Simple Nand metal detector problems...

[Carl-NC]

Oh, Spice definitely came from genuine EE's, it was developed at UC Berkeley by professors and students. But it originated in the 1960's when computers used punch cards for inputs and line printers for outputs; everything was upper case. When I first used Spice it was on a VAX 11-780 character terminal and line printer. There were no graphics at all.

[Dr Vel]

I remember the days. If you ever feel nostalgic I still have copies of it going back to the old 4.77 MHZ 8080A days. Orcad and Layo1 as well. I transferred my floppies (both 5.25 and 3.5's) to CD's and later to DVD's. I read once the Laser burned discs are only good for 10 years or so but I have no trouble reading CD's I wrote on an old SCSI drive 20 years ago so not sure how accurate that information is. I keep a 486 workgroups 3.1 machine around to read some of that old media which includes boxes of Colorado tapes, drive and tapes still working. Makes one wonder why they scare us with fears of data loss as I recently fired up an old RLL and a 10 MB full height MFM both of which still have all my old games intact. Off topic I suppose but after playing my old Duke Nukem and Keen games I gave a little more thought to the Nand dilemma. I just do not see how with His circuit He will ever succeed with such minute amplitudes in the search coil. I wondered about just driving the coil from a more powerful oscillator circuit and using the Nand gates for the rest of it reference and all but I think your first analysis is correct He needs to follow another path. I cannot agree with Qiaozhi however when He said forget the BFO entirely. I think they have their uses. I admit my recently resurrected BH III is my very first BFO but I kind of like it even if I cannot quite figure out why.

[Qiaozhi]

I cannot agree with Qiaozhi however when He said forget the BFO entirely. I think they have their uses. I admit my recently resurrected BH III is my very first BFO but I kind of like it even if I cannot quite figure out why.

Actually I didn't say, "forget about the BFO entirely". What I said was, "In particular, forget about the BFO." ... meaning "forget about the BFO for this application". Jeroi is attempting to improve the existing design to increase depth, as part of his first year engineering studies, but without first understanding how it works and why any attempt to increase depth will be somewhat fruitless.

I agree that building a BFO is a good first step to take when learning about metal detector circuits. However, many people have flogged BFO designs to death over the years trying to improve stability and depth, but without any real improvement. Most people who build BFOs rapidly move on to better designs after a few experiments. The main attraction of a BFO is its use as a learning tool, and the ease of construction.

[Dr Vel]

That makes sense. I probably lucked out in having one of the best ever built for my first one.

[Jeroi]

I made circuit below. Orginal circuit is from Eti561 coil part. This produces about 100-300kHZ signal I guess with amplitude of 2v and 4Vpp. Best part is that it looks driving 6mA thrue search coil and would make quite nice magnetic field. One note is that 3904 is in saturating contition. How will this affect? I made it Linear also but then amplitude is 5v wich is bad for nand gates. If the saturation affects circuit, how could I improve the circuit to be able to obtain Linear mode and 1-2v amplitude. Does the 4110 accept 4vpp to the inputs?

Also asking if it is better to have 9v amplitude to obtain best magnetic field, then reduce the amplitude for nand gates?

Anymore to do is to study how this circuit would connect to nand gates of simple BFO detector circuit? How would I connect the N3 and N4 gates and drive result to N2 input?

I made same values as orginal script in second photo inculded. Is this right way to do it?

[Jeroi]

Have to add that in the second photo it seems that transistor is in linear contition and I am guessing that it would be good and the circuitry would work like charm?

[Jeroi]

Great news! I made Eti561 schematic. Unfortunately I dont have time to test if it works, but the nand gates accept 10v as input so I can have coil with 9volts adjusting resitors. Also I did in 2 days trying to change the coil freguency as my coil ran with normal combonents as 158khz with lowest pot settings. So I finally today changed the 1n to 2n and did put 33n to ground on coil transistor. Now it reduced it to 106khz in loews pot setting wich is marvelous as I can now adjust the freguensy about 20kilos to up and nearly 5kilos to down from 111khz wihc is recommended in the eti561 pdf.

How ever, does anyone have chematic file for eti561 to import to EAGLE layout maker, as the pdf has a bit poor layout image and would need to adjust some cap hole markers and so on?

Thanks. I have frinday and monday time to make the pcb and solder all components and then have a fingers crossed that it will work. Atleast I now got right coil freguensy so it should work to some degree. I still will try to generate best current flow that I can get with this schemas with micro-cap adding some restors in series with coil to make current flow at is best.

