TGSL Experiments

[magma_kostic]

Ivice (IVCONIC)...Potrazi drugog psihijatra , za tebe nemam vremena.
A o ostalom se moze i orbit izjasniti.

Da to sam ja, kako si provalio!? Svaka ti cast! Ha,ha,ha! A od kad to kao obican, polupismeni "htz"-ac sa nizom strucnom spremom, odjednom ti postade ekspert za elektroniku i metal detektore? Neki ljudi koji su te upoznali licno kazu da ne znas da razlikujes otpornik od kondenzatora, a evo te ovde prodajes znanje svima? Svaka cast majstore! Vrlo brzo si napredovao!

[Qiaozhi]

Ivice (IVCONIC)...Potrazi drugog psihijatra , za tebe nemam vremena.
A o ostalom se moze i orbit izjasniti.

Da to sam ja, kako si provalio!? Svaka ti cast! Ha,ha,ha! A od kad to kao obican, polupismeni "htz"-ac sa nizom strucnom spremom, odjednom ti postade ekspert za elektroniku i metal detektore? Neki ljudi koji su te upoznali licno kazu da ne znas da razlikujes otpornik od kondenzatora, a evo te ovde prodajes znanje svima? Svaka cast majstore! Vrlo brzo si napredovao!

OK guys, calm down.
I know you are insulting each other in Serbian. Please do it by PM, and not in the open forum.
Let's all try to be friends here.

[dfbowers]

Not perfect, but I made a crude shield with cookie tin. I did not see enough change to feel it was the main problem.

So I switched my "adjustable" coil for a glued coil that I made a year or two ago.

Here's where I start going nuts. It seemed there was still a low freq oscillation of some kind (at output of LM308s), but a little higher freq than before, and much less sensitive to vibration (like thumping the floor).

So I went to bed to continue next day. Today, the signal looks more random, I'm not seeing the clear oscillation, and not getting regular beeps, just more random ones.

Could there be different electrical interference (EMI) today? Maybe different temperature?

I will re-hookup other coil (adjustable) and see if the beeping comes back.

Then I'll continue testing with coils.

-SB

Simon,

In regards to your low frequency oscillations, I have dealt with that problem early on and it drove me nuts for a while. I definitely narrowed mine down to some sort of electromagnetic feedback between the speaker and the coil. The arrangement of the coil in relation to the speaker in particular. Try a very small speaker, like the ones that you get in greeting cards. Or.. just to isolate, replace your speaker with a 470k resistor in series with an LED. If your problem goes away then you know what it was.
Divide and conquer!

Also, Have you tried to etch your own PCB yet?

Don

[magma_kostic]

OK guys, calm down.
I know you are insulting each other in Serbian. Please do it by PM, and not in the open forum.
Let's all try to be friends here.

No insults in my posts, trust me. Just plain truth. Vladimir's posts are not nice but also no insults in those. Vladimir is Magma. And who knows who else. Many nicks, many profiles here. First he tried to act as Vladimir from Montana, Bulgaria. Now he is Vladimir from Serbia? Also he is lying here, all the time. As admin you should take care about those things. Man can not introduce himself as another existing man, as known Vladimir from Montana. That original Vladimir is educated man, gentleman, expert. This "Vladimir" here is illiterate lying idiot without elementar knowledge in electronics and metal detectors. This Vladimir here is misusing Orbit' poor English ant trying to take advance of his work, showing here it as his own work. I know Orbit personally and i spoke with him many times and especially recently about this. So i apeal to all admins here to take some actions and exclude this magma/vladimir idiot from this forum.
S.V.

[simonbaker]

Simon,

In regards to your low frequency oscillations, I have dealt with that problem early on and it drove me nuts for a while. I definitely narrowed mine down to some sort of electromagnetic feedback between the speaker and the coil. The arrangement of the coil in relation to the speaker in particular. Try a very small speaker, like the ones that you get in greeting cards. Or.. just to isolate, replace your speaker with a 470k resistor in series with an LED. If your problem goes away then you know what it was.
Divide and conquer!

Also, Have you tried to etch your own PCB yet?

