TGSL Experiments

[simonbaker]

thank you simon for answering my question ill watch this topic with intrest
going off topic this is what i am trying to build at moment bandido 11 umax http://www.geotech1.com/forums/showthread.php?t=15725 on strip board

I'll be interested in what you do and learn. Bandido is more difficult than TGSL people say, but they all are challenge to make work excellently.

Regards,

-SB

[woody]

thank you simon for answering my question ill watch this topic with intrest
going off topic this is what i am trying to build at moment bandido 11 umax http://www.geotech1.com/forums/showthread.php?t=15725 on strip board

[simonbaker]

TGSL circuit built on cardboard, hand-wired (cont)

More photos.

Lessons learned:

- Next time, don't leave insulation on top-side jumper wires -- easier to attach test probes.

- Add test point loops on top of circuit board for oscilloscope probes.

- IC sockets with longer pins would be easier.

- Use some kind of pin at trace corners to wrap/guide buss wires.

- Don't use copper wire for buss wire, or make sure clean it first.

- Use 63/37 solder instead of 60/40 solder for quick joint.

- Don't trim parts leads until ready to solder -- they can make useful buss wires.

-SB

[simonbaker]

TGSL circuit built on cardboard, hand-wired

My first TGSL Final does not use PCB. Instead, I printed and glued the layout images on both sides of a piece cardboard, drilled holes, placed parts, and hand-wired the circuit.

It was important to line up both images (top and bottom of PCB layout) on each side of the cardboard as I glued them on. But it was not too hard if you print out the border lines and trim the cardboard after gluing the first image. The holes on both sides lined up quite well.

It was very difficult to hand-wire; took several days. But I was ignorant, had not soldered for a long time. I used bare copper wire from craft store instead of tinned buss wire. The solder would not flow on the copper wire -- caused many problems. If I were less stupid, I would have cleaned copper wire with abrasive. Now I have tinned buss wire (30 awg), much easier to work with.

I also bought new Hakko solder station and was not familiar with the iron, so learning as I built. It was messy, but I could do better next time.

I used a hand drill for holes. It worked well, easy to drill through cardboard; but takes time to drill all the holes. That is one advantage to perfboard designs -- no drilling!

To make the cardboard more permanent, I have acrylic coating I can paint on, but I have not used it yet.

Attached are some photos and in next message also.

-SB

[simonbaker]

stripboard, perfboard layouts

i would like to ask a question but dont want to get my head bitern off well here goes
now that you have TGSL final for all those who don’t necessary have the opportunity to make their own PCB’s can some present TGSL final on stipboard
thanks

I'm very interested in "stripboard" and perfboard type layouts.

My goal is to make a version of TGSL that can be made completely with parts bought at store, no customized PCB, no chemicals, etc. I do not expect it to work as well as PCB, but hopefully quite well anyway.

I have already laid out some perfboard versions of TGSL but have not built or tested so I don't want to publish yet.

My first version is very large, about 3" x 5" (7.6 x 12.7 cm) 40 x 55 holes at .1" spacing ( RadioShack). I based it on TGSL Final but left off some of the .1uF bypass capacitors. I also put much space between components if possible (free line of holes) to make it very easy to place components and test.

(I will probably add a small mod which I find useful for discriminator circuit when certain coils make the oscillator voltage go very high.)

Another version will be more compact, removing extra space between components.

Instead of copper traces, I hand-wire using buss wire. How that is done depends on the perfboard. If the perfboard has copper plated holes, then the layout is done completely with 90 deg and 45 degree angle traces. If you have stripboard, maybe you can use strips for certain traces.

With copper plated holes, the buss wire is soldered down at each bend point to a copper hole.

If bare perfboard (no copper), then easier is some ways, harder in others. You can run buss wire at any angle, but bends are not anchored, unless you put a "pin" in a hole to wrap wire around.

You can use insulated wire for long traces that snake around, for protection against shorts.

