COIN HAWK PROTEL SCHEMATIC

[strujas]

I am working on a board,,soon I or my friend Ivconic will be able to post it here...

[Sean_Goddard]

I have to say how flatered I feel that some of the best minds on this forum are interested in my "little" design.

I'm working on an updated version with an MPU to do things like hotrock ID and target characterisation (this is for when iron is offcentre of the coil and gives a positive response) This system should help eliminate falsing for once and for all IF I can get it to work properly.

Another suggestion is take a look at the Magnum by the late great Andy Flind. See those demodulators or "Flindulators" as me and Andy called them (he lived in my home town, and we were good friends)? try using those in this design instead of 4016's.

The S/N ratio of Andy''s demodulators is MUCH higher (around 15dB) than ANY existing commecial design. The only thing is that his widow owns the copyright to this design and thus any commercial design using it must pay a %age royalty to her.

If anyone would like to stick a micro (PIC or ATMEL) on this design and do some target ID or the like then feel free.

I've already added a switch to switch between 4 or 6 filters for bad ground, extended the GEB to include a BEACH mode, variable the TX power (oh yes, the pk-pk coil voltage on this machine is around 24-30V so TAKE CARE when probing around the Tx oscillator) and I'm working on a PIC controlled auto GEB system and tone ID BUT...........I'd like to see what this system can be made to do in it's current form.

How sensistive can it be made, can the noise level be reduced to make it a "perfect" machine, are the filter response times too slow or too fast, is the GEB effective and does the discriminate have the right range. In the simulations I did it all worked, but there's no substitute for actually building one and I just dont have the time (which is VERY annoying) right now I'm away from home again and just don't have the time to build things when I get home as the family demands all my attention. I feel a little guilty having started this ball rolling then being forced to take a back seat because I can't do any more on the design due to other commitments.

If I can help in ANY way given my limited time, then let me know, and in the mean time a big THANKS to all who've given this idea a vote of confidence by getting involved. I love this forum, I REALLY do, it's the one forum I really feel at home on.

[Delbert grady]

Magnum

I always thought that Magnum was a copy of Garrett Groundhog

[Sean_Goddard]

Nope sorry. I knew the Magnum designer Andy Flind very well, and he was a VERY clever man. Considering the Groundhog was a commercial machine, Andy had no need to copy anything of such a run of the mill design.

Andy is hailed by many as the greatest metal detector designer never to have worked for a large company, but his attitude was that he never wanted the "fame" or pressure that went with that. If you want a professional opinion, email Eric Foster and ask him what he thought of Andy.

The Magnum, IS I can assure you 100% Andy Flinds OWN design, the only similarities with any other design is that there are only so many ways you can skin a cat, if you know what I mean.

[ivconic]

LTC1044


Only one step left,foe me to turn ON CoinHawk and see how it's perform...
I can not find LTC1044 in any near shop...So i done some research and clued that it can be replaced with 7660, pin to pin!
Is there something i should know before try that?
Any hints?
regards!

[Sean_Goddard]

Here's hoping.......

Boy, am I ever going to look an idiot (well bigger than normal) if this doesn't work!

Watch out for the smoke fairies, don't let them escape from those IC's or it'll never work.

ERM... I think I just spotted a GLARING error around the Q6 area!! Not sure how the output is going to get to the audio stage in that config, and not sure how that error got in there. Anyone else see this?

Check out the Eldorado schematic for how it SHOULD look I think.

[Sean_Goddard]

AH NOW I see it.. Amtech your GIF file is wrong in the Q7 to Q6 connection. Q7 base whould go Q6 collector, not the base..sry. The PDF file posted above is the ACTUAL schematic I printed out and posted a LONG time ago on here (forgotten the thread it was so long ago).

Hope this that error on the GIF didnt find it's way onto the PCB I know you guys are working on. Still a cut with a scalpel and wire link will sort that out, but I suspect that other errors may crept into the design.

[amtech2005]

COIN HAWK PROTEL SCHEMATIC

I see That Mistake , Too bad carl Cant Delete that Schematic . Theres only one or two people that can delete a posting on this forum , Carl is one of them . But many Schematics over the years on this forum Have mistakes. Thats why they can be posted , To be corrected and also to be improoved . Eugene

[Geo]

Only one step left,foe me to turn ON CoinHawk and see how it's perform...
I can not find LTC1044 in any near shop...So i done some research and clued that it can be replaced with 7660, pin to pin!
Is there something i should know before try that?
Any hints?
regards!

Hi Ivconic . How are you. Long time to read your post....
What about 1266??? I think that you let it.....
Good luck with CoinHawk.

