TGSL Tuning & troubleshooting

[dfbowers]

Hi Don:

I'm experimenting with coils, cables, and grounding also at the moment, even though I believe the "TGSL standard" wiring is no doubt optimal -- I'd like to understand better what the tradeoffs are and maybe why it works.

-------- USB wire choice ---------

I had a thought, but probably too late to mention it now, and probably not important at all, and it would mess up your coil compatibility --

The thought is that I noticed that many USB cables seem to make a twisted pair out of the green and white wires. I'm thinking maybe they would be the best choice for the RX wires on the theory that the twisted pair is more immune to outside noise, instead of the red and black wires. However, since one is grounded, that theory may be out the window.

--------- Goof??? -----------

I just noticed that on my PCB I interchanged the green and black USB wires compared to yours (dfbowers) -- so I need to go back and redo some tests!!!! Interestingly, it wasn't a disasterous mistake performance wise, but can't proceed until I get compatible with your PCB. I also have to get my head straight on just what the difference was topologically, these things get tricky.

Hold on! I think I just convinced myself that green and black can be interchanged at the PCB, same circuit. So tests OK I think!!!!!! But better to change on my PCB just for clarity. I'm going nuts...

---------- Null -------------

I have always been a believer that the null is not that critical to performance (but willing to be proved wrong anytime -- crude theories usually don't predict the subtleties of MD behavior). Your (dfbowers) MDs seem to confirm that belief. The main danger I see is if your null makes a negative voltage on the C15, C12 capacitors, then the JFets become forward biased at times. Also being off-null theoretically couples the coils a little and changes the properties of the RX resonant circuit - but not much until you get way off.

Looking forward to your tests, always learn something.

-SB

I'm half tempted to try the same cable that I used on my last project.. An extension cable from Whites. .It has individually shielded pairs, other than that, it's similar to a USB-2 cable.

[lunamay]

vladimir..another thing. does it realy matters the shape of "d"?(picture) because i seen many ideas

best regards and thanks for help

[simonbaker]

Simon,

I finished coil # 4 yesterday.. working on coil #5 today. I have enough epoxy and plastic to form one more.

I have been finding that the null point of my coils tends to change a little several hours into the epoxy curing process. They start out at about .2V at pin 7 and end up somewhere around .6 to 1V. Still, they seem to work very well but probably not optimal.

I'm still trying to figure out my wet grass problem and cannot cause noise any other way. I'm not giving up!

Don

Hi Don:

I'm experimenting with coils, cables, and grounding also at the moment, even though I believe the "TGSL standard" wiring is no doubt optimal -- I'd like to understand better what the tradeoffs are and maybe why it works.

-------- USB wire choice ---------

I had a thought, but probably too late to mention it now, and probably not important at all, and it would mess up your coil compatibility --

The thought is that I noticed that many USB cables seem to make a twisted pair out of the green and white wires. I'm thinking maybe they would be the best choice for the RX wires on the theory that the twisted pair is more immune to outside noise, instead of the red and black wires. However, since one is grounded, that theory may be out the window.

--------- Goof??? -----------

I just noticed that on my PCB I interchanged the green and black USB wires compared to yours (dfbowers) -- so I need to go back and redo some tests!!!! Interestingly, it wasn't a disasterous mistake performance wise, but can't proceed until I get compatible with your PCB. I also have to get my head straight on just what the difference was topologically, these things get tricky.

Hold on! I think I just convinced myself that green and black can be interchanged at the PCB, same circuit. So tests OK I think!!!!!! But better to change on my PCB just for clarity. I'm going nuts...

---------- Null -------------

I have always been a believer that the null is not that critical to performance (but willing to be proved wrong anytime -- crude theories usually don't predict the subtleties of MD behavior). Your (dfbowers) MDs seem to confirm that belief. The main danger I see is if your null makes a negative voltage on the C15, C12 capacitors, then the JFets become forward biased at times. Also being off-null theoretically couples the coils a little and changes the properties of the RX resonant circuit - but not much until you get way off.

Looking forward to your tests, always learn something.

-SB

[Vladimir]

Vladimir could you send a picture of your stencil on which you are winding your coils a specialy with geometric parameters (all radiuses etc) ??

best regards;]

....

