48Views9Replies

Author Options:

Pin-pointer metal detector front end not working. Answered

Hi,

I breadboarded the pin-pointer metal detector and am having trouble with the front end. pcounterA reads 11-14 on the serial port monitor.

See previous question on instructibles reproduced below:

_________________

Asked by jont4e 2 days ago

Hi, I bread-boarded your circuit for just the end coils to start, and loaded the sketch. It seems to be pulsing the coil OK but the leds aren't responding nor the buzzer. I uncommented the serial port code and it shows ledthreshold1 always at .97, pcounterA at 12, and calav1 at 81. Is this OK? (The other reported variables are always 1000.) I wound the end coils with some bifilar wire I reclaimed from a microwave HV supply transformer. I assumed that was OK? Do you have a flow chart or description of how the code works so I can troubleshoot? Any help or hints would be greatly appreciated. Jon T

0

Answer from TechKiwi... 2 days ago

The pcounterA value should be much higher 200 to 800 at least. My advice is to focus on one channel. Try reversing the polarity of one of the coils as this often is the problem as the pulseIn function needs to have a clean leading edge to trigger the timing process. Ensure your breadboard connections to the coil are short and well grounded. If this does not work direct message me with a link to some photos of your setup and I will try to assist. I'm just building a second unit so your timing is fortuitous as I can provide some test measurements.

_______________________________

Attached are images of the front end scope traces and a shot of the breadboard. Coils are off screen to the left.

Swapping coil polarity did not seem to change any waveforms.

Hope someone has some suggestions I can try.

JT

9 Replies

user
jont4e (author)2018-04-03

Thanks for the hints. None worked, so I wound a second sense coil separate from the bifilar- wound original sense coil. It worked much better. I got a pcoounterA reading on the serial port of 409 and the calav1 value was 339. Moving a large steel washer close to the coils caused the pcaounbtrA value to drop below calavg1 and the Arduino pin 10 started pulsing.

Looks like it is working!!

I guess tight coil coupling is not a good thing in this case.

Thanks for the help.

JT

Select as Best AnswerUndo Best Answer

user
TechKiwiGadgets (author)jont4e2018-04-03

Yes, I have experimented extensively with double wound coils and separate coils provide the best outcome. Apologies I realise now that you said bifilar earlier.

Also am about to publish a new instructable with 3D Printable case as well as a modified circuit.

Should be up in the next few weeks.

Select as Best AnswerUndo Best Answer

user
jont4e (author)2018-03-30

Hi,

Another picture with more info. Got 4 channels this time with better time resolution.

See notes on image.

Hope this helps you diagnose my woes.

Regards,

Jon T

Select as Best AnswerUndo Best Answer

user
TechKiwiGadgets (author)jont4e2018-04-02

Ok thanks I think I see the problem. There is some residual "ringing" in the TX circuit which is triggering multiple pulses. Put a 150pf ceramic capacitor across the TX coil. This will dampen the ringing. Make sure the 330pf capacitor is across the RX coil as well as this will further reduce the problem incidence. Let me know how you get on.

Select as Best AnswerUndo Best Answer

user

Also in some cases the power supply from Laptops via USB cables can be quite noisy.

Try disconnecting the Laptop from mains and run the cicuit off the Laptop battery to see if this is causing issues.

Either way it is good practice to put a 0.1uf capacitor across the power supply to help address noisy PUS issues

Select as Best AnswerUndo Best Answer

user
jont4e (author)2018-03-30

Hi again.

I just noticed on your eco-friendly detector, the transmit coil has a flyback diode across it as apparently does the perfboard shown in tyour pin-pointer instructible. Does the diode make a difference?

Jon T

Select as Best AnswerUndo Best Answer

user
jont4e (author)jont4e2018-03-30

Sorry, I just noticed you replaced a BJT with a FET and built-in diode.

Dummy me!

JT

Select as Best AnswerUndo Best Answer

user
TechKiwiGadgets (author)jont4e2018-04-02

Yes I found the FET performed better and had a diode built in so was a good fit.

Select as Best AnswerUndo Best Answer

user
jont4e (author)2018-03-31

Just noticed after on-line research, that the pulsein() function relies on the clock frequency. I'm using the 3.3V, 8MHz version of Pro mini. Could that be my problem with pcounterA values being consistently 11-14?

Jon T

Select as Best AnswerUndo Best Answer