Which circuit to build?

[simonbaker]

I am wondering which circuit anyone would recommend that would be most simple but also have:

1. Very good depth/sensitivity.
2. Only basic discrimination -- iron vs. non-iron
3. Good depth on beach, at least dry sand. Wet sand would be a plus.
4. Low power usage.
5. Pretty easy to build successfully.

What do you think?

Thanks,

-S

[Max]

I am wondering which circuit anyone would recommend that would be most simple but also have:

1. Very good depth/sensitivity.
2. Only basic discrimination -- iron vs. non-iron
3. Good depth on beach, at least dry sand. Wet sand would be a plus.
4. Low power usage.
5. Pretty easy to build successfully.

What do you think?

Thanks,

-S

Hi,
depends what you mean for good depth/sens... cause many designs are not so special about depth though they are easy to do.

Other stuff you can get quite easy today...

I think a good candidate for homebrew could be the TGS-light version. It hasn't notch section, so current drain is lower than standard TGS. It's also easier to mount /assembly than standard TGS for at least 40% work and haven't notch switch/pot so it's also easy to use.

You have schematic, 2 PCB designs already made (goldy and Vladimir's PCB)and could also use a standard uMax coil if you have already one handy.

TGS is not easy detector to homebrew... but light version is simpler than standard and you could have success with it. You could use on dry sand, not wet and it will work happy there if not much iron minerals ...(no black sand etc). Consumption will be also nice and lower of many new detectors on the market.

Disc will be perfect: you could use iron/non-iron as you want (just setting pot at fully CCW) , it will reject all iron this way , then you could switch to all metals too if needed.
Sens of TGS-light is the same of TGS, so you'll be happy too.

So, only lack is at point #5... and you must be brave if not enough skilled to mount one, but I will do anyway... you risk just some time and a bunch of components ...maybe something under 30$ total.

Kind regards,
Max

[simonbaker]

Thank you Max I will study the TGS-light information.

Do you think the TGS-light is much better than the Bandido II circuits (especially for depth)?

[Molzar]

Hi Simon,

Have built TGS from original PCB (w/ notch) works well with good performance. Still tweaking it. Next project will be Bandido which is more complex.

I think Max is right TGS Light would be best choice for all your criteria, plus many others have built and can help with any problems / questions. Also much information in this thread!!!

I think you do need good solder / assembly skills and right equipment (multimeter, probably o-scope also).

Also making PCB could be whole project by itself (if you are not experienced).

Also take your time, don't rush..... this is great hobby but takes many patients....


Molzar

[Max]

Thank you Max I will study the TGS-light information.

Do you think the TGS-light is much better than the Bandido II circuits (especially for depth)?

Hi Simon,
it's good question and answer is by me related to my experience with my homemade bandidoII/tgs/tgsl.

I will answer from what I know, from my experience building TGS, TGSlight and BandidoII, so what I will say now is not related to commercially available original TGS or BandidoII.

I made BandidoII as my first Tesoro clone... it was painful experience at first cause I had to make also a concentric coplanar coil for it, but after a while and with help of many members here (Carl, Gary, Ivconic, Qiaozhi and many others) I found I could gain lot of experience tweaking bandidoII. I also bought an original coil then 9x8'' to compare with my homemades.

At the end I made right coils for it, made concentric, DD and other configurations and got nice results. I've used Gary 1st PCB design (large board) for it but that required me shielding it to get higher performances/stability.

Then I made, parallel with others like Ivconic, Nakky, ApBerg etc , the TGS and then the TGSlight too.

Ok... what about performances ?

TGS and TGSlight are exactly the same about performances, once you tune for max sens and tweak on coils. The TGSlight I've made is based on Vladimir's PCB, nice and small masterpiece that maybe miss just a big ground plane on top to be perfect.

The BandidoII is different story... I noticed BandidoII could be tweaked a lot and the use of lower noise components make it ideal for extreme tweaking, though is really complex stuff, it's really complex even getting it work properly.