[ivconic]

Hi


Hi Geo! 1266....i make another...it will need a few components(which i do not have right now) to finish it and check....I'll have to wait a week more...But these days i finished CoinHawk.....Early this morning i tested it.....BAD! NOT WORKING!!! I have proper voltages; +5.05V and -4.38V (7660 works fine)...
But 7805 is getting hot after 20-30 seconds, although it not blows even after 5 minutes....I changed TL074 with TL084 and situation is a bit better...Now 7805 is not hot but warm??????? I am confused...
There is a sound in spkr.....Threshold pot. appears to working cose it can be adjusted....Other controls not working at all. Frequency is to high - 25kHz???? I connected testing coil, i already used with Classic III and freq. was 6.5kHz at the time....at CL3 tank cap is 330nF....At CoinHawk tank cap. is 220 nF......same coil???
I do not see reason for such high freq...Also RX part not responding - no detection....
I'll take a rest for few hours...than see...what to do with this? Scrap it and forget it for good or try to find problem...
By the way,Geo how can i conntact you via e-mail, but not to be rejected as SPAM like it was last time?
Regards!

[Qiaozhi]

Frequency is to high - 25kHz???? I connected testing coil, i already used with Classic III and freq. was 6.5kHz at the time....at CL3 tank cap is 330nF....At CoinHawk tank cap. is 220 nF......same coil???
I do not see reason for such high freq.

Hi Ivconic,

I can see your mistake.
If the CL3 has a frequency of 6.5KHz and the tank cap is 330nF, then the inductance of the coil is 1.82mH.
The transmitter osc in the Coin Hawk is a Colpitts oscillator. The 220nF cap is not the tank cap. The transmit frequency is determined by the 220nF in series with the 33nF cap, which gives a total capacitance of 28.7nF. When connected to your CL3 coil the frequency should be 22KHz.
The tesoro coils have an TX coil inductance of 5.9mH, which gives a transmit frequency of 12.2KHz.
You either need to use a Tesoro coil, or change the capacitor values to suit the CL3 coil.
The frequency of the Colpitts oscillator can be calculated using:
f = 1 / sqrt( 2pi * L * C ) where C = 1 / ((1 / C1) + (1 / C2))

[Sean_Goddard]

Ivconic, try a Tesoro head, that's what it is designed for. Also, what pk-pk value are you getting on the Tx output?

If you connect a 100K and 1K resistor in series then feed thge middle to the RX amp input you should see this signal on the output of the NE5534 (you will have to set up the nulling on the 5534 also before this).

I've been looking at the 4016 chips I used and I think I might have wired them in wrongly. Time to get the scalpel and kynar wire out again I'm afraid. Check the datasheet (4016).

I cant see much else wrong, though I havent had a really good look at it yet and i may have mixed up some of the chip supplies by accident. I thought I'd got it right, but evidently not.

KT315, Rumensat, Bugwhiskers etc maybe you guys could give it the once over, after all another pair of eyes might spot something the rest of us have missed.

[ivconic]

...


You are half way true friend!
It is not even CL3 coil....it is Fisher 8" coil, but i rearranged caps to work with CL3...Anyway you are right, i should use Tesoro coil. But first i have to make it work. Now,it is not working. Situation i explained. not look like it is ok. 7805 should not get hot or even warm. Must be some more mistakes in schematic?
I fixed already mentioned audio mistake....now audio is working. Yet, i tried to conect audio directly to +12 volts. This time audio never stops, and 10 ohm resistor gets very hot! Threshold pot does not have any influence now.
I soldered all back again,like it was and audio works again,threshold pot. have influence.....I checked with Eldorado - same....No mistakes...
Well, i do not know what to think.Either i have doubts in schematic...or might be i made some mistake...I'll check more..
Thanks Quiaozhi!
Best regards!

[Sean_Goddard]

I've checked the PDF file, and I'm stumped though you can remove R62. it was put there to allow the AMOUNT of ground signla to be set relative to the DISC signal to allow a better and more accurate discrimination set-up.

Do you have an oscilloscope? If not you could build another Coin Hawk, and Fed-Ex it to me (using my account) and I'll use PAYPAL to send you the money for the parts.

I'll trace the signal path through the circuit and maybe locate the fault. Of course I'm confident that your PCB is correct .

Please don't take this the wrong way, but I would start by stripping out ALL the IC's (sockets might be an idea) and just letting the TX do it's thing, then working though, making sure the two LM393's are phase shifting to give GND Bal and DISC signals. Next insert the NE5534 and set the output null to AS NEAR zero (relative to GND) as possible (do this iwth the Rx coil disconnected) If all is well and 7805 is still cool, REMOVE C23 and apply 1VDC (limited to 5mA0 to the INPUT of the 4016's, check the outputs ot see if you get a "chopped" signal. If you do, these are OK.

Fit IC7 and look at the output of Pin8 of the TL074 when you wave a coin above the search head. You should see a signal like a human pulse go one way then the other (not sure which). Repeat for Pin1.

I'll have to close this not as it's late but I'll come back to it later. Let me know how you got on, and I'll try to help with the FF procedures.

If I'm teaching you to suck eggs, just tell me and i apologise in advance, but fault location is my bread and butter and I love it

[Geo]

Hi Ivconic. About 1266 maybe another time to try it together. Now about frequency of Coin Hawk Qiaozhi has right. Change the capacitors but the ratio of c2 to c3 must be at least 3/1, for example c2=330..c3=100, c2=470... c3=150 etc. If it don't work don't loose your time. For my opinion it is something near the cl3 or to Bandido. What is the special ??
My email is [email protected] It is the email of my account so it is not for ever but it is very good without rejectimg anything as Spam.
My Regards