TGSL Experiments

[simonbaker]

hi simon

i have a idea! maybe you should mesure every value of voltage and capacintance etc from dfbower's pcb and then copy it to yours:P

seriously maybe that's not so bad idea.. maybe some mesurements..symply voltages and ewerything that can be mesure by multimeter.. could give you(and maybe me:P) som answers about poor working

best regards

Hi lunamay:

That's certainly a good troubleshooting idea to make sure something isn't really wrong.

I know dbbowers and my PCB are very close with TX voltage, power supply rail voltages, even LF353 null voltage (a little different, but I spent weeks analyzing that) and LM308 output noise voltage.

During some tests I trimmed our PCB TX frequencies extremely closely, and trimmed our RX coil capacitor C6 very close, and did tests using his coil for both our PCBs.

It is possible other parts could be a little different, I can try to check using in-circuit measurements.

I may have some "cold solder joints" that are an extra problem, because I hand-wire the circuit rather than use an etched PCB -- yeah, crazy.

I haven't spent much time on the audio circuit, and maybe I should look there to understand why his beep is nicer than mine. There may be difference in our power transistors or also the ground buss resistance. I used a ground buss jumper to directly connect audio ground to battery ground and I think it helped a little.

Noise can definitely affect the beep, especially the edges, so that is a likely suspect, but difficult to tell at maximum depth where signals are so small.

I also don't have the supply "bypass" capacitors on the ICs - perhaps that is a difference also.

I will slowly try to check out these ideas.

Regards,

-SB

[simonbaker]

another thing about freaquency..what is your freaquency on pcb(when signal is going to coil) maybe that's our problem
coil is working in other freaq than pcb.. or maybe i don't understand this problem


real best regards

I'm not quite sure I understand what you mean. By using the same capacitors, I was able to make my PCB and dfbowers PCB have same oscillator frequency with his coil.

Theoretically, by adjusting capacitors, it would seem we could compensate for any variations in our coil inductances. I'm not sure that works out in practice, but it is part of "tuning" your MD.

The important thing I think is to make sure your TX oscillator frequency and RX resonant frequency are separated by the correct amount. What that amount is I can't say exactly, but we know examples that work well.

Note that dfbowers made an excellent TGSL with a non-standard coil. His TX inductance was about 6.4 mH and his RX coil was about 6.85 mH, and his oscillator frequency was about 13.7 kHz.

So I don't believe exact frequency or inductance is important. It is the relationship that is important.

I think the best way to experiment with the relationship is to adjust the value of capacitor C6, up and down by .5 nF steps.

Regards,

-SB

[simonbaker]

hi jerry

do you know this "magic" voltage of coils when balancing? You said something that it's not necesery to make a almos 0 voltage. maybe you know something that can help. i have coils that have 6.5mH and 6.06mH from wire that ivonic adviced and my deep is poor. how about yours?


best regards

I know it's difficult, but you must take your MD out away from all electrical power and test there. Any electrical noise will fool you to believe your MD is not deep.

Your coils sound perfect. I also don't think the exact null is causing you the depth problem as long as you are near minimum. So no "magic" voltage in my opinion.

But maybe you could try this: find minimum with your voltmeter, then slide coils together a little more until voltage rises about 30%. Try that.

If your coils are not potted, any vibration will act like noise and ruin the depth also. You may be able to "temporarily" pot them in clay or other material.

I wish I could test your PCB and coil with a scope -- it would be very interesting.

We will keep trying to make your MD great.

Regards,

-SB

[lunamay]

maybe this cap changing is good idea.
you mean i should use a trimmer cap conected to c6? what is the biggest value of trimming caps?
i will test my md in field because of lot of noise but firstly i have to finish chasis for pcb and buy a acumulator( my idea is to separate power suply from pcb in some kind of backpack this gives me fev advantages:
1 my md will be less havy
2 my power suply can be bigger for eg. 2 or more Ah
3 (maybe) power suply will not affect my pcb

i'm thinking also to change speaker to a buzzer from a computer..it's wery smal and it'll not make this kind of noise like standard speaker.. i didn't mesure the resistance of thiss buzzer yet but this is still idea

still have to make this vaccume coil maker:P

still waiting for more time..bether weather..some money..and more ideas

best regards:]

[simonbaker]

maybe this cap changing is good idea.
you mean i should use a trimmer cap conected to c6? what is the biggest value of trimming caps?
i will test my md in field because of lot of noise but firstly i have to finish chasis for pcb and buy a acumulator( my idea is to separate power suply from pcb in some kind of backpack this gives me fev advantages:
1 my md will be less havy
2 my power suply can be bigger for eg. 2 or more Ah
3 (maybe) power suply will not affect my pcb

i'm thinking also to change speaker to a buzzer from a computer..it's wery smal and it'll not make this kind of noise like standard speaker.. i didn't mesure the resistance of thiss buzzer yet but this is still idea

still have to make this vaccume coil maker:P

still waiting for more time..bether weather..some money..and more ideas

best regards:]

Hi lunamay:

Most trimmer caps I have seen are too small. I use individual caps instead.