The pros of my BandidoII is that goes deeper than my TGS/TGSlight with same coils. The cons are many and many... complexity, difficault components, good but not perfect disc as you find on TGS etc etc some components values are still not sure on its schematic.

So, if I must answer to your question I will say you my BandidoII goes deeper than my TGSlight... but this just after spent months tweaking bandidoII to the limit... much more than time I spent on TGS that's lot of time and about 10-12times more the time I spent on TGSlight project.

The real difference ? About 10% more depth on small stuff... e.g. coins under 3cm diameter. That's it.

If the question is: it's worth a try on BandidoII for a beginner more than on TGS/TGSL the answer will be always NO. Not cause BandidoII is not good to clone too but cause is overwelming about complexity and troubles.

Many people here tryed to make good BandidoII clones but failed and it's not a case, it's really an hard project, maybe not complex as a Relic Hawk clone attempt but really complex, belive me. It's a problem getting the negative voltage right way!

On the other side, I'm sure some that realized good BandidoII-clones will/had find much easier tweaking TGS cause of the great thread and plenty of informations already available. If you have to understand deeper the circuit, if you have to study yourself a solution, if you have to tweak at the impossible level that circuit you'll spend lot of time, money and efforts and you'll have no warranty you'll do it at the end or got better results than with TGS/TGSL.

Many people here will say you BandidoII is bad thing... it's just average one etc etc, and this probably apply to commercial one, but not to an optimized one. Also it's probable that a commercial TGS could be better on average soil than a commercial BandidoII (that has fine GEB tuning), but as I told you, my thoughts are related to cloned/optimized circuits not originals Tesoro's.

Anyway my advise to you is making TGSL, even if not perfect (none is), it's GREAT detector and it's easy enough to get it work without much pain and headhackes, and goes deep as TGS goes, so MUCH MORE than many top of the range VLFs today on the market that will cost you thousands dollars and just a bit under what you can do with a super-tweaked bandidoII, you can live with that 10% less I think.

Kind regards,
Max

[simonbaker]

Thanks again Max and Molzar -- you have convinced me the Bandido II is definitely not for me now. I mistakenly thought the Bandido was a simpler design without looking at it. Easier to build is important for me.

I have some experience with electronics and have an oscilloscope, but I am not very experienced with PCB making. I mostly experiment with simple circuits, some kit building. But I am very excited by Metal Detectors, they are some of the most interesting circuits and what they do is like magic.

The Amigo II is even easier I think, or maybe nobody plays with it much. Do you know how it compares for depth with the TGS Light?

I am reading the TGS thread, but sometimes it is difficult to know when they talk about TGS Light or instead the TGS with notch. But I'll keep trying.

It is great to have so many experienced people on this forum, it is the best!

-SB

[Max]

Thanks again Max and Molzar -- you have convinced me the Bandido II is definitely not for me now. I mistakenly thought the Bandido was a simpler design without looking at it. Easier to build is important for me.

I have some experience with electronics and have an oscilloscope, but I am not very experienced with PCB making. I mostly experiment with simple circuits, some kit building. But I am very excited by Metal Detectors, they are some of the most interesting circuits and what they do is like magic.

The Amigo II is even easier I think, or maybe nobody plays with it much. Do you know how it compares for depth with the TGS Light?

I am reading the TGS thread, but sometimes it is difficult to know when they talk about TGS Light or instead the TGS with notch. But I'll keep trying.

It is great to have so many experienced people on this forum, it is the best!

-SB

Hi,
the best (say easy) 2 ways to PCB are :
1)press-n-peel sheet
2)photoresist processing

1) you simply print by a laser printer the pcb pattern on the press-n-peel sheet (cost each some usd... sometimes you could find one for 50cents or less at "discount" prices) , look here for info: http://www.cibs.co.uk/etch/x7.html

Well... after you print it you just need iron set and a board: place the sheet with toner facing a polished (very well) copper clad board (fiberglass FR4 are the best for this kind of stuff)... the hot will make toner pass on copper and then you'll actually transfer it on the copper layer preventing chemical/acid to etch there where it lye.
Easy as this... BUT take care on thin lines... sometimes p-n-p method is ineffective if too thin lines are present.

