What newer DIY pulse metal detector project can I build instead of AVT-5196 or JABEL 267/297, with PCB artwork I can etch myself?
AVT-5196 and JABEL 267/297 are old Sandbanks-derived projects, so if you want something newer you should look at amateur designs that move more functions into a microcontroller, but that does not guarantee better detector performance [#17679613] A concrete newer option mentioned in the thread is Felezjoo.pi, described as an advanced pulse detector with many settings and iron discrimination, although the build is somewhat problematic [#17689770] Keep in mind that PI discrimination is limited and works more like iron reject than true VLF-style discrimination [#17692766]
I have dug through many threads on this and other forums. I'm looking for a project for a pulsed metal detector for self-assembly. The AVT-5196 or the Jabel J-267/J-297 are often scrolled over, and these projects are a couple of years old. Has anything newer/interesting been developed? I have no intention of ordering a kit, I can make the board myself, so assembly diagrams without any markings for thermo-transfer to the laminate are welcome. I don't have exorbitant requirements, this will be the first "detector".
Simple detectors are either those with frequency deviation detection, or pulse detectors such as the sets indicated. I saw a pulse detector design somewhere without a microprocessor. Digital circuits played the role of the pulse generator. Is this something my colleague is talking about?
Attachments:
Lucaslab schematwykrywacza PI.JPG(217.15 KB)
You must be logged in to download this attachment.
AVT 5196 or JABEL 267/297 appear frequently. These projects are already several years old.
They come from a common source of the Sandbanks detector which is several decades (from 1979) http://www.geotech1.com/cgi-bin/pages/common/...page=metdet&file=projects/sandbanks/index.dat There are many derivatives of the detectors that are changed from the original only when the original parts are no longer available, Jabel includes a microcontroller that replaces the complex timers on the gates, the parameters are not influenced by this, but it simplifies the diagram.
Do not expect that the company making the kits will refine its structures and release new versions - it does not affect the sales results, while the company producing the detectors has to release newer and newer models, regardless of whether it can be introduced essential improvements or not, competitors have new models and you can't be left behind.
If you want something newer, try to look for amateur constructions, there are those in which many more tasks have been transferred to the microcontroller - in line with modern trends, but whether hypothetically greater digital processing capabilities translate into better detector parameters - it's not that simple, because the final effect also affects whether the programmer was able to use these new possibilities and whether these new possibilities, in the face of limitations caused by soil parameters, coils or analogs, give a chance to improve anything related to detection, or just a cool looking display . Besides, there is no other method of comparing amateur constructions than building a few and comparing them in practice, there is no point in debating the diagrams, many Sandbanks derivatives, due to the poor availability of amplifiers without internal compensation (formerly most), use amplifiers with a smaller GBW, but is it really worsens the parameters - I guess no one checked.
In my opinion - build one in popular projects, if you want and time, you will build new ones and improve both the detectors and your skills, the first project will never be perfect, so there is no need to look for the "perfect" scheme.
Added after 31 [minutes]:
Interesting fact, I looked at the AVT store, from a few PI detector projects, only the AVT2874 remained on offer, which contains an obvious error in the schematic, some "transformer" (it cannot be called an electronics) converted BJT to a MOSFET without knowing how to control it correctly, hence on start, everyone who assembles the system correctly will have a problem with the lack of sensitivity and overheating of the coil and the transistor. This is how the market works - someone was not a good electronics engineer, but he had an idea - he changed the diagram, wrote an article and got a few hundred bucks, someone else sent for printing and production with errors, because time is money - checking someone else's projects costs money, you can earn more by working less, in the end, the better product AVT5196 (less reworked Sandbanks) was replaced by the worse AVT2874 (more reworked Sandbanks) people buy, assemble, throw a defective device into the corner (few beginners will be able to fix errors in the project) and the business is spinning anyway.
If you want something new, I recommend the Felezjoo.pi detector
I have had one of the most advanced impulse detectors for quite some time. It has a lot of settings, especially iron discrimination. The construction is a bit problematic but still manageable.
I have had one of the most advanced impulse detectors for quite some time. It has a lot of settings, especially iron discrimination. The construction is a bit problematic but still manageable.
You can give some specific pointers to this detector, because as far as I know, no discrimination can be obtained in impulse detectors. Someone once proposed pseudo-discrimination, but somehow it didn't work. In general, from the way the PI detector works, this is impossible.
Discrimination in PI is possible and has been around for a long time, but of course it won't work as well as in VLFs. Nevertheless, I am satisfied with this equipment. There are several videos on YT with this detector. You should find the discrimination activity there too. Generally this works so that in the 'Iron reject' mode the detector briefly signals the iron as the object enters the range of the coil, but by holding the coil over the iron target, the detector is silent. The opposite is true for color. In addition, the LCD shows "ferro" or "non ferro" according to the sound indications. In every situation, the signal strength can also be seen on the LCD, which makes it very easy to track the target. Signal strength works even when iron discrimination is on.
