Surf pi detector-MOD

[Qiaozhi]

Hello Qiaozhi,

Thanks for your contribution. I also own all of your recently published books. They are a fantastic source of information, and I have already recommended them to other DIY enthusiasts. Inspired by your Arduino Nano project, I'm currently learning how to use the Arduino for DIY projects. The first new project ideas with the Arduino Nano and Arduino Mini Pro have already taken shape. New operating functions for these projects, such as a "one-button operation" with a display, have already been defined and just need to be programmed. The corresponding PCBs for these projects have already arrived from China.

This is an example PCB from one of the Arduino projects I'm currently working on.

Well done.
I love it when people find inspiration from my humble efforts and develop their own projects.

By the way, if you could leave a positive review on Amazon to the books you have purchased, that would be great.
There is some idiot who left a 1-star review on Amazon for the Arduino Nano PI project and for ITMD2, and also a 2-star review for The Voodoo Project. It's clearly the same person, but it makes the reviews look bad.
He also had a bit of a rant about the Voodoo Project for some reason. The attack seems personal.

Even Carl's amazing ITMD3 mega-book has received a 1-star review. Probably the same troll.

[scs-bg]

A problem with all pi detectors is that when working with a loudspeaker, electromagnetic feedback occurs. The reason is the high gain of the input broadband amplifier. This is avoided by not passing current through the loudspeaker coil before or during the reception of a signal from a gating switch. This is done in my modification.

[scs-bg]

I am not satisfied with the performance of a processor-based detector. In practice, analog ones give better results.

[GeoMax]

Hello SCS-BG,

However, a distinction must be made here between "digital" and "digitally processed signals." If the sample and/or TX pulse signals originate from a microcontroller, they are 100% identical to signals coming from an analog circuit. Therefore, the Arduino Nano Surf PI must be identical to the analog Surf PI because in the Arduino Nano project, the microcontroller generates all TX and timing signals.

[scs-bg]

I meant digital signal processing.

[Detectorist#1]

When you use appropriate operational amplifiers or enough fast microprocessors, the only difference is your knowledge for the real process.

[scs-bg]

Many people here need to read a little theory. Then they will know what they are doing. They will not argue about how they drew a circuit or how some super transistor or operational amplifier works wonders. The problem is not in the details but in the conversion of an analog signal into a digital one. It's like comparing the sound of a gramophone record with a compact disc... Let someone make a detector better than a C-scope or a Fischer 1266. I will congratulate him!

[Carl-NC]

It's like comparing the sound of a gramophone record with a compact disc...

That's funny, I have tinnitus so a CD sounds perfectly fine. But I can still hear those ticks and pops on a record that I don't hear on a CD.

There are things you can do in a digital detector design that are extremely difficult or even impossible in an analog design. Narrowband demodulation is easy in digital, very difficult in analog, and the usual wideband demods used in analog design can alias high-frequency noise into the signal. You can do far more aggressive filtering in digital and some very complex DSP. One amazing accomplishment in analog design was the analog ground tracking used in the late 6000 models, it's incredible they ever got that to work. Now with digital, ground tracking is easier and performs better. I like some of the old analog designs, but I could not argue that they will outperform the newest digital designs.

[Altra]

One of the biggest advantages I believe in digital SMF detectors is the multi-modes. With a Equinox you have a coin detector, relic detector, beach detector and a gold prospecting detector all in one compact package. Each mode could be equaled with a analog design but you have four or five detectors instead of one.

[scs-bg]

Sampling frequency What is the fundamental difference between analog and digital signals? In the nature of their sampling. One (albeit only somewhat) correct analogy can be made with the film image. One second on a film strip consists of 24 frames (discrete static images), but the human eye perceives this movement as continuous. This is exactly how a digital signal is structured - the continuous music stream is broken down into discrete points in time. Information about them is recorded in a file, and the playback device, reading the information about them, completes the sound wave in such a way that the listener perceives it as a continuous sound stream. The sampling frequency of most audio files that we currently have is 44.1 kHz, that is, there are 44,100 "key points" of information in one second of recording. At first glance, there should be no problems with the perception of this slightly broken signal, but most acoustics experts are categorical that it is precisely in the sampling of the digital signal that the main problem with its perception lies. It turns out that deceiving the human ear is not as easy as the eye.

[Detectorist#1]

Hi Altra,
Maybe the forum for PI detectors is not right place for write for advantages of Equinox VLF detector with microprocessor control in comparison with simple PI detectors without microprocessor or when this microprocessor is used only for generating of TX and sampling signals (and maybe audio signals). If you have information for PI solution (even with microprocessor control) capable to be so flexible as Equinox - please give to us this information (GPZ7000 is not capable to be so flexible even "different" modes)

[algan]

Hi algan,
No need to do modifications of the schematic (and the board) for using N-ch MOSFET switch. The same improvement for min first delay will be made with 500V P-ch transistor IXTP10P50P and with faster (and with lower noise) first OpAmp type MAX410CPA+.

Today I did a test with an irf9640 transistor and with an IXTP10P50P transistor and with an IXTP10P50P it doesn't work deeper and better, there's even noise with an IRF9640 it works calmly and deeper with a few cm difference I don't have a MAX410CPA to test and say what the difference is compared to the ICL7660.

[Detectorist#1]

Hi algan,
To see the difference, increase peek voltage from 200V to 400V with setting of the time of TX pulse with IXTP10P50P. Also reduce the first delay to 4-5us. Reduce the sample pulses from 40 to 20us and the values of capacitors in differential integrator from 220nF to 100nF. All this have to improve sensitivity to small objects with the same (or better) sensitivity for large objects.

[algan]

Hi algan,
To see the difference, increase peek voltage from 200V to 400V with setting of the time of TX pulse with IXTP10P50P. Also reduce the first delay to 4-5us. Reduce the sample pulses from 40 to 20us and the values of capacitors in differential integrator from 220nF to 100nF. All this have to improve sensitivity to small objects with the same (or better) sensitivity for large objects.

How many should they be please write exactly so I can replace them, and currently they are
Frequency ~616 Hz
Period (T) ~1.62 ms
High level ~1.58 ms
Low level ~0.05 ms
R1 is between pin 8 and pin 7 → R1 = 100 kΩ
R2 is between pin 7 and pin 6 → R2 = 3.3 kΩ
C is between pin 6 and ground → C = 22 nF

[Detectorist#1]

Hi algan,
First, all the changes that I recommend have to be made with help of oscilloscope!
With variation of the value of R2 (3.3K), you have to made the peek pulse on the coil to be near to 400V. Exact value depends of the parameters of the used coil and cable - this is the reason I not know the exactly value.
For reducing of the first delay to 4-5us, you need to change the value of R25 from 15K to 7.5K.
For reducing of the value of the two samples, you need to change the values of R26 and R29 from 51K to 24K.
C5 and C6 have to be reduced from 470nF (220nF in some variants) to 100nF Polyester type.
Also will be useful to reduce the the voltage gain of NE5534 from 1001 to 471 or 221 by reducing of the value of R9 from 1M to 470K or 220K. This will help to reduce the min first delay to 4-5us. For preserve the total gain of the MD, the gain of U3A (LM35, have to be increased respectively by changing of the values of R22 and C8 - from 100K and 100nF to 220K and 47nF (if R9 is 470K) or 470K and 22nF (if R9 is 220K). The settings of "TRESHOLD" Pot V2 will be changed also after these changes - this will be set experimentally.