2) It's the "professional" way... you need presensitized boards (e.g. bungard brand are really cool for this)... and uv lamps...

The boards are made as usual copper clad stuff FR4 but have also a special varnish on them that's sensitive to UV radiations. They are protected normally by a kind of label you must remove just before use.

You need print (ink jet could work too in this case) on a transparent film... then it's like a photographic process... you remove the label on board and put the film on the UV-sens varnish side (usually green or blue coloured depending on photoresist varnish used)... then use a glass to press film on varnish and then leave some minutes under UV lamp(s).

Usually this is done by special apparatus (UV lamps inside, with timer etc etc) but you could also homebrew it with an old scanner if are a bit broke... or similar stuff... just you need some uv-tubes of at least 8w power each... 2 are often enough for small boards, 4 are the standard choice.
Someone like leds... but you need 40-50 UV leds... and for me it's wasted time putting them all together just to make enough power a single small fluo-tube can give.

You have don't look UV light... just put the film , the glass over , close the box and give power... then remove power before opening again. This is the "blind" way... but better blind way that becoming blind like some LED addicted dude! Anyway power is limited and so limited exposure is safe too but avoid if not needed.

Look here for an example: http://www.instructables.com/id/The-...atbed-scanner/

Exposure is often from 10seconds to 4-5minutes depending on many factors like light pressure at film... say that if you use the scanner-kind you need maybe less than 30seconds in many cases. If you use far lamps (e.g. 10cm distance) you could need also 5minutes depending on resist kind too.

The good way is cutting a strip of board and cover with something then give 30seconds intervals for 5 minutes of UV each time... then etch all and see where board is made good to understand time required: once you know for your particular boards you'll use that time always with them and got perfect boards all the time.

Then you'll use a developer solution like photos (but it's sodium hydroxide solution for resist most of the times)... using gloves and glasses to prevent burn skin etc

Wash and etch with chemical solution for copper (there are many... the most common is still today iron perchlorate solution water based).

Copper will dissolve in the etching solution pretty easy in few time (some 10s minutes in worse case).

There's also a way to make your own presensitized boards: http://www.ee.cuhk.edu.hk/~ymfung/ppp/etching.html
but I don't suggest you cause of bad smell and too many chemicals around.

The baking of resist is kind of "art" but its smelly and dirty art... so better avoid if possible.

Look some "all process" tutorials here:
http://www.electricstuff.co.uk/pcbs.html
http://www.computronics.com.au/kinsten/
http://hem.passagen.se/communication/pcb.html

Last step is drilling and polishing ... then you are ready to mount components and solder.

Kind regards,
Max

[ANDREAS]

Hi,
the best (say easy) 2 ways to PCB are :
1)press-n-peel sheet
2)photoresist processing.......

Max

Hi Max.
Difficult work for me.Now I use a Inkjet and manufactures PCBs double layers with sillscreen in 1 hour, very easy and fine analysis .The problem is only one (best calibration machine for printing PCBs)
regards

[Max]

Hi Max.
Difficult work for me.Now I use a Inkjet and manufactures PCBs double layers with sillscreen in 1 hour, very easy and fine analysis .The problem is only one (best calibration machine for printing PCBs)
regards

Hi,
very nice machine!

I'd like to make one too using an old plotter thing but cannot find enough time for little mods required.

I also had on hands a very fine machine some years ago... cause a friend borrow me for a couple of months... it was like little plotter stuff but using a rotating tool to cut tracks on pcb directly... it made also kind of silkscreen on relief... very pretty toy used for proto stuff but I had to get back to owner after few time and then device was lost cause of fire in the building it was employed at that time.

Unfortunately nice stuff like this are often too expensive for the hobby but your solution is what I had in mind for something 15years...

Sooner or later I will do it similar way.

