Houston...I have a problem! (bandido coil)

[Esteban]

Please, see this revised schematic based on Carl's schematic and Gary components. Also, list of materials.

[Qiaozhi]

Some people of Europe has success with the Bandido (hope this is right because has finished the project). In the files there are the same coils, etc. In very detailled parts list you can find replacements, bla, bla, bla. This is the detector based on split PCB of this people (maybe you want all the info) PCBs are in Eagle program and all the documentation in .rar compressed files. You can compare with all has been done by you at this moment. Remember that the second TX coil is winding in reverse, so big TX and receiver are in the same orientation winding and secondary TX (bucking coil) is in reverse winding (or phase).

Nice work Esteban!

[Qiaozhi]

Hi Gary,
"The Frequency on the Recieve coil has Nothing to do with the Inductance or the Parallel Cap."

Yes, I know. But I mean using Mike's instructions for measuring resonance frequency of TX and RX coils when connected to TX oscillator.
TX must give 10KHz and RX+RX1 coils must give 10.5 KHz following this way to making these coils. Then when RX+RX1 are connected to RX pads they gives only same TX frequency (10KHz).
I've measured also with scope not only frequency meter.

Take a look at Mike's instructions, below.

Thanks again.

Best regards,
Max

Hi Max,

Gary is correct, but I think you have missed one subtle point here.
The coil construction article you attached says "I have measured the frequency obtained when I connect the RX coil on the detector oscillator."
In other words, Mike is saying that the RX coil was connected in place of the TX coil and then the frequency of the oscillator was measured. This is completely different to measuring the frequency across the RX coil with the TX coil connected. In this latter case it must be the same as the TX frequency, as there is no other frequency being transmitted. The only difference will be the amplitude of the signal at the RX coil.

[Geo]

Hi Max.
If you connect the scope to the input of IC5b you must see a phase shift when you turn the disc poten.... But connect the scope to -input of IC5b, not at out of ic7b. If you have not a phase shift , then look for the problem. Yes the IC7b is a phase shift (how else to work the discrimination ????)

[Max]

Some people of Europe has success with the Bandido (hope this is right because has finished the project). In the files there are the same coils, etc. In very detailled parts list you can find replacements, bla, bla, bla. This is the detector based on split PCB of this people (maybe you want all the info) PCBs are in Eagle program and all the documentation in .rar compressed files. You can compare with all has been done by you at this moment. Remember that the second TX coil is winding in reverse, so big TX and receiver are in the same orientation winding and secondary TX (bucking coil) is in reverse winding (or phase).

Hi Esteban,
thanks but I've already the Germany-group's stuff and the one from Paul-Panzer...seem nice works.
I'm doing with Gary's pcb design to avoid too layers...at now.
My coil is ok now (nulled under 1mV peak to peak).

Now it's good in all metal but disc doesn't work like expected:

I've found something that seems wrong in the disc section.
I mean at IC7/b pin 1 I get a sinusoidal wave that have a phase shift respect to
the main TX wave, and that's ok.
BUT that phase shift seems fixed , I mean when I turn the DISC pot (100K) I get
ONLY AMPLITUDE VARIATIONS , NOT PHASE VARIATIONS
::confused:: then there is a zero-crossing detector, the comparator IC5/b that so always
switches at the same time intervals (respect to main TX wave) independent from DISC pot
position (because zero-crossing appears always at the same time too).

What's wrong ? I would expect a regulable phase shift there and so different timings of
the gate signal at pin 7 of IC5 not a fixed pulse-train!
I've checked four times...everything is like in schematic and I'm very confused now.

I think that IC7/b is intended as phase-shifter, isn't it ???

Any idea ? Any suggestion welcome.

Best regards,
Max

[Esteban]

Some people of Europe has success with the Bandido (hope this is right because has finished the project). In the files there are the same coils, etc. In very detailled parts list you can find replacements, bla, bla, bla. This is the detector based on split PCB of this people (maybe you want all the info) PCBs are in Eagle program and all the documentation in .rar compressed files. You can compare with all has been done by you at this moment. Remember that the second TX coil is winding in reverse, so big TX and receiver are in the same orientation winding and secondary TX (bucking coil) is in reverse winding (or phase).

