Simplest PI detector possible?

[geoscash1]

Hey bernte_one,

Yea, the parts placement is the same as in that image/pcb. I changed the R1 to a 5k trimmer and for whatever reason it seemed to work better, so Harp added it to the new drawing/pcb.

Qiaozhi had some suggestions for me and I tried all of them yesterday with another board but for whatever reason this particular layout or design does not like any changes.
I tried numerous things, one part at a time but the unit wouldn't work with a single one of them, tried replacing R1/R2/C1, R3, R4/R7, removed C2 and tried a few different transistors and ended up putting it back to the way I did it originally as in the schematic/pcb by harp except I used 5K trimmer for R1 and used an LF412cp for the op amp and a MPSA13 trans.

Its not bad as is but I sure thought with some minor changes it could be plenty better, I'm thinking of changing how the 555 is wired, it was suggested that its an unusual
layout and R1/R2 work in reverse compared to the normal way of doing it. I will keep messing with it since I made 2 of them, I can continue to test and develop with
one of them.

[Silver Dollar]

Well I finally got back to mine and found I missed tying pin 8 on the 555 to plus.
Now I get a continuous tone. I've got to get it going for this weekend as we are
going to the beach!

[geoscash1]

Excellent Silver Dollar, hope to see some video too....

I found that the R1 (1.2K) is no where near correct for this circuit, at least not for mine. I built a little 4 toggle dip switch with 4 different resistors and also did the same with
with 4 different caps, giving me 16 different options for each with 32 combinations total to test R1/R2/C1 combo's.

My first go around mine did the same as yours, continuous tone, so I played with the resistors (R1) until I found where I could get a good variable threshold tone and decent detection.
Its now set up a little different than the schematic with different values in a few places and now am getting 8 inches in air with large coins, 5 inches with smaller coins
and am also getting 2 inches with a piece of 1/2x1/2 pepsi can using a 7-1/2 inch spider/basket coil with 26awg red colored mag wire with 25 turns at 330uH. It is less sensitive with all
of my other bundle wound coils from 300 to 450 uH .

Try a 3.7k to 4.7k at R1 with your current parts, it seemed to be close in that range to what this circuit likes or where it seems to work at. I used 1.2k all the way up to 10k and best result were between 3.3k and 6.8k at R1. I changed R4 to 1k and R7 to a 3.3meg. My current configuration is 560 ohm at R1 and a 100N at C1 and seems to be working great.

[Silver Dollar]

Well I got mine working by using a different coil. I used a 470 ohm damping resistor and a 440 uh coil.
That didn't work so I found my 330 uh basket weave coil. That works but only 2 - 3" detection depth.
I have a 220 ohm resistor and some wire so could make a lower inductance coil. I'd like to get 8" in the
sand at the beach but not sure if this design will do it?

Any one know how to figure out if a certain coil will work on this design? It's tied to the speed of the coil
I'd guess...

[geoscash1]

Hey Silver Dollar,

The coil and R1/R2/C1 are definitely all tied together. While messing with mine I ended up with 560R at R1 with a 330 basket/spider wound coil (although it is open and not potted yet). I have tried
numerous combinations and am right at 8 inches in air with the above with a large coin and small coins (dimes, pennys) being right around 4 to 5 inches. I mocked it up and took it outside
and tested it and with a quarter buried at 5 inches in my soil, it just faintly detects it, it pops off a good signal on any of the coins to 3 inches, then starts to lose its sensitivity, although
if you listen closely you can hear the changes in the audio for the deeper objects.

Current parts setup:
R1 - 5k trimmer (set at 568 ohm)
R2 - 100k
C1 - 100N
R3 - 470R 1 watt
R4 - 1k
R7 - 1.2 Meg
Op amp - LF412cp
Coil: 26 awg magnet wire, 25 turns on spider jig, 7-7/8" diameter, no shield, approx. 330uH.

The coil definitely has a major influence on how this particular circuit works, it likes my surf pi coil (333uH) and a couple in the same
range, but under 300 or over 400 it needs big resistor changes at a few spots to work and it is usually not as good as it is in the 350uH range.
My open spider magnet wire, 330uH coil works better than with my spider coil that is wound with Teflon, silver 7 strand in the same size.