Kind regards,
Max

[simonbaker]

Hi,
the best (say easy) 2 ways to PCB are :
1)press-n-peel sheet
2)photoresist processing

1) you simply print by a laser printer the pcb pattern on the press-n-peel sheet (cost each some usd... sometimes you could find one for 50cents or less at "discount" prices) , look here for info: http://www.cibs.co.uk/etch/x7.html

Well... after you print it you just need iron set and a board: place the sheet with toner facing a polished (very well) copper clad board (fiberglass FR4 are the best for this kind of stuff)... the hot will make toner pass on copper and then you'll actually transfer it on the copper layer preventing chemical/acid to etch there where it lye.
Easy as this... BUT take care on thin lines... sometimes p-n-p method is ineffective if too thin lines are present.

2) It's the "professional" way... you need presensitized boards (e.g. bungard brand are really cool for this)... and uv lamps...

The boards are made as usual copper clad stuff FR4 but have also a special varnish on them that's sensitive to UV radiations. They are protected normally by a kind of label you must remove just before use.

You need print (ink jet could work too in this case) on a transparent film... then it's like a photographic process... you remove the label on board and put the film on the UV-sens varnish side (usually green or blue coloured depending on photoresist varnish used)... then use a glass to press film on varnish and then leave some minutes under UV lamp(s).

Usually this is done by special apparatus (UV lamps inside, with timer etc etc) but you could also homebrew it with an old scanner if are a bit broke... or similar stuff... just you need some uv-tubes of at least 8w power each... 2 are often enough for small boards, 4 are the standard choice.
Someone like leds... but you need 40-50 UV leds... and for me it's wasted time putting them all together just to make enough power a single small fluo-tube can give.

You have don't look UV light... just put the film , the glass over , close the box and give power... then remove power before opening again. This is the "blind" way... but better blind way that becoming blind like some LED addicted dude! Anyway power is limited and so limited exposure is safe too but avoid if not needed.

Look here for an example: http://www.instructables.com/id/The-...atbed-scanner/

Exposure is often from 10seconds to 4-5minutes depending on many factors like light pressure at film... say that if you use the scanner-kind you need maybe less than 30seconds in many cases. If you use far lamps (e.g. 10cm distance) you could need also 5minutes depending on resist kind too.

The good way is cutting a strip of board and cover with something then give 30seconds intervals for 5 minutes of UV each time... then etch all and see where board is made good to understand time required: once you know for your particular boards you'll use that time always with them and got perfect boards all the time.

Then you'll use a developer solution like photos (but it's sodium hydroxide solution for resist most of the times)... using gloves and glasses to prevent burn skin etc

Wash and etch with chemical solution for copper (there are many... the most common is still today iron perchlorate solution water based).

Copper will dissolve in the etching solution pretty easy in few time (some 10s minutes in worse case).

There's also a way to make your own presensitized boards: http://www.ee.cuhk.edu.hk/~ymfung/ppp/etching.html
but I don't suggest you cause of bad smell and too many chemicals around.

The baking of resist is kind of "art" but its smelly and dirty art... so better avoid if possible.

Look some "all process" tutorials here:
http://www.electricstuff.co.uk/pcbs.html
http://www.computronics.com.au/kinsten/
http://hem.passagen.se/communication/pcb.html

Last step is drilling and polishing ... then you are ready to mount components and solder.

Kind regards,
Max

That is very valuable and will help a lot... thanks a million. Finding resources often is the most time wasting problem, so when people like you provide detailed information and references, it is worth gold and helps us all !

My question is: how much better is a PCB for performance than soldering wires? Yes it is much easier, but would wires work OK?

Andreas PCB machine, very slick! We all need one someday...

[Molzar]

Simon,

I have done some MDs this way but they had (many) less components than TGS (or TGS Lite) and were BFOs , not VLF.

MDs are pretty touchy animals about component placement and x-talk signals. You could try wiring, but I think would be very frustrating and performance results may be unknown.......

I think PCB is way to go to get predicitable performance on TGS Lite.

