TGSL Tuning & troubleshooting

[simonbaker]

Hi Simon ,nice work.
I respect it.
Did you grounded the box to the PCB.

Grt Nakky

Hi Nakky:

The exquisite MD was built by dfbowers, not me -- he knows his stuff. I was connecting my PCB to his coil for comparison.

I did not ground my PCB to the tin box.

I personally feel that PCB noise is not a dominant problem. When I disconnect the RX coil, there is not much noise. Most noise problems are picked up by the RX coil itself -- it is a huge antenna that will always be there. Only the Synchronous Detector and filtering can reduce that noise I think, maybe better balancing of coil/cable also.

Regards,

-SB

[simonbaker]

Hi Simon

You must clean the bottom of the plate with a solvent, for example, acetone. Then protect with plastic spray. The flux residue has a high resistance and influence the work of OP amplifiers.
What I noticed on the video clip is quite solid for a basic version TGSL, maybe it is time to do some modifications in terms of increase of sensitivity and stability.
Regards

Thanks, I'll try some acetone.

I want to be clear that the beautiful MD is not mine, it was made by dfbowers excellent craftsmanship.

I was just hooking my PCB (in the cookie tin) to his coil to test it.

I agree with modifications and I'm itching to try some as soon as I feel I have a solid original circuit.

I feel that Vladimir is on the right track with his mods. He has made the response a little slower. I feel the TGS was originally designed as a fast-recovery, shallow-depth MD with a smaller coil. For our bigger coils, I think a slower response should be better, especially for deeper targets. I already have some mods designed, similar to vladimir's, only adjusting different parts.

-SB

[WM6]

I don't really know what to do next. I'm thinking that my lack of depth (without the sun) may be due to the previously discussed threshold problem at the LM308 output region, and so implementing the dfbowers mod and trying different chips with different output offset voltages may open up more sensitivity and more depth.

-SB

Hi Simon, you still can do white PCB paint as Minelab does.

[Qiaozhi]

Hi Simon,

Now you can understand why Tesoro don't bother to do this. Can you imagine the problems in manufacturing? For stability and ease of production, it is much easier to stay with the 20cm to 25cm detection distance. Most of the issues you are experiencing can be put down to component tolerances, and maybe PCB layout and general wiring.

[nakky]

Hi Simon ,nice work.
I respect it.
Did you grounded the box to the PCB.

Grt Nakky

[Vladimir]

I'm going the extra mile...

Flux residue... don't know. What do you do if you have it?

-SB

Hi Simon

You must clean the bottom of the plate with a solvent, for example, acetone. Then protect with plastic spray. The flux residue has a high resistance and influence the work of OP amplifiers.
What I noticed on the video clip is quite solid for a basic version TGSL, maybe it is time to do some modifications in terms of increase of sensitivity and stability.
Regards

[simonbaker]

Holy cow Simon! You are putting quite the effort into trying to figure this whole thing out. The area that you are testing in looks quite desolate. I can't imagine that you would have to go so far out to get a decent air test... but I guess whatever it takes.

Anyway, a quick test would be to pull all your ICs if you have them socketed and swap with mine (if you have not done that already).

Any difference in frequency between your PCB and mine with same coil?

Do you know the exact value of C6?

Did you use Poly caps in some of the more critical areas other than filter and bypass caps?

Did you attempt to match R13 and R14 within 1% or so? (Maybe makes no difference.. but might ..)

Is your PCB free from flux residue?

I'm pretty anal about ESD.. I used to teach a class and would rather avoid headaches later on... So, I'm usually well grounded when playing with IC's and always install them last.

I also test every component before installing just to make sure something is not too far out of spec.

Other than that, things seemed to go together smoothly ..
You probably already thought about the items above, but wanted to throw some ideas out there.

Don

I'm going the extra mile... next I'm going to find some place closer, got caught in road construction on that last run, waiting for the escort flagman...

While I was out there I realized that I should have brought my "trimming capacitor" (had to forget something...) because my C6 is about .013 instead of .015, so that might have been a little tweak.

I definitely want to put in the "dfbowers mod" because that can overcome some LM308 sins, although it doesn't address "imbalanced" biases -- separate sensitivity pots would allow tuning that.

I usually test components too and try to get reasonable ones. I'm pretty sure there aren't totally dead parts anywhere just from looking at waveforms, but swapping parts is always worth a try. Classier IC sockets maybe a good idea next time.

Flux residue... don't know. What do you do if you have it?

Always appreciate the ideas, that's how we keep conquering...

-SB

[dfbowers]

Video link:

http://www.youtube.com/watch?v=Sw8mAVJQStk

-SB

Holy cow Simon! You are putting quite the effort into trying to figure this whole thing out. The area that you are testing in looks quite desolate. I can't imagine that you would have to go so far out to get a decent air test... but I guess whatever it takes.

Anyway, a quick test would be to pull all your ICs if you have them socketed and swap with mine (if you have not done that already).

Any difference in frequency between your PCB and mine with same coil?

Do you know the exact value of C6?

Did you use Poly caps in some of the more critical areas other than filter and bypass caps?

Did you attempt to match R13 and R14 within 1% or so? (Maybe makes no difference.. but might ..)

Is your PCB free from flux residue?

I'm pretty anal about ESD.. I used to teach a class and would rather avoid headaches later on... So, I'm usually well grounded when playing with IC's and always install them last.

I also test every component before installing just to make sure something is not too far out of spec.

Other than that, things seemed to go together smoothly ..
You probably already thought about the items above, but wanted to throw some ideas out there.

