TGSL Experiments

[simonbaker]

Hello,
I have made more coils and a lot of tests.
Not much time so will try to write brief.Have made 4 DD27 and 3 DD19.
Still dont now why but all 27 have poor sens when nulled./around 27cm and usualy last few sm-s are with cracked sound/I had some doubts even about the PCB.But not anymore.
The interesting part is that one of my 19DD have greater sensitivity than all DD27 I made so far.Around 30cm for 1E coin.Disc is ok but cant GBalance device with this exact sensible coil.While all 27 have reduced sens and can be GBalaced very well.What could be the reason?This exact DD19 have great audio.Much more louder and in all area of detection!
The thing that I notice is that when my windings are on former and still not constrinkted are around 4,8 mH and when finished they show 5,98.And Ivconic say that he measur around 5,7mH when not constrikted,so later have 6,0mH.
So I am winding too woosed maybe?

This is very interesting and useful information.

It would be good to see photos of all your coils, along with:

1. TX, RX Dimensions
2. TX, RX Inductance
3. TX, RX Resistance
4. TX, RX Wire gauge/thickness
5. TX, RX Shield arrangements.

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

Ok -- 27 cm is not bad of course, but 30 cm is nice, especially if audio is "fat".

Why one 19 cm coil good??? But not GB??? This is important mystery! You have some useful experience now. Maybe we can get some insight.

Q: what do you mean "cannot ground balance"? How do you test ground balance?

Q: only one 19 cm coil good? What about other two?

I notice with my TGSL that target signal in GB channel is very small compared to Disc channel. Perhaps with some coils GB channel signal is bigger, more like Disc channel, and audio is better, because both signals needed to make audio.

Possible explanation: If target phase becomes shifted so it overlaps both the disc and GB sync pulse phase more evenly, then maybe makes more sensitive detector -- but destroys GB control.

This could happen if the RX coil resonant frequency is not normal distance from oscillator freq, or the the RX coil resistance is changed causing the phase to shift.

But I don't know. I have tried to vary those parameters, but no good conclusions yet.

-SB

[simonbaker]

Hello,
The thing that I notice is that when my windings are on former and still not constrinkted are around 4,8 mH and when finished they show 5,98.And Ivconic say that he measur around 5,7mH when not constrikted,so later have 6,0mH.
So I am winding too woosed maybe?

In my opinion, doesn't matter how loose you wind as long as you get final inductance -- unless neatness matters, but no evidence yet about neatness or ordering of windings. How wires are ordered may affect capacitance of coil -- but does it matter much??? Nobody talks about that anyway.

-SB

[ivconic]

eh...this maybe helpfull to you guys, somehow..
"Fat" audio is most proper term i heard latelly! Exactly!
So...
just few minutes ago i played a bit with my latest CL3 handmade. Guess what? I got horrible and not "fat" audio at it when pcb was straight on table. But when pcb lifted on match box - audio perfect and very "fat"!
Conclusion - hums and interferences on pcb and arround it!
Bad S/N ratio.
UHaze you can perform simillar check. Try to move your pcb away or somehow, lift it over table surface..put something bellow, just to distingusih is it because of bad S/N or something else. Also put long wires to speaker and check, than short wire and check...
One of the reasons why my finished TGSL's works very good is because i use right method to pack pcb and to mount presets on front panel.
Usually i cut and bend Al plate and use it as front panel (plastic layer with engraved texts later added over it). That plate is, logically, connected to pcb gnd. Rear panel....with coil connector - also bended Al plate (smaller one) directly screwed to pcb (fat gnd part).
All the wires from pcb to presets and switches are shortest possible and muffled togather, closest possible.
So...pcb along with plates and mounted/wired presets, switches, are ready than, to use to balance exact coil for that detector.
So..i presume that is one of the main reasons why it is working very good and why i usually dont have problems with coils. Shortly - S/N ratio!
At older Tesoros the same situation. Wires and pcb looks like mess...but all togather closed to good grounded metalic enclosure! Voila!
Newer Tesoros are made on high quality and very small pcb's, usually smt with very correct S/N ratio...
So i presume that most of your problems are actually in just what i explained.
Same thing with minute ago CL3...acting wild and erratic, but when packed in suitable and good grounded enclosure, acts like stone!

Hopefully this will help you a bit..
Regards!

[Ultimate_Haze]

Sorry for dely.
I will put the device in some box and will test again.Funny but I have lift the PCB before some days on a box.Not a big differance))
Thanks for advice.I will try chopper changes also.No matter how I balance this 19DDcoil same results.Just cant be GBalanced.Pitty its so sensitive.

[Ultimate_Haze]

Q: what do you mean "cannot ground balance"? How do you test ground balance?

SB- I use standart procedure that Ivconic have described so many times.I null coil/the best I can/and then use small ferite rod/around 3sm long/and buy rotaiting GB trimer try to reject it.But no matter how I turn it still have loud detection.
Yes audio is fat on this one DD19.Other two are average,noting to do with.Dont now why I wound them almost the same.With e.g. 1-2 turns diferance.Strange is that at the end they have same measurments 6 and 6,5.


