Yet another shielding problem!!!

[James. O]

Hi folks, I have built new coils for my simple VLF IB and I'm having real trouble getting them balanced. The problem seems to be very similar to to the one described in this post

http://thunting.com/geotech/forums/s...ad.php?t=12629

I am using a standard colpitts osc transmitter, both coils are 100 turns at 6.2 inches, Fosc is ~16Khz. One end of my Rx coil goes through a 4.7k into the inverting input of the preamp the other end goes to a voltage divider at Vcc/2 and the non inverting input.

I can't get a completely null signal at the output of the preamp. Now heres the confusing bit. I have been running the transmitter and detector circuits from the same supply and get this problem, but If I power the transmitter from a seperate power supply and attach the shield wire to the negative rail of that supply then i can get a perfect null. If i then connect both the negative rails together the problem comes back! Huh??? Further more if I drop the frequency to around 8Khz i dont get the problem at all whether on 1 or 2 power supplies. Double Huh???

Anyone got any ideas as to what is going on???

[Sean_Goddard]

OK this is a shot in the dark, but I think you may have what is known as an "earth feedback loop" Try separating the grounds with an inductor (say 1mH or a 100 ohm resistor) or use a “star” ground arrangement whereby all the grounds meet as a COMMON point.

[James O.]

Nice one Sean, I think you've hit the nail right on the head. I done some more tinkering last night and come to the conclusion that the coils are forming an armstrong type oscillator through the power rails, but this only seems to occur when both coils are tuned to the same frequency. I have had some succes damping this feedback with an RC LP filter between the RX coil and virtual ground, but so far haven't eliminated it completely. You're right the easiest place to damp this signal is between the transmitter / reciever grounds, i'll give that a go.

Thanks again Sean.

[Geo]

Hi. If you want to null the coil connect the scope at the Rx coil , not at the preampl output. Connect a resistance load at parallel with Rx (1...2k) to reject the selfoscillations. Now try to null the coil. If the tune capacitor is in series with the coil then you will have a big output signal(100...200 mV). If the capacitor is parallel with the coil then the output signal will be 2...10 mv.
Good Luck

[James O.]

Hi Geo, Sorry i haven't been very clear in my previous post about the problem i'm having (i'm still trying to understand it myself!). I know where the null point is and can set the coils there quite easily if i use a seperate power supply for the Tx Osc. The tune capacitor is in parrallel with the coil but i'm not using any load resistor. There seems to be some self oscilation of the Rx coil that is causing the large output on the preamp - but only when connected to the same power supply as the Tx. So it seems to be an earth loop feedback thing as Sean suggested. I'm wondering if this is because both coils are tuned to exactly the same frequency (15.769 Khz), the further i tune the Rx coil away from this frequency the lower the recieve signal drops until i can null quite easily - but this also gives less sensitivity. Does anyone know if -Tuning both coils to EXACTLY the same frequency is a good or bad idea or is there a happy medium????

Suggestions much appreciated

[Geo]

Hi James.
I don't tune Rx and Tx exactly at the same frequency. I tune the Rx 5...10% higher. Once i tuned exactly at the same frequency then i had problems (sorry i don't remember exactly the problem).
Good luck

[Max]

Hi James.
I don't tune Rx and Tx exactly at the same frequency. I tune the Rx 5...10% higher. Once i tuned exactly at the same frequency then i had problems (sorry i don't remember exactly the problem).
Good luck

Hi Geo,
would be resonance problem with same TX and RX.
What's strange is that sometimes works great...like in my last DD coil that uses equals rx and tx coils. Same wire, same number of turns , same shape, same shielding ...all the same and frequencies really close.
But it works great. Don't know why.

Best regards,
Max

[Max]

Hi Geo, Sorry i haven't been very clear in my previous post about the problem i'm having (i'm still trying to understand it myself!). I know where the null point is and can set the coils there quite easily if i use a seperate power supply for the Tx Osc. The tune capacitor is in parrallel with the coil but i'm not using any load resistor. There seems to be some self oscilation of the Rx coil that is causing the large output on the preamp - but only when connected to the same power supply as the Tx. So it seems to be an earth loop feedback thing as Sean suggested. I'm wondering if this is because both coils are tuned to exactly the same frequency (15.769 Khz), the further i tune the Rx coil away from this frequency the lower the recieve signal drops until i can null quite easily - but this also gives less sensitivity. Does anyone know if -Tuning both coils to EXACTLY the same frequency is a good or bad idea or is there a happy medium????

Suggestions much appreciated

Hi James,
think that Geo is right on avoiding same coils. But sometimes it works to have the same. Also in some older projects (e.g. Buccaner and many others) there are equals coils for tx and rx. Think that sometimes input circuitry don't like too much that rx works on resonace with tx. Could be also interactions between tx and rx that may increase in worse ground conditions too.
What to say, follow Geo advice first then if you have time (and wire) you could experiment with equals coils too.
E.g. I found in my last homemade DD coil I have the best results with equals tx and rx ! That means that in some cases this could be the right solution where in other the worse thing to do: need experimentation, so start with RX a bit more inductive.

Best regards,
Max

[James O.]

Sorted !!!

Hi guys, Thanks all very much for your valuable advice. I have finally fixed the problem, and so if you guys or anyone else with a similar problem in the future want to know the solution its this - use 2 caps in series to tune the Rx coil and then connect the center tap to the negative rail. This couples any stray Tx noise to both ends of the Rx coil effectively cancelling it out. My machine is now so sweet, sensitive and stable, it will detect a British 2 pence piece at 14 inches. Hope this helps someone somewhere sometime.

Cheers all, James O.

[Max]

Hi guys, Thanks all very much for your valuable advice. I have finally fixed the problem, and so if you guys or anyone else with a similar problem in the future want to know the solution its this - use 2 caps in series to tune the Rx coil and then connect the center tap to the negative rail. This couples any stray Tx noise to both ends of the Rx coil effectively cancelling it out. My machine is now so sweet, sensitive and stable, it will detect a British 2 pence piece at 14 inches. Hope this helps someone somewhere sometime.

Cheers all, James O.

Hi James,
really interesting. I'll try it out soon on my bandido's coils.

Best regards,
Max