[Max]

YES, C15. Sorry

Your Recieve Frequency "MUST BE THE SAME as the Transmit Frequency".

These Coils make an Air Core Transformer.
The Frequency on the Recieve coil has Nothing to do with the Inductance or the Parallel Cap. It is the Same Frequency.

Probably your Frequency Counter is adding a Parallel Capacitance when Measuring the Transmit Signal, Resulting in a Frequency Change.
This Can't Occur on measuring at the Recieve Coil.

Take care....Gary

Hi Gary,
"The Frequency on the Recieve coil has Nothing to do with the Inductance or the Parallel Cap."

Yes, I know. But I mean using Mike's instructions for measuring resonance frequency of TX and RX coils when connected to TX oscillator.
TX must give 10KHz and RX+RX1 coils must give 10.5 KHz following this way to making these coils. Then when RX+RX1 are connected to RX pads they gives only same TX frequency (10KHz).
I've measured also with scope not only frequency meter.

Take a look at Mike's instructions, below.

Thanks again.

Best regards,
Max

[chemelec]

YES, C15. Sorry

Your Recieve Frequency "MUST BE THE SAME as the Transmit Frequency".

These Coils make an Air Core Transformer.
The Frequency on the Recieve coil has Nothing to do with the Inductance or the Parallel Cap. It is the Same Frequency.

Probably your Frequency Counter is adding a Parallel Capacitance when Measuring the Transmit Signal, Resulting in a Frequency Change.
This Can't Occur on measuring at the Recieve Coil.

Take care....Gary

Hi Gary,
thanks for the help.
Maybe you mean C15 (parallel to the RX coil) not C16. In Mike's instructions there is something about + or - 4.7nF to try on the 33nF cap (C15).
My actual frequencies are:
TX = 10.155 KHz
RX = 10.530 KHz

TX was exactly 10.002 Hz before wounding RX1 then increased. Don't know why - maybe due to force applied by RX1 coil wounded on it.

I can see on another german post that other frequencies could be:
TX 10.15 KHz
RX 11.13 KHz

Are these fine too ? Anyone knows ?
How much critic are frequency ratio and values for TX and RX ???

Best regards,
Max

[Max]

seems now happy !

Hi Max,

I have no experience with uMax or concentric coil making.
-> but anyway here are my thoughts:
*I would start testing the circuit with DD coil - you can achieve null easyier.
and you will have reference to that original coil you are making now.
If both coils will show bad results then should be problem on board.
Maybe you have already some appropiate scrap coils; hook them to the circuit:
1) null them in DD (OO) configuration.
2) with changing caps on Tx get appropiate frequency.
3) with changing caps on Rx search for highest amplitude.

*I found out that searching for highest amplitude on Rx coil works better than trying to get same resonant frequency (even if this should lead to same result)
Amplitude was measured on input pin of preamplifier (ic removed).
Rx coil driven by circuit.

Hi Leto and all,
I'd like to send many thanks to everyone for the useful help and for every hint.
Endly I've rewounded Rx and obtained a good null at less than 1mV peak to peak.
I found also a problem on PCB that was almost invisible without using a lens !
Was a little short due to a small solder metallic "ball" that was "glued" by flux
accross two pads...I've already washed out the excess of flux by a solvent...but
this little thing there was, as I say, almost invisible and impossible to find if
I wouldn't followed signall path using the scope, discovering the "little" mistake.
I have again 33078s as low noise op amps, not yet 33178s.
Much consumption BUT

Now, in ALL METAL wow ! it works fine...pretty good:

2 eur coin at 28cm ---> small output volume boost that is easy to ear in headphones

Like in other VLFs different alloys responds very different here...but:

- a small gold ring (1.5 grams) at 24cm that seems very good considering low tx power!

My coil is now very well nulled, and my actual "recipe" is:

104 turns for TX on 200mm id
155 turns for RX on 120mm id
25 and 3/4 turns for RX1 (0.25 of turn 26 shaped in the middle between tx and rx) 200mm id

(with my components is OK)

very good ! on TX osc I have (as in Mike's "recipe"):
- TX coil gives 10.022 KHz
- RX+Rx1 coils gives 10.506 KHz
- whole RX+RX1 nulled signal on 1Kohm load and 47nf cap is less than 1mVpp
Needs very good shielding too...I haven't at now on and capacitive is a huge problem.