Don

After seeing your post on speaker orientation, I played with that to see. I'm using a tiny button speaker and have not yet found an orientation or distance that clearly solves the problem. Also, I believe I see clearly the oscillations on the scope (very low freq sweep as the oscillations are about 10 hz), and when I disconnect the speaker, no noticeable change on the scope. That doesn't necessarily mean anything, but so far it seems not to be the speaker. The frequency of the beeps can be quite variable, so that does suggest some kind of orientation/proximity effect for sure. Or possibly some part is out of spec and thermally active.

I don't totally trust my anecdotal observations because I may be changing other factors at the same time. I have to redo tests a lot.

At the moment I'm also using a very high Q Rx coil (7 ohms). Maybe part of the problem, I'm going to check various ideas, like adding resistance. Also when you have a high Q Rx coil, the capacitor in parallel is more critical -- a couple of uF off and the gain can go way up, which might also encourage the beeping.

The other possibility is vibrational modes in the building. I think using a glued coil has helped but not entirely, so I'm keeping an open mind. I plan to drag the stuff, including my oscope, out into a park or campground to really try to get rid of EMI and vibrations, but not so easy -- I've found many buried cables in parks before.

No, I have not etched a PCB yet, I'm dragging my feet. But I'll have to make one or get one made before long to really be able to evaluate these designs. Part of my quirky approach is to try to make a MD using as primitive techniques as possible, as an alternative for putzers who just have a soldering iron and some parts from Radio Shack. Just something to contribute, maybe lame. I admire the professional jobs you and others have done, really nice stuff out of a home workshop.

-SB

[simonbaker]

Simon,

In regards to your low frequency oscillations, I have dealt with that problem early on and it drove me nuts for a while. I definitely narrowed mine down to some sort of electromagnetic feedback between the speaker and the coil. The arrangement of the coil in relation to the speaker in particular. Try a very small speaker, like the ones that you get in greeting cards. Or.. just to isolate, replace your speaker with a 470k resistor in series with an LED. If your problem goes away then you know what it was.
Divide and conquer!

Also, Have you tried to etch your own PCB yet?

Don

Just to follow up more on oscillations -

Today I thought they were gone. Detection was fairly good (probably around 20 - 25 cm for US quarter).

Then I removed my oscope probe from the TX oscillator - oscillations kicked in.

I put a pot in parallel with feeback resistor R15 of the LF353 as a crude gain control and turned the gain down.

I still saw the oscillations but when they became smaller, they did not trigger the speaker. But now I think maybe they are there at a low level at least all the time.

- I tried playing with orientation of search coil. It seems if I turn it on edge (perpendicular to ground) and align the plane to point to the PCB, the oscillations drop way down. Can't be sure I'm not just knocking the null off kilter, but I think it confirms that the search coil is feeding back to something, if not the speaker.

Maybe it's all in the packaging at this point. The tin box didn't seem to help shield the PCB, but maybe the test was poorly done. Or maybe the cable to the coil is reacting to the coil.

-SB

[pebe]

Hi Simon,
Just another couple of thoughts. If the search coil is picking up the oscillations they must be going through the LF353 preamp. That has got a gain of over 40 at 10Hz and you could reduce that considerably by inserting a cap in series with the feed from the coil to pin6. A 10nF fitted between C6 and R13 should so the trick.

The other possibility is microphony in the search coil. You say the receive coil has a high Q, so the effect would be worse under those conditions. Can you get oscillations by gently tapping the search coil?

[lunamay]

hi
i was thinking..are all pots lineal or logarytmic??
maybe silly question:P

best regards
lunamay

[porkluvr]

If you are using a feedback FET, using an incorrect FET could cause trouble. The amplitude feedback network can easily become unstable.

The coil current amplitude variations are hard to see in my LTspice screnshots, but the FET gate voltage and the spectrum displays show
possible noise in grand fashion. The only differences are that one circuit uses BF245A and the other uses BF245C.

You would see a similar difference in comparing between 2N4393 and 2N43931.

Differences in coil "Q" and some other things will affect how likely you are to ever encounter this situation and I can't hope to cover every aspect.
I'm just trying to point out a possible pitfall for the builder.

edit: I pickled this post off before I was ready so I had to go back and edit the photos. And, my simulation "zero time" is not actually at turn on but
would occur after the circuit has been running for tens of milliseconds.