Here is another problem: knowing where to run traces. Difficult to mark board -- I am experimenting with a technique similar to making a PCB. Print out layout with Laser Printer on shiny paper and iron it onto the perfboard, just as a guide.

Is this difficult compared to soldering a PCB? Yes! But very possible. Most difficult part is soldering to pins of ICs/sockets. If use copper-plated holes, easier.

**** Note: there is also another way to build a circuit without PCB or stripboard -- this is what I actually did for my first TGSL Final. I made many mistakes, so ugly, but it works. However, it is probably not mechanically sturdy enough for permanent MD, but that can be improved also. I will describe in separate messages.

-SB

[simonbaker]

How much you have now for 1e coin?
Can you post some pictures of the coil?

TX coil about 6.3 mH, RX coil about 6.75 mH
I will try.

What cap value gives you 18,5 at oscillator?

Don't forget, this is just experimental weird stuff!...

I am trying unshielded TX coil with shielded RX coil.

I use 24 awg wire so coils are big thick monsters, only about 7 ohms resistance.

I rechecked TX coil inductance after soaking with glue -- it is actually 5.9 mH now. (That crazy coil drove me nuts... too high, remove windings, still too high, oh well, just put on glue... now lower...??? Maybe glue expanded...)

RX coil 6.7 mH, 7 ohms. But sometimes I add resistor to increase ohms.

Coils are not glued to form -- so can adjust null -- but many chatters because of vibrations -- have to wait to calm down.

Oscillator capacitor of 10 nF (instead of 22 nF) gives about 18.9 kHz.

Don't forget -- reverse the TX leads, then null coil as usual!

I think on bench I got about 20 - 26 cm (more like crack sound) of silver quarter dollar with this setup -- with normal setup maybe 15 - 20 cm and more chatter.

I'm probably exaggerating... easy to believe something is better when it's not. Also, discrimination probably all crazy.

But interesting to try just for fun, I'd like to hear if anyone else has some results.

-SB

[Ultimate_Haze]

How much you have now for 1e coin?
Can you post some pictures of the coil?

TX coil about 6.3 mH, RX coil about 6.75 mH
I will try.

What cap value gives you 18,5 at oscillator?

[simonbaker]

mirror-phase experiment

Ever tried this?

While fooling around with new TX coil (TX coil about 6.3 mH, RX coil about 6.75 mH), I adjusted oscillator cap to give frequency around 18.5 kHz and reversed leads of TX coil to get similar null phase to normal setup.

It made a pretty good detector! - seemed like less chatter and more depth??? Not sure how it affected discriminator yet.

My guess is I was on the other (high) side of the RX resonant frequency... not sure at this point.

Frequencies from 17 kHz to 19.5 kHz had interesting performance.

-SB

[simonbaker]

[quote=Darko;95865]

30-32 cm is very good in my opinion.

Actualy, sensitivity goes even to 35cm's, although with very weak sound ( standard audio part with BC517) and I see that first time when I was turn my PC off!! I was don't believe my eyes when thats happened. Simply, when I shut down my computer, sensitivity of detector raised famously, chaters stops and machine become calm, smoothness and stabile. Well, in the future, when I tend to experiment with my detector, I will send my PC to walk!!

If only small negative voltage when disc or GB pot is turned to some point, probably OK.

I can try to found spot when volatge on c12 is small as possible and at the same time residual voltage of coil goes to minimum. But why that happens? Why I can't get positive voltage on c12? Only GEB pot affect that, disc pot no influence on voltages on c12 and c15. What is consequence if I leave these in this state? Did this will affect performance on the real fild outdoor?

Disc pot should affect voltage on C15 (in disc mode -- no affect in "all metal" mode).

GB pot should affect voltage on C12.

If your detector is good indoors, I don't think these voltages will hurt anything when outdoors, just my opinion.

Phase of null signal affects whether positive or negative voltage on caps. Some coils you can get minimum but not phase for positive voltage. Or you can get phase for postive voltage, but not minimum null.