All the stuff run in "motion" mode ...why the manual talk about non-motion ? where is it ?

Only disc doesn't work well till now...about 1/3 of all metal performances...no good!

Have to fix it, somehow...

Best regards,
Max

[Leto]

Frequency Vs Amplitude

Hi Max,

I have no experience with uMax or concentric coil making.
-> but anyway here are my thoughts:
*I would start testing the circuit with DD coil - you can achieve null easyier.
and you will have reference to that original coil you are making now.
If both coils will show bad results then should be problem on board.
Maybe you have already some appropiate scrap coils; hook them to the circuit:
1) null them in DD (OO) configuration.
2) with changing caps on Tx get appropiate frequency.
3) with changing caps on Rx search for highest amplitude.

*I found out that searching for highest amplitude on Rx coil works better than trying to get same resonant frequency (even if this should lead to same result)
Amplitude was measured on input pin of preamplifier (ic removed).
Rx coil driven by circuit.

[Max]

Hi all,
I get 80mV peak to peak at 20Kohm resistors (output pin 7 of IC7) after C17.
Seems good considering that gain of IC7/a is about 20 --> so I have only about 4 mV
peak to peak on Rx coil or less when I connect to the circuit.

When measured with 1Kohm lad it was 2mVpp.

But I have these results (80mVpp at pin7) only connecting a GND wire to one lead (in my
case the one that goes to pin6 of IC7). Tried different reversing for phase...
but disc doeasn't works in any case.
In all metal I get a 1 eur coin at about 10cm in air test. Again too poor detection.
I've read something about non motion in all metal, but I have always a motion
behaviour ::confused:: when hold a coin fixed to the coil...it stops sounding.
Using auto switch timing are different but even motion detection occours.

Don't know why performance is so small...I think I have to experiment with rx capacitor
C15 adding or removing some nF. But seems strange that I have so few cms detection.

My actual frequencies are now (using Tx OSCILLATOR):
TX coil ---> 10.040 KHz
RX+RX1 ---> 10.432 KHz

So I could try to remove some turns from Rx/Rx1 to get 10.50 Khz ...think that 1 turn on
Rx will be ok to add about 70 Hz. It's so critical ?

Best regards,
Max

[Max]

Phase Shifts are Normal.

In My pinion the RX Coil connects to R26 and R27 as Shown on the schematic. Not to Ground.

This works good using my coil on my Split board design.
(However I have only used my unit for bench tests, Nothing outside on the Ground)

One other thing I recommend doing while Nulling the coils, is to slightly adjust the value of C16 for best Peak detection of the signal.
If your coils or frequency are Not Perfect, (Which I Doubt they ever are) This will Increase the Sensitivity by obtaining a better resonance to the actual transmit frequency.

Take care...Gary

Hi Gary,
thanks for the help.
Maybe you mean C15 (parallel to the RX coil) not C16. In Mike's instructions there is something about + or - 4.7nF to try on the 33nF cap (C15).
My actual frequencies are:
TX = 10.155 KHz
RX = 10.530 KHz

TX was exactly 10.002 Hz before wounding RX1 then increased. Don't know why - maybe due to force applied by RX1 coil wounded on it.

I can see on another german post that other frequencies could be:
TX 10.15 KHz
RX 11.13 KHz

Are these fine too ? Anyone knows ?
How much critic are frequency ratio and values for TX and RX ???

Best regards,
Max

[chemelec]

Hi all,
I've made some modifications on RX1 coil obtaining a very good null:

- 2mV peak to peak at Rx-Rx1 leads when no metal present

I can see on the scope when no metals 2mVp-p, and when big metal plate present about
1Vp-p or little more, that seems ok.

So coil now seems good nulled. But circuit still doesn't works well...unstable...
sometimes I hear cracking sound.

Performance is sometimes better than previous RX1...but still too weak!

I connected first the two leads of RX-Rx1 to RX pads on
PCB ALONE (mean with no GND connection) ---> doesn't work

Then like in Mike's instructions I connected leads to RX but having one connected also
to GND --> doesn't work too like expected but here I can see at output IC7 a different
output...meaning referred to GND and then increasing a lot when a metal is present.

