Surf PI 1.2

[Wasp]

Hi there! Long time lurker here that finally had some time to built my first metal detector.

I've read every single thread under the Surf PI keyword from the forum in the last 2 days and i've saved up a few main threads if i get into trouble later on. I ordered silverdog's Surf PI 1.2 kit a few months ago and i've soldered everything together yesterday, i haven't powered it up yet since i was hoping to build the coil today. I read/know about all the common hurdles and i'm avoiding them swiftly (i have a degree in electronics, i think i'll be alright) .

Surf PI threads are all over the place though and some have outdated problems that could get confusing for noobs .

So today i was about to build the coil and to make sure it was around 200~300 uH like most of threads suggest i used Qiaozhi's coil calculator only to find out it gives me inductance values 3x larger than i expected.

For 250 mm inner radius, 0.5 mm wire thickness and 21 turns it gives me 777 uH (?) when i was expecting somewhat around 230~250 uH. I used another online coil calculator and gave me 237 uH. Am i doing anything wrong? Is Qiaozhi's coil calculator not suitable for this project?

The best result i've seen was from Nezocas with 21 turns, 250 mm and 0.4 mm wire thickness. Since i have 0.5 mm enameled wire, i was hoping to correct (if needed) the number of turns/diameter to reach the same inductance as him.

Also, i know the board is thru-hole plated but do you guys fill them with solder just to make sure? Does anyone had/know any problems regarding this?

Can i get a little help with the coil values?

Cheers,
Wasp

[Qiaozhi]

So today i was about to build the coil and to make sure it was around 200~300 uH like most of threads suggest i used Qiaozhi's coil calculator only to find out it gives me inductance values 3x larger than i expected.

For 250 mm inner radius, 0.5 mm wire thickness and 21 turns it gives me 777 uH (?) when i was expecting somewhat around 230~250 uH. I used another online coil calculator and gave me 237 uH. Am i doing anything wrong? Is Qiaozhi's coil calculator not suitable for this project?

Did you really mean 250mm inner radius, or 250 mm diameter? A 250mm radius coil has a total diameter of 500mm, which is 19.7".
Assuming you actually meant 250mm diameter, then the radius is 125mm which gives an inductance of 342uH with the calculator.

Surf PI threads are all over the place though and some have outdated problems that could get confusing for noobs .

Yes, I agree. That's also the reason why so many members have problems building what is essentially a fairly simple and straightforward design.
If you feel up to it, you could do a great service for the Geotech community and write a Build Document for the Surf-PI. I would do it if I had the spare time, which I don't.
Have a look at the Minipulse Plus documentation as a guideline -> Minipulse Plus REV-C ... if you feel up to the challenge.
The complete Minipulse Plus thread is here -> Minipulse Plus

Also, i know the board is thru-hole plated but do you guys fill them with solder just to make sure? Does anyone had/know any problems regarding this?

There's no need to solder both sides of the board, as the plated-through holes will take care of the connection.

[Wasp]

Did you really mean 250mm inner radius, or 250 mm diameter? A 250mm radius coil has a total diameter of 500mm, which is 19.7".
Assuming you actually meant 250mm diameter, then the radius is 125mm which gives an inductance of 342uH with the calculator.

Oh, that was embarrasing ! I'm used to work with ID and OD as inner diameter and outer diameter almost daily, my brain never read the radius . The value seems about alright now. My apologies!

Yes, I agree. That's also the reason why so many members have problems building what is essentially a fairly simple and straightforward design.
If you feel up to it, you could do a great service for the Geotech community and write a Build Document for the Surf-PI. I would do it if I had the spare time, which I don't.
Have a look at the Minipulse Plus documentation as a guideline -> Minipulse Plus REV-C ... if you feel up to the challenge.
The complete Minipulse Plus thread is here -> Minipulse Plus

As soon as i finish my Surf PI i'll give it a try . I do have some spare time in my hands.

In the meanwhile, i'll post some pics and update as i progress. I coiled a 250 mm diameter, 0.5 mm wire thickness and 21 turns.

Top: http://i.imgur.com/YDB4lLZ.jpg
Bottom: http://i.imgur.com/kFQhWX0.jpg
Coil: http://i.imgur.com/y9cVFzO.jpg

Cheers,
Wasp

[Wasp]

I made an elaborate reply but i think it triggered some spam filter or whatever and never showed up. I guess i'm struggling with this forum mechanics

[Qiaozhi]

I made an elaborate reply but i think it triggered some spam filter or whatever and never showed up. I guess i'm struggling with this forum mechanics

Since you are a new member (and had made only one post) your reply went into moderation because it contained external links. This is an anti-spam measure, and it was waiting for approval from Admin. As you can now see, your post has been approved.
You should be ok from now on.

[Davor]

Regarding coil calculations... the safe way is to calculate the length of wire in a calculated coil, and cut a meter or two more, wind it and unwind to the measured value. For 300uH coil I'd start with ~22m wire, wind a coil, measure, and unwind to ~19m and 300uH as it should be.

[Wasp]

So today i power it up, all voltages seem ok.

Dampening resistor warms a bit, but not enough to feel uncomfortable to touch.

Voltage regulator is bearly over room temperature.

Consumes 10 mA without coil and around 50 mA with coil.