[Vladimir]

Vladimir, thanks for the suggestion, the final adjustment, I knew I put something, but did not know it was, it soon proved, a greeting to all

I just want to help people to a nightmare called TGSL finish and make it usable
Regards

[lunamay]

Vladimir could you send a picture of your stencil on which you are winding your coils a specialy with geometric parameters (all radiuses etc) ??

best regards;]

[Vladimir]

I'm talking about the theory of how signals travel in cables.

Metal detectors, audio amplifiers, and low speed digital data all obey the same laws. Laws that Faraday, Hertz and the others established long before metal detectors were invented, and I don't need experience in metal detectors to know them.

I didn't mention anything about UHF. That is a different ball game. At those frequencies cables become transmission lines.

Of course the laws are invented earlier, but there is one small difference that separates the bad and good metal detectors, because you need experience. I do not think anything bad, just giving you advice.

[geminis74]

Vladimir, thanks for the suggestion, the final adjustment, I knew I put something, but did not know it was, it soon proved, a greeting to all

[pebe]

hmmmm, sound and UHF is not the same as metal detection, you attempt to get practical experience with at least one detector, then, we can advise theoretically.
In this way, without experience, you will create confusion in the current debate.
Regards

I'm talking about the theory of how signals travel in cables.

Metal detectors, audio amplifiers, and low speed digital data all obey the same laws. Laws that Faraday, Hertz and the others established long before metal detectors were invented, and I don't need experience in metal detectors to know them.

I didn't mention anything about UHF. That is a different ball game. At those frequencies cables become transmission lines.

[Vladimir]

Amendment # 312

Amendment # 312

[Vladimir]

No. But I've had plenty of experience with screening of audio signals, and 12kHz is only an audio signal.

hmmmm, sound and UHF is not the same as metal detection, you attempt to get practical experience with at least one detector, then, we can advise theoretically.
In this way, without experience, you will create confusion in the current debate.
Regards

[pebe]

Have you tried it already?

No. But I've had plenty of experience with screening of audio signals, and 12kHz is only an audio signal.

[dfbowers]

Hi Vladimir -- those voltages were at output of LF353 pin 7, not directly at the RX coil leads. Maybe still high, I don't know. It's just the way my coils are. I'll check dfbowers' coil that he sent me...

I just checked dfbowers coil, it is about 1V pp at same point. It seems a little high, but again it is amplified by LF353. I wonder if it has shifted a little during shipping or my handling. I tried deforming the coil head with pressure of my hands, but it is very solid and the null does not change at all -- nice solid coil head I must say!

Regards,

-SB

Simon,

I finished coil # 4 yesterday.. working on coil #5 today. I have enough epoxy and plastic to form one more.

I have been finding that the null point of my coils tends to change a little several hours into the epoxy curing process. They start out at about .2V at pin 7 and end up somewhere around .6 to 1V. Still, they seem to work very well but probably not optimal.

I'm still trying to figure out my wet grass problem and cannot cause noise any other way. I'm not giving up!

Don

[Vladimir]

Well, your 304 posting certainly looks like coax cable - I don't know of a flat braided cable. In any event, the wire that feeds current to the Tx coil is also part of the RX circuit and will cause a voltage drop in both circuits due to the resistance of the screening wire. The Rx current introduced in this way will be in phase with the Tx current so overall the circuit will act like the coils are overcoupled. Obviously, you take care of that when you null the coils.

I wouldn't mind betting that if you separated the screens along their length and only connected them at the PCB, you would need to null the coils again - but the screening would be just as effective.

Have you tried it already?

[simonbaker]

Hi Vladimir -- those voltages were at output of LF353 pin 7, not directly at the RX coil leads. Maybe still high, I don't know. It's just the way my coils are. I'll check dfbowers' coil that he sent me...

I just checked dfbowers coil, it is about 1V pp at same point. It seems a little high, but again it is amplified by LF353. I wonder if it has shifted a little during shipping or my handling. I tried deforming the coil head with pressure of my hands, but it is very solid and the null does not change at all -- nice solid coil head I must say!

Regards,

-SB

Additional note: when checking dfbowers coil, it was attached to my PCB for comparison. I don't know the null voltage with his PCB, would have to open the box to check.

-SB