This could happen if the RX coil resonant frequency is not normal distance from oscillator freq, or the the RX coil resistance is changed causing the phase to shift.

Have try to change RX freq buy adding 1n to this 15n.Nothing changed.Will try other combinations there to test your teory.

Sorry can make pictures now.I have some event and pack in boxes and clear all for now.Only this 19DD and PCB are here.Will unpack next week and post pictures with measurments.

[ivconic]

Q: what do you mean "cannot ground balance"? How do you test ground balance?

SB- I use standart procedure that Ivconic have described so many times.I null coil/the best I can/and then use small ferite rod/around 3sm long/and buy rotaiting GB trimer try to reject it.But no matter how I turn it still have loud detection.
.....

Well... are you sure wires from coil are not reversed? Try to reverse only TX or only RX wires.
If not....than your coil is not balanced well again...
If not....than some component in Disc/Geb circuitry (LM393) is missmatched much ...or bad...

Regards!

[simonbaker]

Q: what do you mean "cannot ground balance"? How do you test ground balance?

SB- I use standart procedure that Ivconic have described so many times.I null coil/the best I can/and then use small ferite rod/around 3sm long/and buy rotaiting GB trimer try to reject it.But no matter how I turn it still have loud detection.
Yes audio is fat on this one DD19.Other two are average,noting to do with.Dont now why I wound them almost the same.With e.g. 1-2 turns diferance.Strange is that at the end they have same measurments 6 and 6,5.


This could happen if the RX coil resonant frequency is not normal distance from oscillator freq, or the the RX coil resistance is changed causing the phase to shift.

Have try to change RX freq buy adding 1n to this 15n.Nothing changed.Will try other combinations there to test your teory.

Sorry can make pictures now.I have some event and pack in boxes and clear all for now.Only this 19DD and PCB are here.Will unpack next week and post pictures with measurments.

Phase problems more likely caused when resonant frequency is too close, so adding capacitor there actually makes it even closer.

You could try adding capacitor to oscillator (parallel to .022u cap) to lower oscillator freq instead.

But perhaps there is some mystery about coils we are missing completely. By studying your 19 cm coils, maybe you can find the subtle difference.

Let's call the coil that detects at 30cm the "deep" coil, the other coils the "shallow" coil.

Some things to test:

1. With dual trace oscilloscope, compare the null signal (LF353 pin 7) against the oscillator signal for deep and shallow coils. What is phase and amplitude?

2. Measure the oscillator freq for deep and shallow coils.

3. Measure coil inductance, resistance for deep and shallow coils.

4. Compare the discrimination for the deep and shallow coils. Use foil or pulltab, etc. Measure where in disc knob setting the rejection occurs. This may depend on distance from coil, so maybe pick point 5 cm closer than max depth.

5. Do the "shallow" coils reject ferrite? If you move ferrite farther from "deep" coil, can you find a setting that rejects it best?


6. This is a very subjective test, but very informative:

- Put oscilloscope sweep on about .2 s/div speed. Adjust for best viewing.

- Put one probe on disc LM308 output pin 6.
- Put other probe on GB LM308 output pin 6.

Watch the waveforms as you move the target over the coil. Try for both deep and shallow coils. Do you see a difference?

It would be excellent if we could understand this -- why two coils, so similar, behave so differently. I think this is an important area to understand.

Anyway, your "deep" coil should be good for hunting gold nuggets!

-SB

[Ultimate_Haze]

Ok,
Simonbaker.
I will do this as soon as unpack all and make setups again.

Ivconic ,
They were connected just the way you show.I try to sweep RX leads nothing change about GB.Few more cm's depth this way.Next will try to sweep only TX.

[simonbaker]

coil idea that almost worked

I tried idea of winding coil on a form that would become part of search head, to save time of binding, etc. First try was unshielded, just for test.

It almost worked. I used thick wire and pressure of wire was too much for coil form at corners, as shown in pictures.

Using lighter wire, winding less tightly, and using more rigid materials for form might make it successful.

Idea is to make two such forms, one for TX, one for RX coil. Then position them in DD arrangement, using support material under one coil to keep it level because it is positioned over the other.

After winding a coil, glue such as super glue can be squirted into coil to make it permanent. For shielded coil, shield is first inserted into notch around form, then wrapped over wires after winding and glue.

Just an idea.

-SB

[ivconic]

Your idea is good. Just pay attention when shield is about to be applied.
I was thinking about some similar arrangement in the past, yet i had a lot of problems with shield.

Ok..one more thing that can be helpful here;
few days ago i tried to make 32cm DD coil for TGSL! Huh! Though job!
I made several coils for that coil housing and when tried to null i got very bad results! Again and again! Bad results; 1e coin barely at 25cm in air!!! Bad!
So first i used 0.25mm wire (0.28mm with resin) and than 0.28mm (0.3mm with resin). With second one i got some so so results...some 28cm for 1e in air, excellent discrimination and excellent stability. But...what's improvement here? I already got 34cm on 1e coin with 27cm DD coil? So why making larger coil with worse performances?
This is first time i tried to make larger coil for TGSL....and most probably the last time!