I have some parts on the way and will be testing some other changes when they get here. There are 2 changes I think may improve
its depth/sensitivity a little and overall capability, I am building another with a few changes to the schematic and how the 555 is wired,
hopefully it works...lol... if not, no big deal, it is a fun and cheap little educational project for me..

[geoscash1]

Also the starting voltage makes a huge difference.

Using the recommended voltage (depending on which schematic is used), one calls out 9v the other shows both 9v or 12v and one shows 12v.
I found it will not work under 8.86v and if you setup with 12v with a specific coil it will make a huge difference especially when voltage has dropped below 10.2v or new batts are installed.

One of my boards was setup up this way, using 12v (8AA alkaline batts, they were older but still carried 11.1v),using a 330uH coil, once the voltage was below 10, it became weak so I put new batts in and once fired up, it no longer worked as it did prior to the new batts. The new ones read 13.2v and now the coil/sensitivity was out of adjustment and a lower inductance coil was needed to make it work like it did. I had to use a 240uH coil with it with the current settings and resistor combinations I used to get it to work as it did before.. it doesn't make much sense to me except in that it may be originally designed with 9v in mind and since my original batts were not new and had 11v it worked, once it had 12v+ applied it was no longer in the range it needed to work correctly.... hope that makes sense... Anyhow, that's what I found with my current setup, and it was the same with either board I made..

[baum7154]

Hey Silver Dollar,

The coil and R1/R2/C1 are definitely all tied together. While messing with mine I ended up with 560R at R1 with a 330 basket/spider wound coil (although it is open and not potted yet). I have tried
numerous combinations and am right at 8 inches in air with the above with a large coin and small coins (dimes, pennys) being right around 4 to 5 inches. I mocked it up and took it outside
and tested it and with a quarter buried at 5 inches in my soil, it just faintly detects it, it pops off a good signal on any of the coins to 3 inches, then starts to lose its sensitivity, although
if you listen closely you can hear the changes in the audio for the deeper objects.

Current parts setup:
R1 - 5k trimmer (set at 568 ohm)
R2 - 100k
C1 - 100N
R3 - 470R 1 watt
R4 - 1k
R7 - 1.2 Meg
Op amp - LF412cp
Coil: 26 awg magnet wire, 25 turns on spider jig, 7-7/8" diameter, no shield, approx. 330uH.

The coil definitely has a major influence on how this particular circuit works, it likes my surf pi coil (333uH) and a couple in the same
range, but under 300 or over 400 it needs big resistor changes at a few spots to work and it is usually not as good as it is in the 350uH range.
My open spider magnet wire, 330uH coil works better than with my spider coil that is wound with Teflon, silver 7 strand in the same size.

I have some parts on the way and will be testing some other changes when they get here. There are 2 changes I think may improve
its depth/sensitivity a little and overall capability, I am building another with a few changes to the schematic and how the 555 is wired,
hopefully it works...lol... if not, no big deal, it is a fun and cheap little educational project for me..

----------------------------------------------

A couple of things that might make a big difference in performance with the spider wound coils:

Don't 'pot' the coil at all. If it is potted with any dielectric it will add capacitance and slow the coil down. If you review the CHANCE PI COIL thread, I have posted pictures that show using polyethylene foam sheet and backer rod to fill out the profile and a cloth tape over wrap, then fiberglass, no flooding of the coil.

Also I would recommend that you optimize the damping resistor to obtain 'critical damping'. See BB SAILOR's paper 'How to make a fast PI monocoil' for the adjustable network that will quickly help you determine how much resistor is needed. From what I can see having wound many of the spidercoils 470 ohms is probably way too low depending on how much capacitance the detector adds. For example my 8" 335uh spidercoil is wound with teflon insulated #24 wire and requires 1060 ohms in circuit for proper damping. The reason your magnet wire spider coil works better may very well be because it is a closer match to your 470 ohm damping resistor. Your #26 teflon insulated coil should be even faster due to less inter wire capacitance and consequently require an even higher damping resistor. I'd try a 1000 ohm damping resistor just for fun to see what happens.

Best of luck!

