TGSL Tuning & troubleshooting

[silverdog]

I can't comment on this issue as my knowledge of electronics is basic

I have just replaced the 4.7uF on my IGSL with 10uF, I can't notice any difference with my test targets, although this is an air test

I will from now on supply 10uF caps for these positions - if anyone who has purchased an IGSL kit wants a replacement - PM me

Silverdog

[dfbowers]

My simulation shows the wrong cap will cost you about 3.4 db gain and raise the filter center frequency 1.7 Hz.

Probably would hurt depth a little; maybe not significant. With our larger coils, I would want to err in the other direction (larger, not smaller cap).

-SB

I have experimented a bit with Eduardo's version (Silverdog's kit) and Ivconic's version.
In the two kits I put together from Andy, I ended up replacing C14, C17, R29 and R31 with the original values specified in Ivconic's version for a couple of reasons.

The extra gain can cause stability problems, particularly in the AM mode. Sensitivity could not be reduced in some cases to get rid of noise.

Also, it noticably screwed with the sweep frequency response to the point that I was missing some targets unless I slowed sweep speed a bit.

On the up side, it did increase sensitivity! It comes down to a matter of preference but there are some trade offs..

Just an opinion.

Don

[simonbaker]

Dear SB,

Fantastically quick answer thanks.

What does it mean with regard to MD operation, sensitivity and whatever, can you say? Better, same or worse?

Regards

Andy

PS Would you recommend me to replace the present caps with either 10µF ones or even Non-Polarized ones?

I would just go ahead an put in non-polarized ones, either a 4.7 uF (and jumper the empty space) or two 10 uF non-polarized in series as per the PCB layout. People have been very successful with the back-to-back 10 uF polarized -- but I personally would use non-polarized until I learn more about the long-term effects of back-to-back polarized.

I will guess that using the "wrong" capacitor (two 4.7 uF in series) would make an MD that would be superior for smaller shallow targets in conjunction with a smaller coil, but it would not respond as well to deeper targets in conjunction with a larger coil.

If you look at another TGSL modification by eduardo and produced by silverdog, you will see those capacitors are changed in the other direction -- he uses 11 uF for that capacitor, and uses a 1 Meg resistor for the LM358 feedback resistor -- this makes an even slower and higher gain response than the IGSL -- maybe better for large coils and deeper targets.

We have not really proved this in the field, although dfbowers has built and reported on eduardo's modification http://www.geotech1.com/forums/showp...postcount=4716 .

So choice of filter characteristics is not a cut-and-dried subject. But for the IGSL, I would definitely use the 4.7 uF or 5 uF value, not 2.35 uF.

Regards,

-SB

[golfnut]

caps

Hi all, Yes the caps are wrong.

I remember it like this.. For the same area plates you now have doule the distance between them = 1/2 the total value.
So knowing this you start with double the value.
You could bodge it with two 4u7 in parallel IN SERIES with two 4u7 in parallel = 4u7 non polarised. Ugly though!

[der_fisherman]

My simulation shows the wrong cap will cost you about 3.4 db gain and raise the filter center frequency 1.7 Hz.

Probably would hurt depth a little.

-SB

Dear SB,

Fantastically quick answer thanks.

What does it mean with regard to MD operation, sensitivity and whatever, can you say? Better, same or worse?

Regards

Andy

PS Would you recommend me to replace the present caps with either 10µF ones or even Non-Polarized ones?

[simonbaker]

Hi SilverDog,

have you looked at the recent posts today regarding the Electrolytic Caps? Is it true that we need 10µF each for those "Dual" caps?

If Yes, I will order them locally here.....probably non polarized as well...

I would guess that depending on case size, non polarized will need to go across both positions....or short one position out (pairwise) when using the other....

Regards

Andy

PS. Does anyone here know exactly what the effects on the MD might be when having the wrong values?

My simulation shows the wrong cap will cost you about 3.4 db gain and raise the filter center frequency 1.7 Hz.

Probably would hurt depth a little; maybe not significant. With our larger coils, I would want to err in the other direction (larger, not smaller cap).

-SB

[porkluvr]

Sorry, I did not read all way to the last post before I made my comment. You beat me too it

Jerry

Good, we have a (small) consensus!

