surf pi 1.2: building my first detector

[geoscash1]

It Possibly Erased the Chip.

I believe so Gary, that's why I asked about that..

[vasileios]

Hi all,

I'm assembling a surf pi kit from silverdog. I have soldered all the components to the board and made a test coil. However its not working. I hope you experts can help me troubleshouting. I already did some measurements with my DMM but I don't have an oscilloscope. I did some soldering and electronics before but this is my biggest "electronics" project so far. I only have basic knowledge as my previous projects were arduino based and the major work was coding. I hope you guys can help me figuring out what is gone on.

power consumption with coil: 91 mA
power consumption no coil: 30 mA

U2 pin4: -4,7V
U2 pin6: 0,423V (this is the lowest I can get by turning the offset)
U2 pin7: 5V


U6 pin1: 1,3V
U6 pin2: 2,53V
U6 pin3: 0V
U6 pin4: -2,35V
U6 pin5: 5V
U6 pin6: 2,7V
U6 pin7: 1,3V
U6 pin5: -4,7

U1 pin1: 1,6mV
U1 pin2: 6,5V
U1 pin3: 12,2V
U1 pin4: 12,5V
U1 pin5: 12,6V
U1 pin6: 6,5V
U1 pin7: 6,5V
U1 pin8: 8,3V

Thank you in advance

Hi,

I built two detectors from SILVERDOG and work perfectly. Bought only the PCB's and procured the parts by myself.

take care to the solders, a special soldering iron with fine tip is required. I used MPSA14 instead MPSA13
regards

[kasperfish]

did you tune the two blue trimmers about to get best performance?

Hi chemelec,

yes I tried turning the offset and delay trimpots without any benefits in sensitivity. I'm building the detector in the living room so maybe I have to go outside where there is less interference. I'll do some tests tomorrow and report them here.

[kasperfish]

So I powered up the surf pi this evening. I cleared all metal objects from the table and tried to get 0V at the opamp out by turning the offset trimpot. I was unable to do this. The minimal voltage I could achieve was in the 400mV range. I guess this is not good enough. I'm using a ~300 microH coil with ~2.1 ohm resistance. I'm using the damping resistor (387 ohm) supplied within the kit. Any ideas more then welcome.

grtz

[Tepco]

You wont be able to adjust zero with coil connected. Disconnect coil and try again. Even then, this is not true zero offset adjustment, transmit pulses are still present, but never mind, not fairly critical at all. For true offset adjustment you need to disable pulses, easiest by temporarily disconnecting MOSFET drain (middle pin) or removing transistor, but no need to mess with this. Using oscilloscope this is routinely adjusted for zero baseline voltage without removing parts.


Focus on minimal delay adjustment, reduce trimmer resistance up to the point when detector stops responding, and then increase just bit above, this is minimal delay for given coil. You will notice how delay setting affect detector sensitivity to different objects Longer delay is less sensitive to small and good conducting objects. From this point, only way to increase sensitivity is to build different coil type, some “fast coil” design, discussed here many times, capable for shorter delay, then adjust minimal delay and damping resistor value for optimal operation with given coil. For this, you will need either oscilloscope, or lot of time wasted if you try to adjust empirically.

[kasperfish]

Hi,

I tried adjusting the offset to get 0V at TP2
with coil: 461mV

when I disconnect the coil the voltage drops a little
no coil:442mV

When I read through the surf pi threads I see that most people even can achieve a negative voltage @ U2 pin6 (TP2). I guess they did their measurements with an O-scope or with DMM without transmit pulses.

With no coil connected my voltage range is between min 442mV (turned trimpot until click) and max 695mv.
I replaced U2 (NE5534P) and the voltage @ TP2 dropped to min 278mV (trimpot turned until click).

I know the coil is very important but I reckon that I should get better sensitivity with the coil I'm using now. There are numerous examples on youtube with good sensitivity using a similar unshielded coil made of enameled copper wire. I understand that the coil specifications are in direct relation with the damping resistor but how do I determine the correct resistor value for a specific coil, what is the relationship? I did read that replacing the damping resistor with a variable resistor (pot) is an option but I guess the ultimate answer is again using an O-scope? I'm thinking of buying one (bitscope smallest version) but since they are not that cheap and I'm an electronics noob, I didn't decide yet. I'm starting little by little to understand how the surf pi works but I reckon that owning an O-scope will make this process easier even if I never worked with such a device before.

I can tweak the sensitivity by turning the delay trimpot. However maximum sensitivity for a 2 euro coin is only 2cm with minimal audio response (barely noticeable). For large metal objects with good conductance (soda can) I can have an audio response from about 20 cm. Indeed I can tweak the sensitivity for specific objects by adjusting the delay but I'm unable to get higher sensitivity then the distances mentioned.

[Tepco]

You have some serious malfunction in your detector, range is unusually low. Normally, with typical coin, like E coin, detection range must be to about coil diameter, even with ordinary coil, faster designs can make difference in detecting some small items, jewelery etc, but this is too bad. Offset adjustment alone is not a major issue, but here some other related failure is indicative.


Most typical mistake can be misplaced component, most probably wrong resistor value installed in wrong place. Metal film resistors used in kit, with multicolored bands can be easily misplaced, for example 1K, 100K or 1M resistor is different in just one color band. Carefully check (desolder one end and measure if needed) resistors around U2 chip, R8,9,11, also check D1,2 for proper orientation. Problem with misplaced component is, if you install one wrong component from kit in wrong place, another component will end up in wrong place in the circuit automatically, producing multiple failure, very difficult to troubleshoot “remotely” this way. With scope this is usually matter of seconds to find, but even without, if you figure out one mistake, another one (another misplaced component) will be obvious immediately. Nice thing about kits, total component count is constant. Naturally, if other factors, like keeping coil on table with metal parts in it (generally, keep any metal at least 2 coil diameter away) or faulty component is not involved. From pictures you posted nothing can be concluded, due to camera color balance system, can be misinterpreted even with naked eyes, check this first and report results.