TGSL oscillator is simply to weak.
Best coil dimensions (dimension/performances) so far for me turned to be 22 to 25cm...

[Ultimate_Haze]

Great idea Simonbaker.
I like it.Have think about something like this.Please keep us informed.What are those materials?

[simonbaker]

Your idea is good. Just pay attention when shield is about to be applied.
I was thinking about some similar arrangement in the past, yet i had a lot of problems with shield.

Ok..one more thing that can be helpful here;
few days ago i tried to make 32cm DD coil for TGSL! Huh! Though job!
I made several coils for that coil housing and when tried to null i got very bad results! Again and again! Bad results; 1e coin barely at 25cm in air!!! Bad!
So first i used 0.25mm wire (0.28mm with resin) and than 0.28mm (0.3mm with resin). With second one i got some so so results...some 28cm for 1e in air, excellent discrimination and excellent stability. But...what's improvement here? I already got 34cm on 1e coin with 27cm DD coil? So why making larger coil with worse performances?
This is first time i tried to make larger coil for TGSL....and most probably the last time!

TGSL oscillator is simply to weak.
Best coil dimensions (dimension/performances) so far for me turned to be 22 to 25cm...

Don't throw it away, maybe good for finding ancient swords!!! Seriously, have you compared coils using bigger targets?

I will try smaller coils too. I am beginner with coils, need to make many bad ones before understanding what's going on, if ever. Maybe gods play with us, bless some coils, spit on others...

Ultimate_Haze: The "form" I made above was foam board for white core and heavy cardboard for yellow outer layer. Cardboard not stiff enough at corners with my thick wire and coil winder (which winds tightly). The foam core is too soft also, wire cuts into it - I tried coating with glue first, which helps but adds steps which defeats the simplicity.

Better materials would help, maybe corrugated plastic sheets or similar. Of course nice plastic channel loops like in Garrett "how it's made" video would be nice! I'm thinking of looking for plastic channel strips at hardware store to bend into D shape, etc. I even wondered if aluminum channel would work, making shield at same time.

-SB

[Ultimate_Haze]

I have made some more windings.This have time try to label all.So have more detailed info later.
And found this.Untill now I have try to make my TX=6mH and RX=6,5mH.
But with this values I have wrong frequency.So I try to make TX that will give me 14,5kHZ.And this is for DD27 100W which show 5,6mH and made RX 106W which gives 6,3mH/All measurments when windings constricted/ . So how come?
I had doubts on wire that have used.
And I tourched poor device so much.But it were ok all the time.
Someting rely misty with those coils.And more I investigate,more questions come not answers.
Will post some pictures with small/22cm when sealed/and DD27.
And will perform the tests that Simonbaker require

[simonbaker]

I have made some more windings.This have time try to label all.So have more detailed info later.
And found this.Untill now I have try to make my TX=6mH and RX=6,5mH.
But with this values I have wrong frequency.So I try to make TX that will give me 14,5kHZ.And this is for DD27 100W which show 5,6mH and made RX 106W which gives 6,3mH/All measurments when windings constricted/ . So how come?
I had doubts on wire that have used.
And I tourched poor device so much.But it were ok all the time.
Someting rely misty with those coils.And more I investigate,more questions come not answers.
Will post some pictures with small/22cm when sealed/and DD27.
And will perform the tests that Simonbaker require

Glad you are still at it; I have slowed down due to other activities.

I really don't know what is most important -- frequency, inductance, etc. My belief still is that with correct adjustments, many different values should be possible. But because difficult to adjust everything just right, we try to stay with proven values. Ivconic, Max, others' experience is what we rely on. But I believe possible maybe to use other values if properly adjusted.

Oscillator frequency of course depends on capacitor as well as coil, so can't just assume 6 mH gives certain frequency. Capacitors off by 20% very common.

Thickness of wire quite important to inductance when you compress/bind the wire at the end. I think thinner wire will make higher inductance when you bind it tightly than thicker wire -- my experience.

But even when oscillator frequency and inductance seems correct, some coils are better than others. This is the mystery that must be explained somehow. I hope your experiments may help. If someone can get two coils that seem very similar but one is "good" and other "bad", then we can really start analyzing and find what is the important difference. May be some very subtle characteristic like self-resonance, etc.

-SB

[Ultimate_Haze]

Hi,
Simonbaker.
Yes,I am still on it.
About the wire,have used 0,25 /bare/ wire for all the tests.And if you mean quality then they all are diferant.So for now I stick with this suplyer.
I think thinner wire will make higher inductance when you bind it tightly than thicker wire -- my experience.
Yes I saw this today too)))
What wire you use?Its seems thicker then the one I use.
so can't just assume 6 mH gives certain frequency
Just try to test some other strategy.Max advice to make TX that gives around 14,5kHz and then add some turns on RX.Basicly the same thing but just wanned to try..
Interesting thing is how Tesoro produce such coils that they can be used on each device?And we "can" make coil that suits best just on one device?
Also I think alot about some kind of comp.coil.Just still havent try.Any sugestion are welcome.How much turns,which wire d,what size former,same direction like tx and rx or oposite,etc...?