Geotech

The CZ3D is a dead end design and financially uncompetitive. It is all-analog, 2 full-sized stacked boards with thru-hole parts, and has 21 trim pots; basically a production nightmare. Yes, we could surface-mount it but even then it's a really big circuit that still needs a big enclosure. Yes, we could digitize it, get rid of the trim pots, and shrink it substantially but that's called "an all-new design." The CZ design is best thought of as a single-frequency (5kHz) detector with the ability to cancel salt. The 15kHz component is weak and not all that useful for detecting targets. I once proposed a variable frequency design that works the same way but the user can vary the primary frequency from a low of 2-3kHz up to the gold range, maybe 50kHz. At lower frequencies the user could turn on salt cancel. The idea got no interest.

I suspect that sequential MF (BBS/FBS) and simultaneous MF (MIQ) end up about equivalent in their raw capabilities. The advantage sequential has is that the analog design is much simpler (no channelization filters needed) but with direct sampling (as with MIQ) all that becomes moot, it's just software. If you were to use direct sampling for FBS then I expect the software timing would become a major headache (it's not continuous), whereas it's simple for MIQ. Therefore, I think the evolution toward direct sampling is why Minelab is now using simultaneous instead of sequential. It's just the natural evolution of metal detecting. Better recovery speed would have helped 50 years ago but what mattered back then was depth and ground balance. Today, separation is the Hot Topic so everyone focuses on that. One day it'll be something else. Personally, I'm hoping for foolproof bottle cap rejection.

There is nothing inherent in BBS/FBS that prevents it from having fast recovery. There are 3 major elements to recovery speed: hardware filters, DSP, and how audio is processed. Hardware filters are easy to change (or eliminate, as the new designs do), DSP depends on the processor speed and what kind of filters you implement, and audio is just a matter of choice in how you generate it. Back when the Sovereign, Explorer etc were designed there was not the emphasis on recovery speed that there is now, most of the emphasis was on depth. And they used much slower processors, so fast recovery may not have been a realistic goal. But it could be done now.

Hello FLOT, I live in Republic, WA which is at -17° today and lots of snow on the ground, both poor conditions for detecting.

Yes, patents & copyrights are different but if NM violated copyright (by ripping and directly using ML code) then it certainly would have been mentioned in the complaint. The reason that a distributor is named at all is because this is a US patent; it only applies to an infringing product made in the US (which the Legend is not) or sold in the US (which it is). Because it's sold in the US, Minelab needs to state an infringing seller in their complaint, and they only need one. Assuming that this patent was not granted anywhere other than the US, Nokta could continue manufacturing the Legend and selling it anywhere except the US and that would not violate the patent.

Dutchman, your information on microcontrollers is about 30 years out-of-date. Modern uC's have incredible processing power and speed and the standard language used by all of them is C. Some people still program them in assembly but they are the laggards who would be unhirable in the open market because no one does that any more. Pimento, you are correct, a lot of the Bounty Hunter and Teknetics models are written in assembly because they are derived from legacy designs. For the TekPoint/FPulse and the new revised Bounty Hunter pinpointer I used C++, and I think they are the only modern designs we've released so far (32-bit uC & C/C++). All my other 32-bit projects are in C++ and the few 8-bit projects I'm doing (for my book) use C. PSPR, you are correct, at least concerning one manufacturer. All the detector companies that have released new models in the last 2-3 years are using C or C++. Back in college (when PCs were new) I knew a guy who had a Jive translator on floppy disc; you type in a sentence and it would give the Jive translation. Loosely based on the Jive talkers in the movie Airplane!. Using a bunch of precompiler #defines the same could be done to make C look like Jive. Probably someone has already done it.

There is no question that the lawsuit concerns the Legend's use of SMF and, in particular, how MF signals are demodulated. I downloaded the complaint from PACER and it's very explicit. Yes, it's possible to rip the code from an Equinox and then use that code, but it would be a seriously tall order and probably harder than just writing your own code. Also, that would be a copyright violation and the complaint makes no mention of it.

Most of the newer designs use ARM Cortex micros like the STM32 line (my favorite). Programmed in C or C++ (my preference).

It's the exact same PCB.

I assume Invision lets you put new registrants in moderation, meaning their posts won't display until you approve and remove them from moderation. Next step is to start building up a .htaccess file with spammer IP addresses.

That means they are likely using the new LE Audio standard. The nRF5340 is the same chip I was testing with. LE Audio adoption has been slow, so far I've only seen earbuds and not headphones. But they're coming, and then you'll have lots of options. One cool thing about LE Audio is that it streams left & right separately, so earbuds do not need a wire between them.

On the audio, AptX-LL is not part of Bluetooth, it is a custom Codec owned by Qualcomm and the chips are produced only by them. Bluetooth does support "standard" audio but the latency is too high. Starting with BT 5.2 there is a new low-latency audio called "LE Audio" and it is part of the BT standard. I've used a Nordic chip running LE Audio and compared it to AptX-LL and Garrett's proprietary wireless. LE Audio is dead even with AptX-LL which means that Qualcomm's AptX-LL will likely be discontinued. Garrett has slightly better latency but all 3 are low enough that you won't notice. Minelab only says the new headphones are "better than standard Bluetooth" which could mean they are AptX-LL, LE Audio, or proprietary. My guess is proprietary.