[Vladimir]

Today I tried the coil from MAGNUM MD (2x61 turns, DD coil) and adapted capacitors in RX and TX, for the frequency 14.5 kHz. This combination is completely eliminated noise.

[simonbaker]

hi
i was thinking..are all pots lineal or logarytmic??
maybe silly question:P

best regards
lunamay

Both. Some pots are linear, some are logarithmic. You need to look at the specs when you buy a pot. If the pot says "audio taper", I believe that means logarithmic.

I think linear is best for the TGSL pots, especially disc control. Log would work, but be more difficult to use.

-SB

[simonbaker]

Hi Simon,
Just another couple of thoughts. If the search coil is picking up the oscillations they must be going through the LF353 preamp. That has got a gain of over 40 at 10Hz and you could reduce that considerably by inserting a cap in series with the feed from the coil to pin6. A 10nF fitted between C6 and R13 should so the trick.

The other possibility is microphony in the search coil. You say the receive coil has a high Q, so the effect would be worse under those conditions. Can you get oscillations by gently tapping the search coil?

Very good ideas, I'm trying to isolate what it is. There is definitely some microphony because it does react a lot to vibrations and touching, and vibrations cause oscillations of similar frequency. I suspected house/bench vibrations at first, and that may be the cause or part of the cause. I tried holding the coil in my hands to see if that changed the oscillation frequency. Difficult to assess because my hands add some noise, but it seemed that there were still underlying oscillations of same frequency.

So far what seems to reduce the oscillations is to turn the coil on edge (perpendicular to ground) with the plane pointed in a certain direction. I first thought it was relative to the location of the PCB, but I tried moving PCB to other direction, and it seemed that the PCB location didn't matter. That is weird, and perhaps indicates the oscillations are picked up from outside source or other interaction with the coil.

I tried turning the PCB to different orientations to see if that had same effect as turning the coil to different orientations, but it did not seem to have same effect on oscillations. So maybe doesn't involve coil-pcb interaction.

But don't trust my coil-on-edge experiment yet because I'm holding with my hand which is not steady. I need to build a support to really test. And also the coil on edge is probably deformed a little by its weight, which could affect the null -- the oscillations may disappear but also the sensitivity, which is no solution. But I find that when I put the coil on edge, only one direction reduces the oscillation, so I think there is some clue there.

Your point about the gain of the LF353 at 10 Hz is interesting. The question is: is the 10 Hz oscillation loop a 10 Hz analog signal, or is it a 14.5 kHz signal being phase modulated at 10 Hz somehow through a feedback loop?

The speaker would be a great suspect, but I disconnected the speaker and still see the oscillations.

Sometimes the oscillations are clearer and more steady than other times. I noticed that my scope does not trigger consistently on the 10 Hz oscillations for some reason, I don't know why. I think maybe it triggers on higher frequency noise riding on the 10 Hz signal.

I will try removing the LM308 chips to see the effect on the oscillations and perhaps where the feedback path originates from, if it is feedback.

Lots of tests to do. I suspect it is possible to get rid of the beeps by boxing the detector and just jiggling wires and such, but I'd like to understand it so I'll keep experimenting.

-SB

[simonbaker]

If you are using a feedback FET, using an incorrect FET could cause trouble. The amplitude feedback network can easily become unstable.

The coil current amplitude variations are hard to see in my LTspice screnshots, but the FET gate voltage and the spectrum displays show
possible noise in grand fashion. The only differences are that one circuit uses BF245A and the other uses BF245C.

You would see a similar difference in comparing between 2N4393 and 2N43931.

Differences in coil "Q" and some other things will affect how likely you are to ever encounter this situation and I can't hope to cover every aspect.
I'm just trying to point out a possible pitfall for the builder.

edit: I pickled this post off before I was ready so I had to go back and edit the photos. And, my simulation "zero time" is not actually at turn on but
would occur after the circuit has been running for tens of milliseconds.

That's interesting. I agree the feedback JFet could cause low frequency phase oscillations. I know about something called "limit cycles" which could easily happen like that. I don't actually use the JFet, I'm following a circuit almost identical to TGSL Final. But I think maybe the same thing could happen anyway because the Transistor saturation is the nonlinear element that stabilizes the oscillator (I believe), and a similar effect could happen.