"Those who do not study the past are going to screw things up again."
Socrates

"UN-BELIEVABLE!"
Ruby Rod



edit: Note to builders: the original TGS and TGSL used 4.7uF bi-polar capacitors in the first filter section. Using two 10uF caps in series (back-to-back) was suggested as a work-around, in case somebody could not locate the 4u7 bi-polar units. Using two 10uF series capacitors is not the best way to go but should be OK so long as you use two reasonably high quality capacitors of the same make and model. Matched, quality components should help assure the final product will be a viable substitute.

[der_fisherman]

Hi SilverDog,

have you looked at the recent posts today regarding the Electrolytic Caps? Is it true that we need 10µF each for those "Dual" caps?

If Yes, I will order them locally here.....probably non polarized as well...

I would guess that depending on case size, non polarized will need to go across both positions....or short one position out (pairwise) when using the other....

Regards

Andy

PS. Does anyone here know exactly what the effects on the MD might be when having the wrong values?

[simonbaker]

I should point out that on the TGSL we used two 10 uf caps back to back to give an approximation of a 4.7 uf.... It was actually five but with tolerance on electrolitics being so wide, it was close enough.

Two 4.7 uf in series back to back is going to halve the value. It should still work but will raise the freq. of the filter I would think.

Jerry

You're correct, wrong value will mess up the filter (assuming you like the filter the way it is).

I'm always amazed back-to-back electrolytics work as well as they do. I would have thought that a static voltage across them would slowly make one of them lose its dialectric and become like a wire, so the capacitance might slowly change.

Because non-polarized are easy and cheap where I am, I play it safe and use those.

-SB

[Jerry]

If you want to replace 4.7uF bi-polar with two back-to-back electrolytics, you should DOUBLE the value of the two capacitors to 10uF for a series grand total of 5uF. If you put two 4.7uF capacitors in series, their combined total will only be 2.35uF.

Sorry, I did not read all way to the last post before I made my comment. You beat me too it

Jerry

[Jerry]

Glad to hear it arrived safely

The 4u7 caps are arranged back to back to replace the Bipolar shown on the schematic, thats why C15 for example is shown on the pcb as C15a & Ca5b
The little lug on the 2N2907 is Emitter, the legs if you look are arranged like the pcb placement
Bd140 - with the writing facing you its ECB

good luck with your build

Silverdog

I should point out that on the TGSL we used two 10 uf caps back to back to give an approximation of a 4.7 uf.... It was actually five but with tolerance on electrolitics being so wide, it was close enough.

Two 4.7 uf in series back to back is going to halve the value. It should still work but will raise the freq. of the filter I would think.

Jerry

[porkluvr]

Glad to hear it arrived safely

The 4u7 caps are arranged back to back to replace the Bipolar shown on the schematic, thats why C15 for example is shown on the pcb as C15a & Ca5b
The little lug on the 2N2907 is Emitter, the legs if you look are arranged like the pcb placement
Bd140 - with the writing facing you its ECB

good luck with your build

Silverdog

If you want to replace 4.7uF bi-polar with two back-to-back electrolytics, you should DOUBLE the value of the two capacitors to 10uF for a series grand total of 5uF. If you put two 4.7uF capacitors in series, their combined total will only be 2.35uF.

[der_fisherman]

SilverDog IGSL - TGSL Coils

All - got my Silverdog IGSL-TGSL kit thru.

Positives
Very prompt delivery

Silverdog had written the resistor values on the packing strips which was handy.

Good quality Branded IC's like ST etc, some caps made by Wima - well respected brand

Pleased to see a little speaker in the kit too.

I tallyied the parts and it all was there to the letter.

Solid pcb, massively better than bodging something yourself.

I agree entirely.

Not sures
10k trim pots, 20k were supplied - should be ok though.

Could actually be better, in the configuration they are used in, it effectively gives an even finer adjustment. In some other positions, the correct values are more important, but not those two.

Cmos Ics loose in polythene bag - better chance of survival in antistatic foam, bag, or in ally foil wrap.

Mine were all pushed into the relevant chip sockets, far more work of course. When I send chips, either I cover a bit of Styro-foam with Aluminum (Aluminium for Brits!) foil, or use the correct anti-static foam....