Unless you have HEX without a counter The detector is difficult to calibrate. You have to carefully select the parameters of the LF357 amplifier and sometimes you need to buy even a dozen or so pieces to choose the right one. But there are ways too The detector itself requires quite a lot of experience with the PI system, but once you get to know it, it can amaze you.
I asked if you have any specific bearings, such as pcb, schematic, tuning. The YouTube video says nothing. When you say discrimination you are probably thinking about discrimination, which is not the same thing. Discrimination completely eliminates the reaction to iron.
I have all the materials but in English. I will. Hex will unfortunately be limited to 300 actuations. In this case, it can be called discrimination because iron can be omitted in the way I wrote above. It's just a bit different than in the VLFs.
Attachments:
Nowy Dokument tekstowy.txt(15.15 KB)
You must be logged in to download this attachment.
Kopia PCB1.jpg(107.11 KB)
You must be logged in to download this attachment.
Keys.gif(39.54 KB)
You must be logged in to download this attachment.
fuse.jpg(106.8 KB)
You must be logged in to download this attachment.
FelezJoo-PI-Schematic.jpg(439.24 KB)
You must be logged in to download this attachment.
FelezJoo-PI - Schematic.pdf(138.63 KB)
You must be logged in to download this attachment.
FelezJoo_PI_v3_35.hex(58.13 KB)
You must be logged in to download this attachment.
felezjoo pi - part list.pdf(395.07 KB)
You must be logged in to download this attachment.
2.png(182.17 KB)
You must be logged in to download this attachment.
1.png(269.12 KB)
You must be logged in to download this attachment.
You're welcome. You will probably fire the detector 200 times during the launch I propose to bring out the kanda socket, you will be able to upload soft without removing the prock.
Hello, I know it's been 2 years since the last post, but I haven't found anything newer, so I have to hang up here. I've been looking for something simple, but with discrimination, for several weeks now. It may be PI, it may be VLF, but it would be nice if it could be tuned without an oscilloscope. I do not have an oscilloscope within my range, and even if I did, I do not know if I would be able to operate it, because I am rather an amateur. I did a few BFO projects, a few PIs, got some practice with PCB etching and reading schematics, but it's probably not enough to do something difficult. I was looking for some relatively simple projects on the forum, and I don't think there are such VLF's, and I haven't found any detailed PI with discrimination. There is quite a lot about ALL, but apparently the tuning of the coils is a pretty good job, so despite the simplicity of the scheme, I have to let it go. I found a lot on the net about Terminator 3, a few Tesoro, Sniffer, etc., but all of this is too advanced for my capabilities and they all require a background that I do not have (mainly this oscilloscope). I don't want to buy anything ready-made, because I'm even moderately interested in walking in the woods or the beach, I'd rather fight with a soldering iron and a PCB. I also do not want to do anything to earn money, I already have something to live on, and electronics is just a hobby, so I will not hurt anyone. If any of my fellow professionals recommended some amateur VLF or PI with discrimination for alignment with a common meter, it may be without frills and even with a poor range, I would be very grateful. Regards.
PI detectors, as a rule, do not discriminate, I read that someone who combined with the analysis of the beginning / end of the pulse, but this is not a way for everyone, nor a way to build a well-functioning and simple device.
IB (VLF) detectors, as the name suggests, are based on coil balancing, do you expect a design that does not require it?
The oscilloscope is a very useful piece of equipment, some circuits cannot run without it. A used analog oscilloscope can be purchased for about two hundred. In the case of the VLF detector, there is a chance that it falls within the acoustic band and a sound card will be used instead of an oscilloscope.
PI detectors, as a rule, do not discriminate, I read that someone who combined with the analysis of the beginning / end of the pulse, but this is not a way for everyone, nor a way to build a well-functioning and simple device.
I wasn't sure, but I guess it is. I definitely do not know enough to plan something like this by myself.
Quote:
IB (VLF) detectors, as the name suggests, are based on coil balancing, do you expect a design that does not require it?
Maybe it can be balanced with a frequency meter alone? It's a simpler gauge, and easier to access.
Quote:
The oscilloscope is a very useful piece of equipment, some circuits cannot run without it. A used analog oscilloscope can be purchased for about two hundred. In the case of the VLF detector, there is a chance that it falls within the acoustic band and a sound card will be used instead of an oscilloscope.
I think even the vast majority of VLFs touch the acoustic band, the ones I read about mostly in the 5-25kHz range. I am considering the use of a sound card, I have already figured out how to convert the input to different ranges. Are the integrals suitable? Or should it be external? In your opinion, is a single-channel oscilloscope enough? Or do you need a 2-channel? For now, the Russian Terminator 3 seems to be the easiest one: it has quite passable documentation and instructions to start it, and it's based on Tesoro. Have you seen him? Have your own opinion? Regards.
✨ The discussion revolves around DIY pulsed metal detector projects, specifically the AVT 5196 and Jabel J-267/J-297 models, with inquiries about newer designs. Participants share insights on the limitations of older models, the potential for microcontroller integration, and the challenges of achieving discrimination in pulse induction (PI) detectors. Suggestions for alternative designs, such as the Felezjoo.pi detector, are provided, highlighting its advanced features and settings. Users express a desire for simpler projects that do not require extensive calibration or specialized equipment like oscilloscopes. The conversation also touches on the availability of resources and documentation for various designs, including the Terminator 3 and a PI detector with discrimination capabilities. Generated by the language model.