I could not get the eti561 coil schematic work with the simple nand detector so swiched completely to eti schema but had this freguensy issue with it also. Unfortunatelly did not have time to check if I wil get any sound. So at final note, is this chema audible with normal headphones? Or does it need pro 200ohm headphone as simple nand and surf pi does?

[Jeroi]

I did a layout for eti651 with my cap values for 25cm outer dia of 30 runs and inner dia 12,5 against 55 runs coil. 33n and 2n in the circuit to make 106khz and can make more higher with course pot. Image, pdf for printing and circuit zip if one wants to edit circuitry attached.

[Jeroi]

Here is photos of my coil work:





You can see from second photo that I added also alufoil to coil and wired thrid cable into it.

Will post some photos more when coil and detector is finished.

[Jeroi]

Here is some more photos of my coil work. Unfortunately, the amount of silicone in it is too much because I did not design my coil to be used with trash can lids but omg they fit to my coil with very tight fit but the downside was that it had that groove in it. If I would have had this design in the beginning, I would have done 1cm shorter diameter coil and then gut that groove of and clue only the inner side of the lid together.



Here is photo with the tube. It is going to be attached to crutch. The positive side is, that coil is 100% water tight up to meter with this drain pipe. The connectons on the coil has rubber tighteners so it allows removing the pipe and coil. Later you could produce different type of coil and fit it here also.



Again will post some more pictures of our metal detector when it progress.

[Jeroi]

The layout image has a bug. The software had an bug that it did notice that the output pin of first nand gate must be connected to 10MEG leg. I suddenly noticed why it's leg is not connected to anywhere. I the removed wiring on schema and put wirings again and then program did see that there is missing connection. So admin, give me right to edit the photo.

Also why editing more than 30min is not allowed? I my admin career on many forums, I haven't seen any forum with this kind of limitation.

[Qiaozhi]

Also why editing more than 30min is not allowed? I my admin career on many forums, I haven't seen any forum with this kind of limitation.

The editing time is 60 minutes, which is more than enough time for you to correct any errors. If you were allowed to edit any past posts without a time limit, then continuity within the discussion could be lost, causing a lot of confusion. This is common practice throughout many other forums.

Please make a new post with your edited photo.

[Jeroi]

Hello, I need some advise. I get good signals from both oscilators on eti561 mobo. How ever the final chip the fliplop outputs only some millivolts and then cannot open the transistor which I have 2n3904.

I have coil kiloherts arround 111khz to 125khz and the output should have good feedback if coil is adjusted more than 111khz. How ever I get no sound from output. The transistor base dont get more than millivolts and thus don't amplify sound at all. How ever I get few khz signal from output, but some microamps output.

How come the final stage flipflop dont output more ac amplitude to get the transitor working?

My headphones is potted one and can be adjusted from 4ohms to 800 ohms and if you put both speakers to mono you will get up to 400ohm resistans. So resitans should not be issue here.

How ever, I had a bit weak batteries so tomorrow will test new ones. I can hear sound if I touch voltmeter on headphones trace kind of low herz breaking signal similar when speakers are connected to live receiver.

I found two mistakes on the mobo, one was small trace on nand gates that I did cut, then I found audio voltage and main voltage connected because autorouter and cut that one also. The first fix gave my ntsc oscillator to work and second one did not affect about nothing to sound output stage.

The main issue is because final flipflop dont give enough voltage out, the transitor does not work, and speakers are dead.

I can see trough oscillator that output varies about 3khz depending if I touch metal or free air meaning that the flipflop at least does output audible khz signal but it's not amplified at all.

So take a look about the circuit. My mobo gives that 890khz to flipflop. Also my coil provides 111-126khz. The reason is low voltage ouputs.

Chips are:

4013BN
4001B dont remember maybe this one right.

Both transitors are 2n3904's.

If somebody have ideas why I lack proper ouput from flipflop to audio stage, please point me about it.

[porkluvr]

I don't know it this has anything to do with your problem but any unused inputs on your CD4xxx chips should go somewhere.

IC4B has a few pins to take care of. I'm talking about SET, RST, D, and CK. Either tie them to ground or take them high, but do not leave them floating.

[Jeroi]

Orginal circuit left them floating also. I think my 4013 in the output stage is blown. It just wont output any 101010. Only trash looking 4meghz signal output.