Dan

[geoscash1]

Hey Dan,

Thank you for the great advice...I actually had planned on doing the coils just like yours (cloth wrap with foam backer) when I get the supplies to finish them,
I'm just using them as test coils for now while I mess with these different boards.

I went and swapped my 470R 1 watt for a 1k 1/2 watt damper (unfortunately I don't have anything higher than 1/2 watt so that is what I installed) and you are right on the money..
Since my R1 is a 5k trimmer I did have to make adjustments to get it in a good ballpark, so now instead of 560R it is reading in the 3.6k range but has made a world of difference..
Although it is getting a little warm it added 2 inches to the air test, so the mag wire spider coil is sounding off at 10 inches with a quarter (faint at 12) and 8 inches with a nickel (faint at 10),
a penny and dime approx. 6 to 7 inches and a solid signal at 4 inches with my folded piece of 1/2"x1/2" pepsi can.

It has a bit of a crackling threshold depending where I set it but I think once it is out of my garage and out in the real field it will settle down some.. I think I will keep R1
as a trimmer instead of using a permanent single resistor and then add an off board pot for adjustability, it seems to work with multiple coils as long as I can make
adjustments to R1.

Oddly enough, the Teflon wire spider coil outstretched the trimmer, so I swapped the 1k damper for a 1.5k 1/2watt and it is close to the mag wire spider coil but doesn't quite
get the same result in air as the mag wire coil and the coil actually gets warm along with the fet.. Not to sure why they are so different..


I'll continue to play with it and make small changes here and there, I have some parts on the way to try out and see what difference it makes with this little board..

I am really in the dark as far as what it is actually doing, where the timing, pulse rate or frequency is (terrible with math...lol) since I don't have a working scope, but its fun and a great
learning experience trying different things and seeing the changes it makes, I am to the point where I have a good idea and/or know what it will do with minor changes
in certain areas so it has definitely helped me understand so much more.

Thank you again for the advice, much appreciated..

[Silver Dollar]

I had used a 39k and 68k so when it didn't work I added 2k to the 120k and it started working with the 330 uh coil.
The second op amp compares a signal with a delayed signal so depending on how fast your coil is changing the 10uf
might make a difference. Also making the 120k a pot might allow some adjustment too. It would be nice to have it
adjustable to a coil so you could use most anything.

I wound up a Tyvek coated wire wrap wire coil last night but made it 450 uh as most other PI projects can use this
value. I hope to get it working with that coil. It would be nice to know what values influence the operation so you
could fine tune it to your coil and use. I'd like a low power unit so might try finding a bipolar transistor to fire it.
I'd like to get it to run at 9v too...

[geoscash1]

Silver Dollar, Which op amp are you using?


Don't use or lose the LM358, it is weak for whatever reason, use a TLO72 or TLO82, you will be surprised at the difference, if you have an LF412 it works even better..
There may be others that work better but these are the ones I have used and tested so far.. They are all the same pin out so no changes needed in the trace/wiring.

[baum7154]

Hey Dan,

Thank you for the great advice...I actually had planned on doing the coils just like yours (cloth wrap with foam backer) when I get the supplies to finish them,
I'm just using them as test coils for now while I mess with these different boards.

I went and swapped my 470R 1 watt for a 1k 1/2 watt damper (unfortunately I don't have anything higher than 1/2 watt so that is what I installed) and you are right on the money..
Since my R1 is a 5k trimmer I did have to make adjustments to get it in a good ballpark, so now instead of 560R it is reading in the 3.6k range but has made a world of difference..
Although it is getting a little warm it added 2 inches to the air test, so the mag wire spider coil is sounding off at 10 inches with a quarter (faint at 12) and 8 inches with a nickel (faint at 10),
a penny and dime approx. 6 to 7 inches and a solid signal at 4 inches with my folded piece of 1/2"x1/2" pepsi can.

It has a bit of a crackling threshold depending where I set it but I think once it is out of my garage and out in the real field it will settle down some.. I think I will keep R1
as a trimmer instead of using a permanent single resistor and then add an off board pot for adjustability, it seems to work with multiple coils as long as I can make
adjustments to R1.

