lowest output and it is sensitive when the metal passes the other pair of coils.

One thing that is not clear from your posts is how the coil arrangement is connected. Is there a TX coil on one side with a separate RX coil on the other side of the archway? It appears this may be the case as two sets of leads go out from each pcb. Are the TX and RX coils all set to give the same polarity field and the correct polarity output from the three RX coils? If this is the case then an object placed in the field of any coil will give a weaker but reversed response in the other two channels. This would be the same whether all TX coils pulse together or are sequentially pulsed. This effect cannot be shielded against. One point about the leads from the PCB's to the coils. They really should be coax cables but they appear to be just wires which would contribute to interaction and noise problems.
I would use square or rectangular coils sitting tight to one another. In the TX this would be equivalent to one TX of three times the height as the currents in the adjacent sides, when all pulsed together, cancel out any developed reverse magnetic field. Each RX coil should then have the same polarity signal but be considerable stronger when the target is in the receive zone of that coil.

Eric.

[Carl-NC]

Are the TX and RX coils all set to give the same polarity field and the correct polarity output from the three RX coils?

Eric brings up a crucial point. You want all 3 TX coils polarized in the same direction, and then all 3 RX coils also polarized in the same direction. You can test the TX coils by looking at the output of a single RX coil that is placed in each zone (even using an oscope). Then test the RX coils by placing a target in each zone and ensuring they all produce the same polarity result.

Here are some ideas for coil arrangements:



It sounds like you are using the "GOOD" arrangement. It will work, but you will have dead zones between the coils. The BETTER design is what Eric suggested; make the coils rectangular and with no gap. But there is no reason to have 3 TX coils, so the MORE BETTER approach uses one big TX coil.

MORE BETTER is not BEST because there are more improvements you can make. Even with the RX coils jammed together it will still have weaker detection at those points. So overlap the coils "a bit" to get even sensitivity at the transitions. "A bit" is trial & error, unless you are good with magnetic FEM software. Another improvement is to make each RX coil a "figure-8" coil (aka butterfly) turned sideways. This offers a first-order induction balance (not very important) and far-field EMI rejection (VERY important). Another improvement is to duplicate the entire TX-RX coil system, but pointed in the other direction. This doubles the zones from 3 to 6. Finally, you can interpolate between coil zones and create even more (virtual) zones. If you interpolate between top/middle and middle/bottom this gives you 5 vertical zones. If you interpolate between left & right you get a middle horizontal zone. This can turn a 6-zone detector into a 15-zone detector. This is commonly done in walk-thru designs.

[DJSUOL]

Thanks for your help. But why the square coils do not have dead zone? What is the advantage of it?