I cut two pins from an IC socket and solder them in parallel with C6. Then can add separate capacitors there.

But to test smaller values, you would have to replace C6 with a smaller cap, like 13 nF, and then add different caps, like .5 nF, 1 nF, 1.5 nF, or closest values, etc.

Of course, there is danger of adding noise and loose connections by adding socket. But it allows you to test. When you find the best value, you can remove socket and solder in correct cap(s).

That is my idea.

I agree it takes time to get the money and everything finished. Persistance is the key.

Regards,

-SB

[Jerry]

Lunamay, the specs you gave for your coils are very good and should produce a working detector. All I have done so far is build one set of coils that produce good results so all I can talk about is those coils. (20cm diameter, tx 5.79 Mh, Rx 6.23 Mh)

I measure the rx voltage at J2-1. I find that measuring at the output of U101 on pin 7 to be really fussy and touchy in setting the coil null. (J2-1 voltage times about 47) I connect a DVM at the scope rx channel so I get both a "picture" and an actual voltage readout. You can also read from the scope directly if it is calibrated.

The first thing to do is find the spot where the Rx voltage is minimum or near zero. Then move the coils a little closer togeather. On my coils, the Tx/Rx 20 degree phase shift falls right about where the null voltage increases to 12 Mv but I get real good air test results with 4 Mv so it isn't critical. I have not found a particular spot that seems to be the "sweet" spot.

Hope this help

Jerry

btw..... Simon: I am in Northern Wisconsin. Our elevation is about 1,200 ft msl.

[simonbaker]

btw..... Simon: I am in Northern Wisconsin. Our elevation is about 1,200 ft msl.

Looks like nice spot. Reminded me of Vermont where I will take a short trip in a few days and hopefully will see some leaves (the ones left).

Cheers,

-SB

[dfbowers]

Qiaozhi had an idea which we should test but my setup is not too good. I think dfbowers and Jerry are probably best equipped as far as I know, but anyone with low noise environment is invited.

Basically it is a tweak to one of the filter sections which raises the gain a little. Also slightly raises bandpass.

Since this falls under the general area of "filter tweaking", I'm putting in my own request also. I've already done my "insane" variation which I haven't reported on well due to time constraints, but initial tests indicate it is too far out to really interest anyone seriously, I think.

So here are two mods to try if anyone is willing.

1. Qiaozhi's mod: change C20 and C23 from .022 uF to .01 uF.

2. SB's mod: change C14, C17 from 4.7 uF to 10 uF (or back-to-back 20 uF electrolytics).
change R31, R29 from 470k to 1 Meg.
change C18, C21 from .47 uF to 1 uF.


I'll try to experment too, but will take me a while to setup and test.

The test is to see how the detection depth and overall detection quality is.

-SB

Simon,

I can do the mods and provide some feedback. I still need to replace my one LM308 first. Stay tuned..

Don

[simonbaker]

Simon,

I can do the mods and provide some feedback. I still need to replace my one LM308 first. Stay tuned..

Don

Excellent, I trust your setup more than mine.

I'm going to put a whole bunch of sockets in for the capacitors and resistors in the filter section on my last PCB so I can experiment (and curse flaky connections).

But my limitation is the darn noise around here as well as not having a whirly-gig target machine. So will be interested in your results, bench and field.

I'll be out of town for a week or two starting Oct 6, so that will explain the blissful quiet in these threads during that period. I will of course resume chatter and immerse myself in all the various tests and findings on my return.

Should be getting nice and autumnal in your neck of the woods these days. Enjoy some apple cider for me! I've sure been appreciating your TGSL you sent me for everything I test. Invaluable.

Regards,

-SB

[simonbaker]

hiding

Dennis the Mennis / Danny coaxed me out of hiding with a pm.

Had a nice trip to see foliage, enjoyed taking pics.

Got way behind in work and other priorities, so nothing to contribute right now, just so busy. Summer seems to have more time.

But I'll always try to make progress when time permits, even though old worn-out topics.

The subtle differences between my and dfbowers MD still amuse me; hard to do any tests unless very low-noise environment; probably best plan is to make a PCB using etching (like normal people) and see how it works. Every PCB is an adventure.

I also want to continue with playing with the TGSL response dynamics by varying resistors and capacitors, because I believe the standard values are just a little too fast (because designed for smaller concentric coil). But these are very minor things.

Also I need to complete a MD with pole, coil, etc., as well as get better at making coils.

I also got the Texas Instruments LaunchPad development kit for playing with the MPS430 microcontroller chips. Very inexpensive, and who knows, maybe some cool stuff can be done (but not too sophisticated like moodz et al). Big manual to read though.

So for now, I'm just lurking and enjoying reading other peoples posts. But I hope to hear that people are fixing problems with their TGSL and testing them far from noise sources. Always interesting.