Oddly enough, the Teflon wire spider coil outstretched the trimmer, so I swapped the 1k damper for a 1.5k 1/2watt and it is close to the mag wire spider coil but doesn't quite
get the same result in air as the mag wire coil and the coil actually gets warm along with the fet.. Not to sure why they are so different..


I'll continue to play with it and make small changes here and there, I have some parts on the way to try out and see what difference it makes with this little board..

I am really in the dark as far as what it is actually doing, where the timing, pulse rate or frequency is (terrible with math...lol) since I don't have a working scope, but its fun and a great
learning experience trying different things and seeing the changes it makes, I am to the point where I have a good idea and/or know what it will do with minor changes
in certain areas so it has definitely helped me understand so much more.

Thank you again for the advice, much appreciated..

----------------------------------------

Glad to help!
Your distance with the 8" magnet wire coil is about what I get with my 8" coil on similar targets i.e. nickle, dime, and penny.

The warming issue may be due to the transmit pulse length and the small wattage of your damping resistor. As far as the teflon wire coil not getting the same distance I think the jump to a 1.5K damping resistor is too much. If you have a 1200 ohm resistor try it or add a 100 ohm in series with the 1K ohm damper you used at first. I think the answer is in the range of 1050 to 1250 ohms. Alternatively two 1/2 watt, 2.2k resistors in parallel gets you 1100 ohms and gives you a 1 watt resistor.

Another possibility exists with the heating. When winding spider coils it always occurs to me that it would be very easy to accidentally reverse winding direction when jumping a slot after every 5 turns. If winding direction gets reversed I think it would cause havoc. The check on this is if you have verified inductance with an LCR meter vs the calculated number of turns to get to your target inductance. There would also be visual indications of reversal in counting the number of wires in all the slots. For 25 turns it should be 5 in each slot for 30 turns 6 in each slot. (taking into account the 5 each, 1 slot jog offsets of course) When winding a spidercoil I always mark both sides of the coil form for winding direction. Also mark every slot jog location and identify the feed wires connected to the innermost and outermost windings. This is important to insure that the outermost winding is tied to system ground so that the self shielding will work properly. You might want to make sure that the outermost winding on the teflon wire coil is tied to ground.

By the way I have found that no other shielding is needed with this coil geometry. Even operating at 8us delay this coil never sees my hand gripped completely around it. No additional shield layer means a faster coil and no additional signal attenuation.

Regards,

Dan

[Silver Dollar]

I tried a TL082 and LM358 they seem about the same. Maybe you'll see a difference when
you start getting some distance out of it. My new coil at 450 uh didn't work too well so I
wound another. This one came out to 340 uh. I hope it works a bit better. I'll give it a try
tomorrow. I have some other op amps I'll try but imagine you need the right coil and damper
resistor combo first. I think the design is a bit insensitive but with careful tuning can yield
acceptable results. I'm not sure if you even want a fast coil with this design....

[baum7154]

I tried a TL082 and LM358 they seem about the same. Maybe you'll see a difference when
you start getting some distance out of it. My new coil at 450 uh didn't work too well so I
wound another. This one came out to 340 uh. I hope it works a bit better. I'll give it a try
tomorrow. I have some other op amps I'll try but imagine you need the right coil and damper
resistor combo first. I think the design is a bit insensitive but with careful tuning can yield
acceptable results. I'm not sure if you even want a fast coil with this design....

--------------------------------

What is the minimum sample delay with this detector? If it will go to 10 or 15us it wouldn't hurt to have the fast coil.

[geoscash1]

I really don't know baum71, I have no way to test other than trial and error and volt meter.... I tried using a 555 software for setting R1/R2 and C1
but it never works with the stuff it says it should... so I just keep making changes until it works like a metal detector with the parts I have or I order
what I want to try with it......lol..

I can send you one of these two I have made if you want and you can check it out and measure everything and see what it is actually doing timing wise..

[baum7154]

Thanks for the offer, however my limited time is spent on optimizing the CHANCE PI and spider coils so I don't think I could get to it any time soon. In my mind it does not hurt any thing to have a fast coil on the detector. It only means that the coil allows you to look for targets sooner than a slow coil. The sample delay is really up to the detector electronics package.

Have you looked at any of the handheld scopes on EBAY? Some might be pretty good for detector work.

Dan