Fun with nulling

[morganton]

I'm in the process of trying to build a coil for my TGSL. I wound two coils that ended up with the following specs after binding, superglue and electrical tape:

TX Coil:

99 turns of AWG 30
5.80 mH
22.4 ohms
14.85 Khz when connected to oscillator

RX Coil:

104 turns of AWG 30
6.40 mH
23.9 ohms
14.14 kHz when connected to oscillator

These are off spec with dfbowers guide, but I was hoping they still might work.

I also have the correct voltages for LM358 pin 4 (-6.1V) and pin 8 (7.86 volts)

I arranged to coils so as to get as close to zero volts on pin 5 of U103 as I could:



But when I did that, the phases of J1-1 (TX) and U101 Pin 7 were separated by a lot more than the 20 degrees lag recommended by the guide:



So I probably know the answer to this question, but I'll ask it any way. Are my coils enough off spec to cause the separation between phases shown in the scope screen shot?

[morganton]

Oh, man I just read an old post that I may have my RX leads reversed. I'll try that tomorrow. Gonna feel pretty silly if it's that . .

[bernte_one]

it is often so, you dont see the forest because there are too much trees

two possibilities:
just flip one side of the coil or Change one Connection
good luck

[morganton]

After reversing the leads on the RX coil (duh), I got this phase relationship between TX and RX.

This is much better. RX lags TX by about 6us or about 32 degrees.

On the bench, the coil rejects a ferrite rod slightly before the mid position of the ground balance pot.

It detects a U.S. quarter at about 23cm.

Not terrible, but I think I can do better by winding some more coils and trying to get closer to the specs.

[morganton]

I wound some more coils:

TX:
6.09mH

RX:
6.59mH

I was able to overlap the coils to get the proper phase relationship with RX lagging TX by 20 degrees (4us):

(Yellow is TX, Green is RX)


And the "null phase" (U101 pin 7 and TR5 gate) seems correct:


I am able to reject a ferrite rod with the ground balance and the discriminating circuit works properly.

Detection of a U.S. quarter is only about 20cm on the bench, but I'm thinking that I may have a lot of EMI in my house that is affecting the detection range.

The puzzling thing is that adjusting the sensitivity pot (10K) doesn't seem to make any difference. The voltage on pins 2 and 6 of the LM393 correctly varies between about 0 and .035 volts as the pot is adjusted. I also tried the other LM393 on the board, but it didn't help the problem.

Is there anything else to check regarding proper operation of the sensitivity circuit?

[morganton]

Video of air test

I wound some more coils with about the same inductances and hot glued them into a housing.

I'm still not getting any change in sensitivity from the sens pot, but it detects fairly well.

Here is a video showing an air test:

https://youtu.be/7weAtXZnztI

[ashefr10]

hi, my coils are too poorly performing no more than 22 cm for a bronze coin is not terrible and the same sensitivity knob thing it is not working. I do not know, or the problem may be.

[morganton]

hi, my coils are too poorly performing no more than 22 cm for a bronze coin is not terrible and the same sensitivity knob thing it is not working. I do not know, or the problem may be.

ashefr10, the kit I received from silverdog had TL071's instead of LM308's for U104 and U105. I just received some LM308's and I am going to try them, although the discussion on the forum seems to say it doesn't make any difference in the sensitivity setting.

There is also the mod of changing the sensitivity pot value to 100K and connecting one side of the pot to -5V instead of ground. I will probably try that too.

Otherwise, I have no idea why the sensitivity pot doesn't work. Everything else on the detector works the way it should, although I wish I had more depth.

Dan

[ashefr10]

it is certain that the tgsl he sorely lacking in depth, I bought NJM2068D I will put in place of the LF353 we'll see, it is not easy to break through this detector.

[morganton]

Null change after potting

I potted the coils using MG 832B. I did one pour, rather than multiple pours.

Below are screenshots of the null phase relationship when the coils were only hot-glued in the housing and after the coils were potted.

I have also attached a photo of the bottom of the coil after potting.

The null voltage increased to about 1.15 volts from the previous 400mV. However, the phase relationship remained the same.

The depth of the detector may have actually increased slightly.

So, it looks like the null phase relationship did not change enough after potting to affect the overall operation of the detector.