However, the Synchronous Detector should ignore phase noise from the oscillator, especially at low frequencies. But maybe the high gain picks it up even if miniscule.

I have wondered if the high-Q RX coil could be almost resonating with noise at it's resonant frequency, and causing "beat frequencies" that come through some how. I don't think so given the numbers.

Since I think I'm seeing coil orientation involved in the oscillations, I have to keep checking that interaction. It is certainly possible that there is a mechanical oscillation being driven by the coil in some kind of a fashion. I'm a little suspicious of the cable.

But actually, the oscillation is very susceptible to an oscilloscope probe being placed on the TX oscillator -- sometimes the probe causes the oscillations to start, sometimes it reduces them dramatically. There is a clue there somewhere.

-SB

[simonbaker]

It Arrived!

Coils from dfbowers arrived. Many thanks Don!!!!

One Euro coin is on it's way from Stefano - many thanks Stefano!!! Dennis the Menace also offered to send Euro - great thanks there too. Appreciate the community spirit very much - it's fun and it will help a lot.

-------------------

dfbowers craftsmanship is very high - my workbench is in awe. Coils arrived beautifully packaged.

I was amazed how light the package was.

His technique for winding and gluing coils is quite unique as described elsewhere, resulting in square-shaped cross-section. The coils are very light and thin (not shielded), and nicely stiff and hold their shape well.

Coil diameter is about 26.5 cm.

From my measurements, wire thickness is 10/1000 inch (.25 mm, 30 AWG) with enamel insulation.

--------------------

Along with the coils is the plastic coil shell he made, I presume using his home-built plastic vacuum forming machine. Very, very light and attractive, with a nice molded mounting bracket for the pole. Maybe he can describe how he fixes the coils in the shell.

--------------------

It seems he has implemented the designs of Ivconic, Max, and others excellently and has found their designs to work as advertised. He says his tests indicate these coils give air depth around 29 cm with his TGSL PCB. I will now work to achieve that somehow with mine.

------------------

I did something dumb, and removed coils from the shell to photograph them, and forgot to clearly note the orientation he set up for me. However, I think I put them on the floor in same orientation, and hopefully I can restore that orientation from my photos and by nulling and testing.

-------------------

It will take some time to see how the coils work with my PCB and to use them to troubleshoot and tune some of the issues I have. My main issues are:

1. depth
2. beeping (motorboating)

For now I am just taking some photos to put here and some preliminary measurements.

-----------------

Dfbowers kindly attached a USB cable to the coil (as he normally uses). I cut the USB connector off the other end as he suggested and prepared the wires and took photos (see attached).

The USB cable has a really robust outer shield consisting of three parts: braided silver outer shield, copper ground wire, inner foil shield. The four signal conductors inside do not have separate shields.

(I had previously tried a cheap S-Video cable from China and was disappointed that there was no shield at all, only a ground wire and signal wires. So much for trying to save money.)

-------------------

I took some preliminary measurements of the coil-cable combination. This is what my meter measured:

coil1: (black wire -to-shield) 25.7 ohms, 6.82 mH.

coil2: (white wire-to-shield) 24.5 ohms, 6.35 mH.

red, green, and shield wires are connected together at coil head.

So it seems coil2 is the TX coil and coil1 is the RX coil.

--------------------
QUESTION

This is a question for coil experts: dfbowers grounds three wires together at the coil head - one of the TX coil, the cable shield, and one of the RX coil leads.

My question is: is it normal to ground the TX lead to the shield at the coil head, or only at the PCB, or should it matter? At least we know it works well for dfbowers!

--------------------

I look forward to playing with these coils in the weeks ahead. Anything I learn I'll be sure to post here, maybe something will be useful for others.

Thanks again Don!

-SB
　

　

[dfbowers]

TGSL airtest, 1e @ 30cm

Simon,

I uploaded a short video. I tested the coils that I sent you and performance is about the same as the posted video.

About the grounding.. I normally ground the shields to the point where the three wires are joined, just like Ivconic has documented in his TGSL coil making PDF.


Also, interesting inductance measurements you have taken. It tells me that maybe my method is a little off.. but I did measure the coils before I attached the cable, so maybe that is a variable. I will see if it makes a difference next pair that I make.
Don

http://www.youtube.com/watch?v=fwU0iweAimA