TP1 and TP2 looked like this before some tweaking (TP2 was all squiggly, not sure why yet): http://i.imgur.com/SvLqemu.jpg

After regulating the offset to get TP2 to 0 Volt: http://i.imgur.com/x4o3mWd.jpg

Regulated the delay/tune and started testing a few object. Detects small coins at 10 cm, a small block of copper at 12 cm and an aluminium can at +20 cm.

The sound is really faint though. Volume doesn't seem to effect the sound at all. I had to put the speaker in my hear to hear properly. I was expecting it to be louder.

A lot of stuff was triggering the alarm. Touching the coil, moving the coil, a lot of random things, etc, i suspect the lab i was on has way too many electrical stuff working and affects the detector. Since i was a bit disappointed with the depth on coins, i'm going to try to tune it again tomorrow in a better environment.

Never thought it would be this hard to tune it, either that or the coil isn't too good. We'll see.

Cheers,
Wasp

[6666]

The sound is really faint though.

Did you connect it correctly, those little buzzers are polarized.
Mine overloads the buzzer if volume turned up too high.

[Wasp]

Did you connect it correctly, those little buzzers are polarized.
Mine overloads the buzzer if volume turned up too high.

Yeah, i noticed. It was polarized correctly. I also tried a PC motherboard speaker, same thing, very faint.

[Wasp]

Hi again,

So i retouched all the sound weldings with the hot iron and the buzzer and volume now work properly. It seems some bad welding was introducing high resistance to the circuit.

Pin 6 of NE5543 is dead on 0 Volt.

The alarm still goes off randomly every few seconds in bursts, sounds like morse code, even though i'm testing it in a solid wood table with as little around as possible. Even with the threshold at the minimum, it still goes off, but less frequently. I can detect a thin gold ring at 8 cm, 50 euro cent coin at 15 cm and a small block of copper at 18 cm when the threshold is at the minimum.

If i raise the threshold it's almost contuniously going off and everything i try to detect blends with the random beeps and gets confusing even though it's still detecting.

Tuning the delay anti-clockwise reduces sensitivity. Tuning the delay clockwise (like most of people report the best results) gets to an area where suddenly the sound stops. The only way to get any sound again is by raising the threshold more than half way only to hear a constant tone but the coil detects absolutely nothing. No matter how i regulate the threshold, it doesn't detect anything. I need to tune the delay anti-clockwise to regain detection and random alarm bursts every few seconds.

I believe my signal has some sort of noise, like the image i posted before of the oscilloscope and interferes with the detection/threshold. I have no idea what could be introducing this noise signal.

I'm going to have to use the oscilloscope on every IC leg and try to figure where it comes from.

Any ideas where to look first?

Cheers,
Wasp

[6666]

So i retouched all the sound weldings with the hot iron and the buzzer and volume now work properly.

Good work.

Test your MD far away from household noise, pc's tv's thermostats, etc etc.

[Wasp]

Good work.

Test your MD far away from household noise, pc's tv's thermostats, etc etc.

Thank you 6666!

I tested it outside and the false alarms have reduced quite a lot. The chirps it still makes once in a while can now be distinguished from the actual metal on coil by somewhat the tone, and volume.

Never thought it would be this sensitive to noise from household electronics.

I tuned it the best i could but i still think the performance is lacking. I'm thinking since i used thicker wire, it's resistance might be too low. I measure 1.7 ohm in my coil.

I still have plenty of 0.5 mm enameled copper to try more coils.

Any advice on coils for a better performance? More turns? Diameter?

Cheers,
Wasp

[Silver Dollar]

Well enameled copper is part of the problem. You want low capacitance via spacing
between wires and the enamel allows them to be close. Teflon covered wire is better
but you could make a basket type coil with the mag wire;

[Wasp]

Hello again,

I've followed Tepco's post and tried a PVC insulated cable coil. I wanted to do a 30 turn coil but i quickly ran out of cable and only made it to 25 turns.

Here is the second coil i've made: http://i.imgur.com/eIrzsbQ.jpg

And here are it's specs: 0.9 ohm, 25 turns, ID 125 mm, OD 205 mm, double 0.25 mm copper (wired in parallel).

Damping resistor gets warm and i can now use the full trimmer range of the delay with the best results close to the full clockwise regulation.

I'm using a thin gold ring as my smallest detection target and a 150 grams copper block as my biggest detection target (i'm assuming anything in between in size/volume can also be detected). It can detect them both at around 15 cm max which i think is an improvement from the first coil.

Well enameled copper is part of the problem. You want low capacitance via spacing
between wires and the enamel allows them to be close. Teflon covered wire is better
but you could make a basket type coil with the mag wire;
[ATTACH]30937[/ATTACH]

Thank you for your reply.

It looks very interesting but a search for "basket coil" in the forum brings only indirect threads and it's hard to find a thread dedicated to the subject.

Do you know where i can go and read on the subject? I need some guide lines. How do i calculate it's inductance? Do i make an aproximation on it's diameter? Has anyone made a successful basket coil for this project? If so, what kind of specs should i go for?

Cheers,
Wasp

[Silver Dollar]

Here's some info I have found in various places;

I wound a teflon mono coil then reused the wire to make a basket (same length ~ 65 feet for 300 uh).