Coil for the Barracuda and SurfPI

[gryse]

Hi

So i've tried and I think failed at my first attempts of winding a coil for both the barracuda and surfPI

The coils made were a 190mm basket weave with .63 enameled copper. Resulting in L = 153uH This was connected to the surfPI and does (semi) work, could detect at depths of up to 15cm on small coins.

The other coils.
190mm fastcoil with .5 enammeled copper resulting in L = 159uH
250mm coil with o.5 pvc wire resulting in L = 257uH

So the question: Can I use these coils or should I rather just dump them?
I've read that the Barracuda coil should be 450uH and 1.5ohm and I believe the surfPI should be about 250uH?
Then I have a scope to use but do not know that or where to measure and what it should look like... Is there a video or tutorial somewhere? It seems that I should "Just tune the dumping R with a scope, but no one explains how to do this... This is if I should keep the coils...

[Tepco]

Better try to add more turns to existing coils. Even with lower inductance, 150-160uH this will work only coil peak current and battery consumption will be unnecessarily high. Larger one with 257uH is OK. Good side is, lower inductance coil will tend to be very fast, but to exploit this careful adjustment is needed, minimal delay and damping resistor, it certainly need to be changed slightly. Scattered somewhere on this forum, adjustment procedure is posted many times, including animated .gif of scope picture during adjustment, but interestingly no one compiled “tutorial”, adjustment or troubleshooting guide, always solved on case to case basis.

[gryse]

Hi Tepco

Thanks for the reply. I will then try to find whatever I can on the forum and compile the tutorial as my first contribution

[Tepco]

I find couple of scope pics posted time ago on some other forum, will repost here with explanations, can be helpful.

[Tepco]

I posted this eons ago somewhere, but can be useful to repeat. Test point is pin6 5534chip, preamp output. Connect scope, set to 1 or 2V\div, horizontal scale 5-20uS, adjust trigger etc to get image similar to what is below, interesting part is negative going spike after TX pulse end. Usually you will get something similar to surf1 picture, only duration of this part may vary slightly depending of coil type. Now, varying damping R value (use stereo 2x220 or 470R potentiometer and some 150R fixed resistor in series) observe waveform change. At too low value, decay time will increase, coil is overdamped. Increasing resistance, waveform will change, at very high value ringing will appear like on surf2 picture, this is underdamped. Somewhere between, optimal adjustment is reached, with fastest decay and no ringing, like on surf 3. Only, this shortest period can be different, this was some very fast coil design, 8uS as on picture is hard to get, ordinary bundled coil will be 25-30uS, faster designs maybe 10-15 or so. One regular trick is to partially unerdamp coil, like on surf4 picture, detector will work fine with this and last uS can be squeezed out. In essence this apply to all detectors, only on some, waveform can be inverted.



With this fixed, next is minimal delay adjustment, first sample must start as soon as possible. Easiest way is to monitor integrator output (first half LM358 after analog switch, pin 7 on silverdog PCB). Will be about zero, or less than volt in any direction at longer delay setting. Now decrease delay, pot and trimmer and observe voltage change, at shorter delay it will start to rise, at one point will reach some 3.8V saturation, chip is unable to output more, and detector will stop responding. Increase delay a bit and this is speed limit for coil used, few hundred mV or so less than saturation point. This output is very sensitive, less than mV of change will trigger audio when metal is detected.

[kenny444]

Hello Tepco can I ask please about surf3.jpg. Is the image almost perfect with the offset voltage as near to zero as possible. Thank you. kenny

[rankx0]

With this fixed, next is minimal delay adjustment, first sample must start as soon as possible. Easiest way is to monitor integrator output (first half LM358 after analog switch, pin 7 on silverdog PCB). Will be about zero, or less than volt in any direction at longer delay setting. Now decrease delay, pot and trimmer and observe voltage change, at shorter delay it will start to rise, at one point will reach some 3.8V saturation, chip is unable to output more, and detector will stop responding. Increase delay a bit and this is speed limit for coil used, few hundred mV or so less than saturation point. This output is very sensitive, less than mV of change will trigger audio when metal is detected.

I have a problem with Surf PI Pro, I have managed to fix coil damping, but after that, turning delay trim doesn't affect LF358 pin 7 voltage. It is always 3.8V. I watched signals with an oscilloscope, and all the pins of 4066 / 4093 are active. I really have no idea what could be wrong, so if anyone can help, please do.
Here are the oscillographs from both ICs.

[rankx0]

I have changed LM358N and CD4093, also tested delay trim and R25, they are all OK. Could it be problem in CD4066?

[rankx0]

And yes, finally, replacing 4066 made thing work.

[surfdetector]

I posted this eons ago somewhere, but can be useful to repeat. Test point is pin6 5534chip, preamp output. Connect scope, set to 1 or 2V\div, horizontal scale 5-20uS, adjust trigger etc to get image similar to what is below, interesting part is negative going spike after TX pulse end. Usually you will get something similar to surf1 picture, only duration of this part may vary slightly depending of coil type. Now, varying damping R value (use stereo 2x220 or 470R potentiometer and some 150R fixed resistor in series) observe waveform change. At too low value, decay time will increase, coil is overdamped. Increasing resistance, waveform will change, at very high value ringing will appear like on surf2 picture, this is underdamped. Somewhere between, optimal adjustment is reached, with fastest decay and no ringing, like on surf 3. Only, this shortest period can be different, this was some very fast coil design, 8uS as on picture is hard to get, ordinary bundled coil will be 25-30uS, faster designs maybe 10-15 or so. One regular trick is to partially unerdamp coil, like on surf4 picture, detector will work fine with this and last uS can be squeezed out. In essence this apply to all detectors, only on some, waveform can be inverted.





With this fixed, next is minimal delay adjustment, first sample must start as soon as possible. Easiest way is to monitor integrator output (first half LM358 after analog switch, pin 7 on silverdog PCB). Will be about zero, or less than volt in any direction at longer delay setting. Now decrease delay, pot and trimmer and observe voltage change, at shorter delay it will start to rise, at one point will reach some 3.8V saturation, chip is unable to output more, and detector will stop responding. Increase delay a bit and this is speed limit for coil used, few hundred mV or so less than saturation point. This output is very sensitive, less than mV of change will trigger audio when metal is detected.

These are very good example images for reference.