Cheers,

-SB

[Jerry]

Welcome back Simon! It has been pretty quiet in here without you. I have been very busy as well but now winter has arrived so I should have more time to continue. I have everything I need to make an oven for making coil shells just a matter of assembling.

Jerry

[simonbaker]

Welcome back Simon! It has been pretty quiet in here without you. I have been very busy as well but now winter has arrived so I should have more time to continue. I have everything I need to make an oven for making coil shells just a matter of assembling.

Jerry

I envy you and dfbowers and others who can make coil shells -- you can get really commercial quality. I think my wife wants the turkey to smell like turkey, so I'm not molding plastic yet -- a custom oven is the way to go (if you have space). Look forward to seeing what comes out of it! (don't forget the paprika...)

-SB

[Dennis the Mennis]

Noise Fieldtest

Inspired by someone who went to the mountains for a field test ( I guess at boiling temperatures ) and someone who went to the beach ( same temp ? ), I've decided to make my own fieldtest ( only at freezingpoint , must be out of my mind) due to lower detection depth as should on my TGS : 1 eurocoin > +/- 20 cm , probably caused by noisy workshop and sometimes 'motorboating' due to this noise.
The average workshop noise measured on the output of the LM308 is 40 - 60 mV ac

So gathered everything from my checklist : coils ( not potted yet but glued with hotmelt ) , borrowed a computer PSU ( 12v DC > 230 v AC ) so I could take my ancient scope in the field , batterypack for clean powersupply and Arctic Equipment etc etc.

The only living soul who passed was a shepherd with his herd of sheep and me arriving from the east ( is this a kind of christmass carol ?)

With great expectation I've started my test and got really disappointed by the results :
the noise in the middle of nowhere was even higher as at the workshop and continues 'motorboating' / falls detection - fals alarm .>> 70 mV ac

Even with the other set of coils ( shielded and unshielded ) same results, no matter the nulling ...quite a cold turkey

Back home ,after I was melted , reconnected everything for further investigation and guess what ....
noise at workshop about 30 mV and stable with a great detection depth : I've got about 29 cm on a 1 eurocoin in air .

What he heck is happening? It seems that somewhere internally noise is generated causing all these unwanted side effects.


Anyone with a clue ?

What should be considered as a normal noise average ?

Does the TGS operates below freezing points with windspeeds at 5 Bft ?

Can sheep cause side effects ?

The leads from the pcb to the commands at the front of my housing are quite long but shielded ( audio cable ) , is that a problem ?

regards ,

Dennis the Mennis

[dfbowers]

Inspired by someone who went to the mountains for a field test ( I guess at boiling temperatures ) and someone who went to the beach ( same temp ? ), I've decided to make my own fieldtest ( only at freezingpoint , must be out of my mind) due to lower detection depth as should on my TGS : 1 eurocoin > +/- 20 cm , probably caused by noisy workshop and sometimes 'motorboating' due to this noise.
The average workshop noise measured on the output of the LM308 is 40 - 60 mV ac

So gathered everything from my checklist : coils ( not potted yet but glued with hotmelt ) , borrowed a computer PSU ( 12v DC > 230 v AC ) so I could take my ancient scope in the field , batterypack for clean powersupply and Arctic Equipment etc etc.

The only living soul who passed was a shepherd with his herd of sheep and me arriving from the east ( is this a kind of christmass carol ?)

With great expectation I've started my test and got really disappointed by the results :
the noise in the middle of nowhere was even higher as at the workshop and continues 'motorboating' / falls detection - fals alarm .>> 70 mV ac

Even with the other set of coils ( shielded and unshielded ) same results, no matter the nulling ...quite a cold turkey

Back home ,after I was melted , reconnected everything for further investigation and guess what ....
noise at workshop about 30 mV and stable with a great detection depth : I've got about 29 cm on a 1 eurocoin in air .

What he heck is happening? It seems that somewhere internally noise is generated causing all these unwanted side effects.


Anyone with a clue ?

What should be considered as a normal noise average ?

Does the TGS operates below freezing points with windspeeds at 5 Bft ?

Can sheep cause side effects ?

The leads from the pcb to the commands at the front of my housing are quite long but shielded ( audio cable ) , is that a problem ?

regards ,

Dennis the Mennis

29cm isn't too far off. I was getting just over 30cm until I came up with a mod. Now 36 to 45cm depending on how noisy the environment is.
http://www.geotech1.com/forums/showt...=14655&page=10

Question: Can you advance your sensitivity to the point where it chatters or does it motorboat?

The only times I have seen motorboating is when I get the coil too close to the speaker or I get close to 60Hz power.

Divide and conquor! Maybe we can try a different coil. Where do you live?

My experiments have grinded to a halt lately. Gone hunting (relics) and have been using mine lately!!


Don

[ApBerg]

Hello Dennis,

Long leads can give problems, keep them as short as possible! For screening take care that only one side from the screen is connected to the pcb ground, do not twist wire's that go to the pots.

In the past I made a lot different teso's, at high sense they can give some noise but not at a level that it is a problem..

Koud hé...

Regards.

Ap