TESORO GOLDEN SABRE

[manny77]

Manny wrote:

Speaking about tx oscillator, i see not perfect shape of sine wave (distortion on the top). I know that early in the forum someone investigated that issue. Anyway i don't see if a clear solution for that problem was found. Is a good, almost perfect, sinus desirable? IGSL, if i'm not wrong, has a variable resistor for R3 to adjust sine distortion. Can i apply that to tgsl edu? Which are the benefits?

Jerry wrote:
Yes you can apply that to the TGSL. I tried it a couple years ago. It will clean up the TX waveform for a nice clean sine wave. However, I was not able to see any difference in performance in either air tests or in the field.

Jerry

Today I did some mods to my TGSL.
Changed R3 fixed 47K to trimmer 10 turns 100K. Tuned to find a good sine waveform.
I can't notice depth improvements as you already tested Jerry. But I've examinated received waves that are much more clean so discrimination is slightly better.

Silver Dollar wrote:

I found a difference on my Bandito. A pure sine wave got a bit more in air test but
A slight distortion is better for falsing. It hits the rail and stops amplitude variations.

For now I tested it only in air. Hope less falsing as you stated SD.

Disc pot was causing beeps when rotating before this mod. With oscilloscope i saw the wave changing and producing jagged profile, very ugly.
Now LF353 pin 1 is almost SINE WAVE. Only cutted on lower part because not symmetric rails (+8V and -6V). This seems to be not a problem anyway...disc pot now act like charm, without beeps on entire rotation.

During TX refiniment with new Trimmer R3 i notice that amplitude could change a little and frequency too! I saw that rotating it to make worse sine (with distortion on top) frequency goes lower (from 14.3 KHz towards about 13.7 KHz). Testing with an euro coin while rotating trimmer in that direction, detection distance is a little better. BUT waveform is worst! Probably a little lower frequence is tuned better with RX...or amplitude of tx is little (very few) more... I honestly don't know. Anyway for me 14.12 KHz seems the best value. So i adjusted trimmer to perfect sine...rising freq...Losing 2 cm of more detection.
To mantain this little more gain i achieved (previously, with lower freq.), I thought to add a little more capacitance to C2. Inserted a capacitor bank tester in parallel to it, trying to fit the right value. 1nF seems to lower freq. exactly to 14.12 KHz. And NOT CHANGING MY PREVIOUS PERFECTLY ADJUSTED ROUND SINE WAVE! So I win a couple cm of detection... Nice.

After this mod I installed my brand new NJM2068 instead of LF353. See post #6 from Davor here:

http://www.geotech1.com/forums/showt...9-TGSL-EDU-Mod

It give me some 2cm more (not 5), but I have to test it outside...in real field too. Feelings are good...

Total gain in depth today after these mods: about 4-5 cm IN AIR. Total distance for 1 euro coin: 31 cm (32 broken signal...very faint. 30 almost perfect beeps).

My coil is 27cm, homemade, you can see it in my avatar
Nulled at 280mVpp (pin7 U101a).

[Silver Dollar]

I was just playing with the minelab oscillator and managed to get it to run at 14.7 khz
just like the original TGSL one and using the AB/C configuration it only draws around 4ma!

I needed 0.124 uf to tune my 1 mh coil to 14.7 khz. It puts out 19.3V peak to peak and
draws 3.9 ma in AB/C mode (normal mode is 9.9 ma). I had another coil that was a little off
somehow so had to adjust the resistor to the -9v to 770k (from 681K) to make sure it stayed
running.

It's a nice oscillator. I could adjust the frequency from 9 khz to 15 khz and the amplitude
stays the same. Current is up there without the AB/C mod but it traps a lot of current in the
coil (150 ma!) and with the mod the power use is reduced when the negative supply comes up.

[Gwil]

Can someone tell me where I can find a description of the "AB/C configuration"? I tried a search but nothing came up - I must have been doing something wrong.

Thanks.

[paku]

what is that

[Davor]

AB->C is a mod on Musketeer IGSL that turns a Tx oscillator from class AB to class C after oscillation is established. It improves power consumption and also signal purity to a small degree. The concept may be extended to other oscillators as well.

[Silver Dollar]

Opps I had meant to post the schematic;

Original




AB/C

[Gwil]

Thanks.

[paku]

can someone explain me about this Jumper Setting

[sinclairuser]

it looks like a board revision.

[paku]

i have see on TGSL EDU schematic have 10u on that schematic why on the silk board 22u and how can replace with 4.7u bipolar is same electrolytic caps right

Schematic
http://s28.postimg.org/96dr9ae6l/TGSL_EDU_Copy.png

[mschmahl]

If you connect two 10 uf back to back you get 5 uf bipolar. So you can use one 4,7 uf bipolar instead of two 10 uf normal electrolytic caps.

[golfnut]

If u put two 10uF in series you get 5uF - to make them bipolar rather then polarised you need to put the negatives together or the positives.

If one pos and one neg join, it is is a 5uF polarised still.

[Silver Dollar]

The original was 4.7uf nonpolarized. Zoom in on this picture to see;



The 22 uf's give 10 uf and thus lower the bandpass frequency. That can get you more depth
at the cost of more chatters. It was a mod by Eduardo...

[paku]

so witch one is better for stability
4.7uf
10uf
22uf

[Silver Dollar]

It depends on your opamp. It opens the input frequency down from 7.2 hz. If your amp is noisy
down there it will get chattery. I went to 5.7 uf with a bit of depth increase and no significant
chatters. You have to sweep slower with larger values there or you'll miss some targets. It's
best to stay with 4.7 unless you have special needs...