Question is : what's the correct wiring for RX leads to circuit ??? I think one must have
GND connection...near zero for op-amps, away from saturation voltages...anyone can confirm ???

I can see also phase shifts when metal is near...but disc seems doesn't work like
described in manual.

Any idea ? Any suggestion welcome.

Best regards,
Max

Phase Shifts are Normal.

In My pinion the RX Coil connects to R26 and R27 as Shown on the schematic. Not to Ground.

This works good using my coil on my Split board design.
(However I have only used my unit for bench tests, Nothing outside on the Ground)

One other thing I recommend doing while Nulling the coils, is to slightly adjust the value of C16 for best Peak detection of the signal.
If your coils or frequency are Not Perfect, (Which I Doubt they ever are) This will Increase the Sensitivity by obtaining a better resonance to the actual transmit frequency.

Take care...Gary

[Max]

Hi all,
I've made some modifications on RX1 coil obtaining a very good null:

- 2mV peak to peak at Rx-Rx1 leads when no metal present

I can see on the scope when no metals 2mVp-p, and when big metal plate present about
1Vp-p or little more, that seems ok.

So coil now seems good nulled. But circuit still doesn't works well...unstable...
sometimes I hear cracking sound.

Performance is sometimes better than previous RX1...but still too weak!

I connected first the two leads of RX-Rx1 to RX pads on
PCB ALONE (mean with no GND connection) ---> doesn't work

Then like in Mike's instructions I connected leads to RX but having one connected also
to GND --> doesn't work too like expected but here I can see at output IC7 a different
output...meaning referred to GND and then increasing a lot when a metal is present.

Question is : what's the correct wiring for RX leads to circuit ??? I think one must have
GND connection...near zero for op-amps, away from saturation voltages...anyone can confirm ???

I can see also phase shifts when metal is near...but disc seems doesn't work like
described in manual.

Any idea ? Any suggestion welcome.

Best regards,
Max

[Max]

Hi Max. Sorry but i am not at good condition ( i am very tired) to look for Bandido bad operation. But 2 Euro you must detect at 22 cm on air with 20cm coil. I saw it at a detectors test. I believe that at "all metal" mode you must detect the coin at 28....30cm .
Bye

Last edit.....
Minelab ( Sovereign XS, Sovereign XS-2)
All MetalPomiary dla $25DETEKTORCEWKAAIRGROUNDGarrett GTI 20009.5" 37 cm 25 cm Garrett Treasure Ace 300 8.5" 17 cm 13 cm Garrett GMH CXII 8.5" 24 cm 18 cm Garrett GTA X 1000 8.5" 25 cm 17 cm White's Spectrum XLT 9.5"Blue Max 30 cm 21 cm White's 6000 Di Pro 9.5"Blue Max 35 cm 22 cm White's Quantum XT9.5"Blue Max 26 cm 20 cm White's Classic SL III 9.5"Blue Max 24 cm 15 cm White's Classic SL II 8" Blue Max 19 cm 14 cm Fisher 1225 X5" 17 cm 12 cm Fisher 1235 X8" Spider 21 cm 16 cm Fisher 1265 X8" Spider25 cm 17 cm Fisher 1266 X10.5" Spider 29 cm 19 cm Fisher CZ5 8"Spider 32 cm 21 cm Fisher CZ 20 10.5" Spider 36 cm 24 cm Tesoro Lobo Super Traq 10" x 5" elipsa 28 cm 20 cm Tesoro Toltec 10.5" 26 cm 18 cm Tesoro Golden Sabre8" 21 cm 16 cm Tesoro Bandido8" 22 cm 14 cm Tesoro Silver Sabre10.5" 24 cm 17 cm Tesoro Sidewinter8" 20 cm 16 cm Tesoro Cutlass8" 19 cm 14 cm Tesoro Amigo7" 17 cm1 3 cm Discovery Baron Module base 8" 32 cm 23 cm Minelab Sovereign XS8" 24 cm 18 cm Minelab Sovereign XS7.5" 26 cm 19 cm Minelab Sovereign XS-2 10" 33 cm 22 cm

Hi Geo,
thank you for the useful infos.

Best regards,
Max