Also again... MDs are not easiest projects to build........

Molzar

[Max]

That is very valuable and will help a lot... thanks a million. Finding resources often is the most time wasting problem, so when people like you provide detailed information and references, it is worth gold and helps us all !

My question is: how much better is a PCB for performance than soldering wires? Yes it is much easier, but would wires work OK?

Andreas PCB machine, very slick! We all need one someday...

Hi,
you can try with wire-wrap also... not that it doesn't work... most of the time it's ok for MDs also, but not complex kind.

Me personally breadboarded many times entire circuits of this kind... with controversial results: sometimes very nice work ... sometimes with really furstrating situations.

In good VLF I will at least choose some protoboards like veroboard and use solder joints if I cannot or don't wanna make real etched PCB.

For example I've made a very tricky and pesty PI MD (the GoldscanIV by Eric Foster) replica using just protoboards... and it works! But then I made it again on real PCB thing and now is much more stable and "easy".

Sure you can make all old projects like e.g. sandbanks or twinloop or e.g. magnum MD on veroboards... if you want... but keep in mind that's sometimes more furstrating trying to fix troubles after circuit already works (somehow) than making pcb first and spending some extra time and efforts before soldering stuff.

To me it's not a big problem... I usually made RF circuits using not etched PCBs soldering pins on ground layer and making flying connections where needed but it's not the way I prefer deal with MDs... cause when you have to remove stuff or replace it's a huge mess with flying joints.
Also, you have to be very skilled with soldering iron and solder stuff to make things good without burning components.

At the end it's your choice and nobody here will tell you you MUST do it in a way or another...

Kind regards,
Max

[simonbaker]

Hi,
you can try with wire-wrap also... not that it doesn't work... most of the time it's ok for MDs also, but not complex kind.

Me personally breadboarded many times entire circuits of this kind... with controversial results: sometimes very nice work ... sometimes with really furstrating situations.

In good VLF I will at least choose some protoboards like veroboard and use solder joints if I cannot or don't wanna make real etched PCB.

For example I've made a very tricky and pesty PI MD (the GoldscanIV by Eric Foster) replica using just protoboards... and it works! But then I made it again on real PCB thing and now is much more stable and "easy".

Sure you can make all old projects like e.g. sandbanks or twinloop or e.g. magnum MD on veroboards... if you want... but keep in mind that's sometimes more furstrating trying to fix troubles after circuit already works (somehow) than making pcb first and spending some extra time and efforts before soldering stuff.

To me it's not a big problem... I usually made RF circuits using not etched PCBs soldering pins on ground layer and making flying connections where needed but it's not the way I prefer deal with MDs... cause when you have to remove stuff or replace it's a huge mess with flying joints.
Also, you have to be very skilled with soldering iron and solder stuff to make things good without burning components.

At the end it's your choice and nobody here will tell you you MUST do it in a way or another...

Kind regards,
Max

Thanks Molzar and Max I appreciate the advice. I was wondering what is the best way to experiment with modifications. MDs are very sensitive, so sloppy circuits may not work -- makes it difficult to experiment sometimes. I will try to learn PCBs. We really need a cheap PCB maker machine.

Do you have an opinion on best/cheapest software for designing PCBs? And also circuit simulation?

Cheers,

-SB

[Max]

Thanks Molzar and Max I appreciate the advice. I was wondering what is the best way to experiment with modifications. MDs are very sensitive, so sloppy circuits may not work -- makes it difficult to experiment sometimes. I will try to learn PCBs. We really need a cheap PCB maker machine.

Do you have an opinion on best/cheapest software for designing PCBs? And also circuit simulation?

Cheers,

-SB

Hi,
you can use LT SwitcherCad (LTspice) simulator that's free. www.lt.com

For PCB you can try ExpressPCB software... really essential one... and with no print , save etc...

Or you can find a copy of e.g. SprintLayout software... that's easy and good for small stuff.

Kind regards,
Max

[Qiaozhi]

I can also recommend: http://www.pad2pad.com/