TL;DR: About 200 is what a used analog oscilloscope costs, and “PI detectors, as a rule, do not discriminate.” This FAQ helps DIYers choose between AVT‑5196, Jabel J‑267/J‑297, Felezjoo PI, and VLF options, with build tips and gotchas. [Elektroda, jarek_lnx, post #19790229]
Why it matters: It speeds your first successful build and helps you avoid flawed schematics and false expectations about discrimination.
Are AVT‑5196 or Jabel J‑267/J‑297 still worth building today?
Yes. They remain solid beginner PI choices derived from Sandbanks. Jabel swaps complex gate timers for a microcontroller without changing performance. Do not expect kit vendors to publish improved revisions. Build a popular design, test, and iterate. “The first project will never be perfect.” [Elektroda, jarek_lnx, post #17679613]
What is the origin of AVT‑5196 and Jabel PI circuits?
Both descend from the Sandbanks PI detector, first published in 1979. Many later versions only changed parts for availability. Jabel moved timing to a microcontroller to simplify the schematic. That change eases assembly rather than improving sensitivity or depth. [Elektroda, jarek_lnx, post #17679613]
Which PI kit should beginners avoid due to schematic errors?
Avoid AVT2874 as published. Its BJT‑to‑MOSFET change lacks proper gate drive. The result is low sensitivity and heating of the coil and transistor at startup. Choose AVT‑5196 if you want a kit‑style PI without this issue. [Elektroda, jarek_lnx, post #17679613]
Can a PI detector discriminate iron at all?
Basic PI detectors lack true discrimination. Some designs analyze the start or end of pulses, but that path is complex and uncommon. It is not a route to a simple, reliable build. “PI detectors, as a rule, do not discriminate.” [Elektroda, jarek_lnx, post #19790229]
How does Felezjoo PI indicate iron vs non‑ferrous?
Use Iron Reject mode. Iron gives a short entry beep, then silence when you hold over the target. Non‑ferrous behaves oppositely. The LCD labels targets as “ferro” or “non ferro.” It also shows signal strength to aid pinpointing. [Elektroda, MADMAX1978, post #17692766]
What makes Felezjoo PI challenging to build?
Calibration is demanding. You must select a suitable LF357 op‑amp, often from many pieces. The detector rewards experience with PI systems. “There are ways” to succeed, but expect effort during tuning.
How‑To: Select LF357 for Felezjoo
Buy multiple LF357 parts for binning.
Test each candidate in‑circuit and note stability.
Yes. Shared HEX files may include a counter that limits about 300 actuations. You will need to reload firmware after reaching that limit. Project documentation mentioned is in English. Plan your workflow around this constraint. [Elektroda, MADMAX1978, post #17695390]
Any reflashing tips while bringing up Felezjoo PI?
Expose a Kanda ISP header so you can reflash without removing the MCU. Expect many programming cycles during bring‑up, even around 200 per builder reports. This saves time and reduces mechanical stress. [Elektroda, MADMAX1978, post #17698380]
Can I tune a VLF without a dedicated oscilloscope?
Often yes. VLF detectors can sit in the acoustic band, so a PC sound card can substitute as a basic scope. If you prefer hardware, a used analog oscilloscope costs about two hundred. Buy one for circuits that truly need it. [Elektroda, jarek_lnx, post #19790229]
Is Terminator 3 a good first VLF with discrimination?
One builder found Terminator 3 the easiest to start. It draws on Tesoro concepts and has workable documentation. It still needs careful setup, but suits a patient hobbyist willing to learn. Review its startup guides before etching. [Elektroda, Majster109, post #19790922]
Where can I browse DIY metal detector schematics and ideas?
Check curated hobby resources. The miriskateley.com collection lists many do‑it‑yourself detectors and notes. Use these for learning and inspiration, then verify before building. [Elektroda, Krzys55, post #17683134]
Any open PI with discrimination to download?
See the “2nd Chance PI” archive. It provides a PI design with discrimination features for hobby study and builds. Read the documents and code before assembly. [Elektroda, pawol, post #21226732]
What’s the best learning path for my first detector?
Build one proven project and compare results in the field. Then improve both the circuit and your skills on the next build. Skip endless diagram debates without measurements. “There is no point in debating the diagrams.” [Elektroda, jarek_lnx, post #17679613]
How do I avoid coil overheating on first power‑up?
Do not copy AVT2874’s MOSFET drive as published. The flawed gate control can overheat the coil and output transistor and cut sensitivity. Choose AVT‑5196 or correct the gate driver before testing. Monitor temperature during bring‑up. [Elektroda, jarek_lnx, post #17679613]
The detector is difficult to calibrate. You have to carefully select the parameters of the LF357 amplifier and sometimes you need to buy even a dozen or so pieces to choose the right one. But there are ways too 
The detector itself requires quite a lot of experience with the PI system, but once you get to know it, it can amaze you.