[kasperfish]

Hi Tepco,

I double checked and triple checked the resistor values. I did this by comparing the color codes with the resistors from my unassembled kit (I bought two because I was expecting troubles). I measured the actual resistance before soldering the resistor in place. I'm pretty sure all resistors are correctly placed and had the correct resistance before soldering (maybe a resistor started malfunctioning afterwards?). The diodes are oriented the correct way as well: the band marking on the diode matches the orientation of the marking on the board. D1 and D2 are oriented in opposite direction (D1 black band left; D2 black band right). D3 and D6 are oriented also according to the markings on the board.

I also double checked the values of the polyester film capacitors. All are placed on the correct position. However I received two types of 100nf capacitors with different text on it (10 and 0.1 100-). Both types measured the same capacitance (+-100nf) so I used them interchangeably.

Maybe I shorted something out while touching the board. For example when I touch the mosfet (iron part with hole in it) I get an audible signal. I discovered this when I checked if it was getting hot. Should I give up this board and start building my spare kit? I mean if the only solution is desoldering all resistors, capacitors, etc and test them if they still do their job... this is a lot of work. Not that I'm lazy but this doesn't sound very efficient.

[kasperfish]

Hi all,

after reading trough the forum and verified that all components where 100% certain on the right place. I start replacing the ICs one by one. It seemed that U7 was not working properly. After replacing this part I was able to get ~0V @TP2. I knew this was a good sign and tested the detector with a 50 cent euro coin... whoehoe +-20 cm. I'm so happy after all those hours reading and testing. I'll try to make a new video with the performance. Thank you all (especially Tepco) for the kind help.

[kasperfish]

Short video of the improvements.

[kasperfish]

Hi all,

I have an other question I prefer to ask it in this topic rather then creating a new one for each question I have. Who knows how many questions will follow

So I was thinking about the on off switch. A reed switch seems a very good option to keep the housing waterproof. I wonder what type of reed switch I should use? On ebay there are plenty to choose from. So what type of reed switch is fine for switching a surf pi 1.2 with a 12v power supply.

My second question is somewhat different but also involves the power supply. I want to add an arduino nano to the board. What would be the best for its power suply? Tap into the 5V on the surf pi PCB or connect the arduino to the battery with its own voltage regulator? I know it will depend on what I want to do with the arduino and how much current it consumes. So in fact my real question is how much extra can we load the 5V power line of the surf pi? Thinking a little further, maybe its possible to answer this question myself. Please follow me to see if I make some mistakes...

When I look to the data sheet of the 78L05 http://pdf1.alldatasheet.com/datashe...NEC/78L05.html I can read that the output current is in excess of 100mA. My detector consumes about 60mA so I have 40mA to use for other stuff. Is this correct?

I'm sorry to ask these noob questions but I have no one else to learn this stuff from. I'm doing my best to learn as much as I can on my own but sometimes a little feedback can be really helpful.

grtz

[Tepco]

Few answers: If you want to use RRED relay to power on detector, use largest possible one you can find. Small relays can latch due to high inrush current during large 1000uF cap charging period. For small REED you can add some extra circuit, one MOSFET and one resistor do provide actual power switching, controlled by small REED. Time ago, I used some specific high voltage REED tubes, some 5cm long, and somewhat larger magnet on belt around casing.


About powering extras, detector consumption is some 60mA, but actual current needed from 78L05 is much smaller. Lot of extra stuff can be powered with it, but not recommended if not necessary. Better use original board regulator from 12V, to avoid mixing digital and analog part power, also possible decoupling problems.

[kasperfish]

thanks for the info (again)! I'm waiting for parts to arrive so meanwhile I made a few more coils. My best sensitivity for a one euro coin is almost 30 cm with a flat spiral coil (inner diameter 13.5cm; outer diameter 21cm; pvc insulated wire, see picture). I noticed that the initial beep, when powering up the device takes a lot longer to fade out compared to the normal bundled enameled copper wired coil. Can someone explain why this happens? I'll try to update with more pictures/video this week.

cheers!

[vineeee]

Hi all,
i build my first metal detect Surf pi 1.2 . It works (get detection about 20-25 cm for hammer) but not the way it shuld. I have some issues with trashold potentiometer. Potentiometer is good i checked with ohm meter, but has no functionality if i unplug them is the same, i can tune it up or down detector works the same all the time. With multimeter i measured U3 pins, i will post this later because i have it on other computer, but for sure U3 pin 1 does not change with metal at coil or not. Can anybody quick identify the problem ? Mybe bad soldering, i only soldering at back site of pcb, but later i look to pictures and realized that PCB is two sided, on front not much but also some. Now i do not know if mybe this is the problem.

My coil measured with RLC meter:
L = 300 microH
C = -86 microF
R = 2 ohm

I bought PC scope i get nice pulse for offset (U2 pin6 set from 0.21 to -0,1, if up or down pulse disappear, but detector still works. Mybe some little performance is lost but still works, mybe someone explain me why?
Sorry for my not perfect English.

Thanks all for help!

[vineeee]

U3 no metal:
1 -0,72
2 0
3 0
4 -5,1V
8 +5,4V
7 0.82V
6 -0,08V
5 -0,08V

U3 big metal "hammer" at coil

1 -0,72
2 0
3 0
4 -5,1V
8 5,4V
7 4,1V
6 -0,04V
5 0,23V