C59 & C60 need to match each other in value to 1% or less. Otherwise GB will be noisy. Reg was shorting C50 which does eliminate a high-pass filter stage. I've never tried it but see no downside to doing it.

There are 3 modes: Vibe only Beep only Beep+Vibe For each mode, there are 3 sensitivity settings.

F-Pulse has no volume control, it always runs at max volume. If there was anything I could do to make it louder I would but it is limited by the mechanical enclosure. Same with the vibe strength. I am utterly shocked that a factory reset would have any effect whatsoever on volume or vibe strength. No idea how that could be but if it worked, then good!

I had a QED (PL3) on loan for a while and, honestly, I was a bit disappointed with it. I considered it a little better than the TDI, but not as much as I expected. I found it to be noisy and temperamental, and the weird 2-step threshold adjustment was just awful. This particular unit didn't have the 3D printed parts I've heard about, instead I found the mechanicals to be perfectly reasonable. In any case, the Impulse Gold was clearly outperforming the QED in the tests I was able to do. Given that Garrett did a really nice job on the Axiom (way better mechanicals than we can do) I would expect to be south of the Axiom price. Some PI detectors have simpler designs that are cheaper to build. I designed the TDI-SL to sell for $999 MSRP and it could have been sold for $600 with good margins. Probably the Sea Hunter, even with that enclosure, is not expensive to build. The Impulse AQ (and Gold) are difficult to build, especially for FTP which is accustomed to mass assembly-line production. We are still struggling with the AQ, which is why it's been such a challenge to ship detectors. That's one of the key issues to address on the Gold before it goes into production.

The HaHa emoji is probably appropriate, too. That's what most folks think of when they think of FTP.

Hartley oscillator, which is pretty tolerant of the coil tap. But getting the 3 coils more closely matched may be more critical.

There is an "Impulse Gold" in development. I field tested it in June and it works pretty well, sees sub-grainers about the same as the SDC. There were some minor issues that needed addressing and FTP moves at glacial speeds so that's still in the works. Otherwise, there are other developments on the VLF/MF side but I have no exposure to those.

This is sad news indeed. I worked with Eric a lot and had hoped to visit with him one more time but then Covid arrived. Eric's willingness to share his knowledge is a key reason Geotech was created and also why I have spent most of my time focusing on PI technology. He was not only generous, he was a very kind man. I will miss him greatly.

I'll betcha I'm the only other person here who found this interesting. I had a White's BFO with a Triplet coil and was curious what made it a Triplet coil, but not curious enough to destroy the coil to find out. Thanks for the tidbit, curiosity satisfied.

Without knowing for sure what the mods were, it's just a guessing game. Since the TDI PCB is mostly surface mount trying out those guesses is difficult and potentially damaging. You really need to find a modded TDI and look inside.

Sorry, I don't know what was involved in the threshold mod. When I was designing the SL Reg suggested a modification to the design of the SAT amps that resulted in a smoother threshold, so I suspect this might be the mod he was doing to the TDI. However, he also had a mod for the SL whereby he removed the changes to the SAT amps, and I don't know why.

For those who forego headphones in order to hear snakes and such, I use "transparent" Koss UR40 headphones. They are "transparent" because they don't block outside ambient sound so you can still hear everything around you but the detector response is right in your ear. I've played around with those tiny speakers and tested maybe 20 different ones. I thought they sucked and would make for very poor detector audio, even with the speakers pointing at you. It's possible they could be made better with a proper acoustical chamber design (in the same way Bose can make a mundane speaker sound phenomenal) but that's not my expertise and probably not a skill with any detector manufacturer. So the answer is to brute-force it with a big speaker, and that means putting it where it will fit, usually pointing out the back.

MAP pricing was created back in the 1970s to deal with mega-dealers like Kellyco. It was intended to create a level playing field between them and the mom&pop dealers. You could not advertise a detector for less than MAP but you could sell it for less than MAP. To skirt the advertising restriction dealers often added a bunch of accessories "for free." Usually they were junk, but it's the illusion that counts. Detector companies usually have the policy that they will not compete with their own dealers. So if the company sells direct, it is usually at MSRP and not MAP. I think what may have happened is that the T2 had been listed at MAP and now it's been raised to MSRP. Typically MAP is 80-85% of MSRP. Their price is correct. This is a special edition "audiophile" version of the coil. It uses oxygen-free copper Litz for the winding, the housing is carbon nanotubes formed under the heat and pressure of the sun, the cable has double defluorinated Teflon coaxial cables using niobium-germanium conductors with a polymorphine SuperFlex jacket, and a titanium connector housing with beryllium contacts. It performs exactly the same as the standard coil but allows the owner to sneer at those with lesser equipment.