I believe the shielding of the coil affects the null a lot. Very conductive shield can make null phase shift, make difficult to get postive voltage on capacitors I think. But not completely sure, I'm studying coils now because of these differences.

I think best strategy is to try to get positive voltage on C12 by moving a little off the null point, but not too far. Even if voltage is between -.7v and 0 should be OK I think.

Again, if your detector is working great, I don't think it matters anything else. Your detector seems to be very good, congratulations. Interesting point about computer nearby -- I'll pay attention to that now! Wireless transmitter especially...

Regards,

-SB

[Darko]

[quote=simonbaker;95839]30-32 cm is very good in my opinion.

Actualy, sensitivity goes even to 35cm's, although with very weak sound ( standard audio part with BC517) and I see that first time when I was turn my PC off!! I was don't believe my eyes when thats happened. Simply, when I shut down my computer, sensitivity of detector raised famously, chaters stops and machine become calm, smoothness and stabile. Well, in the future, when I tend to experiment with my detector, I will send my PC to walk!!

If only small negative voltage when disc or GB pot is turned to some point, probably OK.

I can try to found spot when volatge on c12 is small as possible and at the same time residual voltage of coil goes to minimum. But why that happens? Why I can't get positive voltage on c12? Only GEB pot affect that, disc pot no influence on voltages on c12 and c15. What is consequence if I leave these in this state? Did this will affect performance on the real fild outdoor?

[simonbaker]

During coil nulling voltages on c12 and c15 are in plus, sens on 1e coin is about 30-32cm, but after readjusting ground trimmer for rejecting ferrite rod, voltages go to minus.No changes in sensitivity. For to reach plus voltages again I was trying to null coil again but then voltages go to minus and it is not possible to get good null and plus voltages on c12 and 15. Where I make mistake? If I want to get good null then I cant get plus voltages and vice versa until GEB trimmer set to reject ferrite?? 27cm coil 5.9mH and 6.4mH.

30-32 cm is very good in my opinion.

If only small negative voltage when disc or GB pot is turned to some point, probably OK.

If square wave voltage at JFet gate suddenly goes lower a little as you turn the pot, then not so good, means JFet is forward biased. I think this should be avoided if possible.

But if performance is good -- can't argue with voltages!

-SB

[Darko]

GEB should only affect voltage on one capacitor, C12 I think, is that true?

If your MD works fine with negative voltages, then you have no problem. Small negative voltages may be OK.

-SB

Yes.It is true. Only C12 have negative voltage, around -0.80V. C15 go to +0.700+V.
Thank you.
Best regards,
Darko

[simonbaker]

During coil nulling voltages on c12 and c15 are in plus, sens on 1e coin is about 30-32cm, but after readjusting ground trimmer for rejecting ferrite rod, voltages go to minus.No changes in sensitivity. For to reach plus voltages again I was trying to null coil again but then voltages go to minus and it is not possible to get good null and plus voltages on c12 and 15. Where I make mistake? If I want to get good null then I cant get plus voltages and vice versa until GEB trimmer set to reject ferrite?? 27cm coil 5.9mH and 6.4mH.

GEB should only affect voltage on one capacitor, C12 I think, is that true?

If your MD works fine with negative voltages, then you have no problem. Small negative voltages may be OK.

-SB

[Darko]

c12, c15 voltages

During coil nulling voltages on c12 and c15 are in plus, sens on 1e coin is about 30-32cm, but after readjusting ground trimmer for rejecting ferrite rod, voltages go to minus.No changes in sensitivity. For to reach plus voltages again I was trying to null coil again but then voltages go to minus and it is not possible to get good null and plus voltages on c12 and 15. Where I make mistake? If I want to get good null then I cant get plus voltages and vice versa until GEB trimmer set to reject ferrite?? 27cm coil 5.9mH and 6.4mH.

[nakky]

Yes I will try it. How deep is your TGSL? Which schematic?

Regards,

-SB

Kurtie's, 1 euro coin 28/30 cm.Grt Nakky.