Tweet
Actually is not exact 8.5 kHz... it is approx 8.7kHz...
-
-
Hi Don:Originally posted by dfbowers View Post
There was a version of TGSL that had a "threshold" pot in that area of the circuit. It seems worth revisiting that because that U107a is a very peculiar type of noise filtering in my understanding. It seems designed to eliminate short target responses and let through longer ones. However, it also shaves a constant time interval off all responses. Playing with R44 changes the amount of the time interval it shaves off (and also changes the size of pulse allowed through).
The senstivity pot seems more like a true "threshold" to me. However, I guess you could play with the U107a components to allow a time interval so long it would block virtually all practical detection signals.
I have never fully understood the value of "after detection" filtering vs. "before detection" filtering -- could you achieve almost the same effect with a little more low-pass filtering before detection, or is it really a radically different animal to low-pass filter the pulse train after detection? It is certainly different, but is there a better way? Shaving time off target signals is not really what we want, but is a price we pay for filtering out some "chatter". Another price we pay is that noise that breaks up a target signal may make short pieces that never get past the "after target" filter. Maybe we'd rather hear a broken beep instead of none at all.
Probably the interplay between a "threshold pot" and the sensitivity pot is confusing because the difference in the way they work is subtle. I think it's worth checking out again though.
BTW - I always used +8v for U107 V+ . Need to simulate to see if using +12v there changes the audio
-SBComment
-
I forgot to think -- what is inductance of RX coil? Maybe it is larger than our home-builts???Originally posted by ivconic View PostYes, i was confused too! Actually it was like this:
i removed previous TX from a pcb and put another, different one. I had no idea what would be TX LC tank values. I knew TX coil is 1mH (TS1000) but was to lazy at that very moment to calculate L. So i made pretty wild guess and put 0.33 (ooops! secret a bit revealed!!
) and switched it ON. It started to work just SPLENDID from a start! It's a kind of luck and nothing else. Later i really did experimented with RX cap (15nF) to see if some more benefit can be gained.. but i had no progress. 15nF seems to be most proper choice again.
How is possible to work such good on 8.5kHz?? I don't have a clue, honestly speaking!
Audio pitch from 4024 divisions were to low so i simply picked another (higher) outputs...
And that was all....
-SBComment
-
ISGL@ 8.5KC
Ivconic, do you feel that running the IGSL @ 8.5KHZ, it will be less sensitive to gold?. Although it seems that many of the vlf manufacturers are running at lower (6.5) etc.freqs. I think the two or three tone idea is great, I think I can react better to different tones than I can react to meters, leds, or digital readouts for discrimination. It should be no harder to wind the coils and null them for 8.5 as 14.khz!Comment
-
Simonbaker i like what you just wrote! (previous post to Don)All the troubles are starting just at that spot! As you maybe remember (if you followed TGS thread from the beginning); at some point i tried to achieve something by experimenting with hysteresis at last comparator. At the time i didn't understood whole concept well..so pretty soon i gave up of experimenting more!
But now, designing and playing with IGSL, again i faced same old issues at that stage!
If signal is weak - some internal oscillations can appear and chatters can randomly occur. If signal is a bit more amplified - again some falses can randomly occur!?
I changed gain in chains by rising up some resistances and previous chatters simply disappeared!
But it is not good solution at all. There is a point from which more amplification simply starts to be destructive to usable signal. All the troubles are starting and ending at last comparator. Simply as that. Chatters can occur only when there is no detection signal at all. At even weakest signal from previous stages - chatters simply disappear. So TGSL (and first IGSL version) can produce occasional chatters: if Sense is over "9" and coil is in air, away from soil. Once coil lowered close to soil - everything goes on it's place and detector is rock stable.
So i was thinking to pay more attention there and try to fix that issue in the future. Actually; i think i solved it now with this last version of IGSL. Nothing really changed at later stages - only TX is replaced with a bit stronger one!
Comment
-
It is Minelab TS1000 (24.5cm) coil. TX=1mH, RX=14.7mHOriginally posted by simonbaker View PostComment
-
True. I realized that after 8 months of using XP Gold Max Power and walking with it on many different terrains and soils.Originally posted by Wiltran View Post
XP is cool... but way out of being so "clear" and accurate, as maybe seems on first glance. So i experienced with XP many "mixed" audio responses so far. It has 3 separate tones.. but most of the time it is playing wild mixtures of those 3 tones. Simply because those were real conditions and real soil - not some testing bed at workshop.
Than i realized that you simply can not beat real ground conditions.
So i learned how to construe all those possible tone "mixtures" that XP can produce, under various soil conditions.
Living with that - i got idea for IGSL!
Comment
-
A 15nf cap with 14.7 mH coil gives a freq of about 10.72 Khz. With the transmit of 8.76 Khz. This fits very nicely with the 14.5/16.1 Khz relationship of the TGSL. This makes me wonder if you can take most any coil and check resonant frequency of the receive coil and then tune up the transmit oscillator 1.4 - 1.9 Khz lower and make it work with either TGSL or IGSL circuits.Originally posted by ivconic View Post
Very interesting thread I am very busy taking care of wife's health concerns right now otherwise I would be doing the experiment rather than just asking the question.
JerryComment
-
simple Q
Ivconic, could you please address the .1uF cap at the +power supply? I have wondered what adding .1uF to the existing 220uF electrolitic does??
Also another question on the two negative supply sources Tesoro uses.---- Is the 4024 neg supply circuit you use on the IGSL better than the ICL7660 negative source used on some Tesoro classics??? If so please explain. I have tryed to find that info in the Tgsl thread but it is so huge it is hard to search.
Thanks friends, TBComment
-
Ivconic You are viewing "private messages"? To be able to answer? Sincerely ...Comment
-
Good thinking!Originally posted by Jerry View Post
Along with Minelab TS coil i also tried Excellerator 36cm DD coil made for Fishers 12xx serie. Not good results!
I changed TX cap few times, experimented a bit - but i couldn't achieve good results as with TS coil.
Excellerator coil has TX=~2.4mH and RX=~10mH.
So this just justify exactly what you pointed...
Comment
-
Actually ... nothing! Except occupying some space!Originally posted by turtlebowl View Post
It is removed at last IGSL version because i needed that space now!
Usually i add to many components on my prototypes to occupy space and to preserve "holes" for some later experiments in vivo.
Instead soldering some modifications in "air" over pcb, later i can use those spare places on pcb.
I prefer 4024 "method" rather than 7660. It is more easy to control 4024 and later to mod it if needed.
Comment
-
Originally posted by smity View Post
?? I got no new pm's? If you sent me one; please repeat again.Might be that i deleted it by accident...Comment
-
-
I am wondering if, with lower frequencies, the 90 degree DISC range will be enough for discriminating all metals, or if we need more range. Some practical tests would be useful to answer that.Originally posted by Jerry View Post
-SBComment
.
Comment