All 4u7 caps on the schematic are clearly markrd - Non Polarised Electrolytics
Supplied 4u7 caps are Polarised! (normal electolytics)?

If you read the TGSL docs, it shows how you can either use non-polarized ones, or two polarized ones "back to back", a method that was new to me too, but far cheaper than the non-Polarized ones I believe....even though you need twice as many....

I can look these up.
2N2907? Oscillator transistor was metal can type - doesnt match silkscreen image - orientation?
BD140 pin out not known - could go either way?

Getting the relevant data sheet makes the 2N2907 easy to insert, I can send you a picture and a data sheet if you want of it installed, if that would help?

Same for the BD140. But do count the pins "from the front", that is, not the side where the exposed metal is seen for a heat sink. Again I can send both the data sheets and pictures of my PCB to show orientation if you wish.....you must of course get it right and we always help each other.....

All in all a good effort.

Agreed, SilverDog is very well organized and accurate. ('cos his name is Andy too!!!)

Regards

Andy

[silverdog]

All - got my Silverdog IGSL-TGSL kit thru.

Positives
Very prompt delivery

Silverdog had written the resistor values on the packing strips which was handy.

Good quality Branded IC's like ST etc, some caps made by Wima - well respected brand

Pleased to see a little speaker in the kit too.

I tallyied the parts and it all was there to the letter.

Solid pcb, massively better than bodging something yourself.


Not sures
10k trim pots, 20k were supplied - should be ok though.

Cmos Ics loose in polythene bag - better chance of survival in antistatic foam, bag, or in ally foil wrap.

All 4u7 caps on the schematic are clearly markrd - Non Polarised Electrolytics
Supplied 4u7 caps are Polarised! (normal electolytics)?


I can look these up.
2N2907? Osillator transistor was metal can type - doesnt match silkscreen image - orientation?
BD140 pin out not known - could go either way?


All in all a good effort.

Glad to hear it arrived safely

The 4u7 caps are arranged back to back to replace the Bipolar shown on the schematic, thats why C15 for example is shown on the pcb as C15a & Ca5b
The little lug on the 2N2907 is Emitter, the legs if you look are arranged like the pcb placement
Bd140 - with the writing facing you its ECB

good luck with your build

Silverdog

[der_fisherman]

Component values

I remember years ago it wasnt uncommon to get resistors from known manufactures/suppliers to the radio/tv trade etc where they were sending out resistors with wrong colour codeing, what a pain in the *** that was, tryed to get into the habbit of checking each component where possible before installing, but to this day I still forget.
Another one to look out for is pin outs ID, This particular project uses mainly T092 packages, and that in no way tells you its EBC 123, there are 5-6 pin out combinations, and that is another reason why we are getting so many conflicting storys between say TL071/BF245. Note: you can purchase transistors with same part ID same package, but different pin outs.
Try contacting some of the suppliers and often even they don't know, so again its another pain in the ***.
The best way I found of working out pin out connections is first of all get you multimeter and find the Base which set on 20 mega ohms range, you will find continuity between base and emitter and base and collector one way very much like a diode reading, the polarity will depend on whether its a PNP or NPN, the emitter to collecter reading normally will read O/C on 20mega ohms setting.
So with this first simple test you will know if its PNP or NPN and you know your base conection.
To work out the collector/emitter conections I plug the transistor into the hfc socket on one of my multimeters and swap between the E-C leads until I get the most sensable reading which is normally the lower readings of the two.
If anyone here has any better ways of doing it please bring it on here because its a very important subject that far as I know has been discussed enough.
Hope this helps a little.
Regards

I still check all transistor data sheets (still get it wrong sometimes ), but I check caps, inductances and resistors with the meter(s), its far too easy (forgetting for a moment wrongly marked devices as you correctly pointed out) to put a 4K7 where a 47K should be as red and orange are sometimes, especially under artificial light, difficult to sort out.

Due to an inadequate brain, I can almost never read a cap value and be certain I have it right......the cap meter is then the only way to go for me (never regretted making a Cap meter "addon" many years ago and buying one about 10 years ago with a larger range).......I have been too often wrong with values x10, x100 or even x1000 different to what I want......DUUUUHHHH!!

Regards

Andy