Don

[simonbaker]

Mountain Test 4 continued again

Video link:

http://www.youtube.com/watch?v=Sw8mAVJQStk

-SB

[simonbaker]

Mountain Test 4 continued

Additional photos, video links.

-SB

[simonbaker]

The Unbearable Being of Lightness

Re: 4th trip to mountains -- more questions than answers...

My last trip to the mountains proved dfbowers TGSL capable of 30 to 34 cm when get away from environmental noise. (In my workshop, dfbowers TGSL gets 20 to 25 cm depth.)

I had intended to test my PCB with dfbowers coil, however, I had a screw-up where I short circuited my battery supply.

So I took a 4th trip to mountains today. I carefully isolated my battery pack, and also took along long headphone wires for better audio recording.

First I replicated dfbowers TGSL at 30+ cm. I noticed that what I previously referred to as "mid" or "vertical" position on his sensitivity pot is actually quite toward end position, just a little more to go. "Max sensitivity" (usable) is a small turn further. After max position, I get some chatter.

Nexted I unhooked dfbowers coil from his box and hooked the coil to my PCB which I had sprawled in an open cookie tin.

I do not yet have the "dfbowers sensitivity mod" implemented yet, but I plan to. However, even without the mod, I noticed that if I turned my sensitivity pot up, at a certain point I got chatter then motorboating then almost steady beep. So I carefully dialed in what I thought was the max sensitivity for my test, on the edge of chatter.

First, this surprised me because dfbowers machine, which has the mod, only has a little chatter at full sensitivity. My machine, without the mod, was going crazy if I turn the sensitivity up not even full. Well, ok.

My first test seemed encouraging. My beep seemed a little shorter than dfbowers, and it seemed I had to wave the Euro a little faster, but I felt I was getting some detection at around 28 to 30 cm.

I kept testing and testing, playing with DISC and sensitivity. I found sensitivity was critical, if I backed off, quickly I was in the 20 to 25 cm range.

I was annoyed by the crazy beeping if I turned the sensitivity up too high, so I tried putting the cookie tin lid lightly over the box holding my PCB. Now there was no crazy beeping, it was quiet for the full range of the sensitivity pot. But the air depth was no longer good, maybe 15 to 20 cm! Giveth with one hand, taketh with the other...

I go back to lidless configuration. Now I'm having trouble getting 30 cm at all. I hook dfbowers PCB back to his coil. Now his MD is only getting 25 cm at best. I'm sweating in the sun, dust forming a grimy plaster on my skin...

I notice dfbowers box and coil, being black, are pretty darn hot in the sun. I decide to put them in the car and break for lunch.

After lunch (turkey & swiss on sourdough), I fire up dfbowers TGSL again. Good, back to 30 cm. I hook up my PCB to his coil and am able to get a sort of scatty 28 to 30 cm, but not as clean as dfbowers.

I'm not happy with the way my PCB goes unstable at high sensitivity so I decide to experiment with the cookie tin top again. I put it on the box and again, it becomes stable. I hate that because I don't really know the cause.

Then I try something out of the blue that doesn't make much sense. I try seeing how high I can lift the lid before it becomes unstable. Then the shocking truth comes out.

I can lift the lid as high as I want as long as the shadow of the lid falls on my PCB.

!!!!

It's not elecrical shielding that's occuring -- instead it's the effect of blocking the sunlight on my PCB!!!!!!!!!

So... when the sun hits my PCB, it is less stable, but capable of greater air depth.

I prove this by covering my PCB with a sheet of paper. The sensitivity pot no longer causes the instability, but again the air depth reduces to around 20 cm.

I don't really know what to do next. I'm thinking that my lack of depth (without the sun) may be due to the previously discussed threshold problem at the LM308 output region, and so implementing the dfbowers mod and trying different chips with different output offset voltages may open up more sensitivity and more depth.

The problem is my workarea is too "noisy" for good air tests. I hope to use dfbowers MD to find a more nearby location than the mountains where I can get lower noise for testing.

Some sample photos and links to video for fun.

-SB

[dfbowers]

Dear Friends ,

- Can TGSL detect only gold and reject other metals ?

Man.. I donno. Gold can show up on about half the VDI scale on some detectors.. depending on how big it is. Dig everything!!

[simonbaker]

Dear Friends ,

- Can TGSL detect only gold and reject other metals ?

The TGSL ("lite" version of TGS) does not have a "notch" circuit for that. I think the original TGS circuit does. Check beginning of "Tesoro Golden Sabre" thread.

-SB

[raff33]

Cool! Thanks Simon.. Now you might not think that I'm nuts anymore. . Divide and conquor!! Now we need to come up with a noise reduction scheme.. Isn't that strange that detection range just sucks sometimes for no apparent reason? We all need to build Faraday cages around our workshops!! I was in a lab once in Harrisburg that had one.. It looked like a chicken coop.

Don

Dear Friends ,

- Can TGSL detect only gold and reject other metals ?

[simonbaker]

mountain test video

Here is some sub-standard video of the test of dfbowers TGSL in my trip to mountains. Very hasty because noise from cars intruding. I also did poor job of centering target under the search head.

Depth not perfectly show because of camera angle, but I think its clear that with sensitivity control in vertical (Mid) position (starting video), 30 cm is just achieved. With max sensitivity, just a small amount past mid position, detection is clearly beyond 30 cm.

Ruler is 12 inch, with 30 cm just before the end. Target is 1 Euro.

http://www.youtube.com/watch?v=PH6kjdP18Fc

-SB