Multiple frequencies good, explain this if you would

[49er12]

You no it all sounds great , running multiple frequencies at the same time. However folks think about this say your equinox 800 running on the 5 frequencies at same time sounds awesome, except please consider your sweep speed and separation, ground conditions for this to actually Get the most of this information in a second or two, my opinion there is no way In heck considering what I just mention You will get the proper information needed , and if so this machine must have an hyper speed processor to give the amount of information in that second or two, I’m no scientist but this machine would cost more if the processor really worked this fast and most importantly ACCURATE. my opinion to much room for error when bragging about multi frequencies, they better work exactly like suppose to or what’s the sense. I hope you all understand what I mean. You sweep in 2 seconds, ground conditions, separation and depth, I’m more interested in accuracy, not saying that can’t work in tandem but processing speed I question.does anyone no what type of processor is used in these so called multi frequencies machines and do you think I’m correct in some fashion. Thankyou

 

Amazon Forum Fav 👍

Minelab PRO-FIND 35 Waterproof Pinpointer Metal Detector for Experts with Tone ID & Adjustable Sensitivity - Grab it through Amazon!

[hound]

I'm trying to think of a way to explain this to a non-electrical engineer. It's a signal processing problem. You're severely underestimating modern microcontrollers, and overestimating the complexity of the problem. Spread spectrum has been around since the the later parts of WW2. Transmitting multiple frequencies is easy. The receiving part is easy. The response in simple is a number for each frequency that represents signal strength. That's all done in RF on the signal side and measured by an analog to digital converter or ADC. You get a value for each frequency that then goes through their weighting system and spits out a value that tells the user interface if there's a signal and what indicator to give. It's just a series of math problems. That's a simple problem that digital signal processors were created to solve. That can work on audio, video, radio, or other problem domains. It's built to do math in hardware very fast unlike a normal processor that would do the problem in a number of small steps taking up CPU cycles. Modern microcontrollers can have built in DSP extensions which make them very good at analyzing signals quickly.

Unlike a desktop CPU that's doing countless tasks constantly a microcontroller has a much more dedicated job. In the case of a metal detector. It'll update the user interface, and I include sound as part of that, read signal values, and put them through weighting faster than your brain will perceive any pause. If it's using a modern ARM with DSP extensions it's probably idle most of the time waiting for the relatively slow radio chain to give it a new signal value, and then once in a while updating the user interface for the glacial slow sack of meat using it. It has zero problems doing this.

Transmitting and receiving multiple frequencies is trivial, kids toys do that these days. The difficult problem for a multi-frequency detector is making sense of the signals you get. The processing power is there it's the development of the equations/algorithms to read and interpret that data that takes time and a lot of testing to get right.

Something like the AT Max is using a relatively low powered microcontroller, only as fast as it needs to be to process a single signal chain and update the interface. I doubt even the Equinox is using a very high end microcontroller part. Any updatable detector is likely using an ARM microcontroller with built in USB on the chip and self update capability so you're having to flash the firmware with a dedicated programmer.

There are more recent parts that even incorporate machine learning into a low cost microcontroller that's only a few dollars. You could possibly have a detector you could teach to work on a particular target in hundreds of different scenarios and have it ignore anything else.

 

[49er12]

So it’s the micro controller you stated low powered, so considering iron, ground mineralization, pull tabs and everything else in that close proximity of a coil sweep , you would or wouldn’t admit it’s relatively impossible to react to all that and then some in that second or two of a coil sweep over such a small area correct, I’m trying to understand from your prospective. Is the micro processor, chip or controller capable of responding accurately in just a short notice or you saying also the equinox being more expensive Isn’t using a high end part either to justify the cost and accuracy, thankyou sir please understand from my end not being a engineer either, thankyou

 

[49er12]

Yes it’s like you said above and my concerns also, it’s not the transmitting and receiving the issues, it’s how the detector in multi frequencies can determine all that information on a constant moving device as in motion detecting Yes your point , that’s my question can it perform in such a short manner accurately when in constant motion

 

[Jason in Enid]

So it’s the micro controller you stated low powered, so considering iron, ground mineralization, pull tabs and everything else in that close proximity of a coil sweep , you would or wouldn’t admit it’s relatively impossible to react to all that and then some in that second or two of a coil sweep over such a small area correct, ......

just, just stop. Modern computer chips can perform over a billion calculations per second. The problem of modern detectors isnt about processor speed, its about the engineer's ability to understand the complexities and create a program to analyze and interpret.

 

[hound]

I'm saying it doesn't need a high end part to respond to all of that. All of those things are processed in exactly the same way very quickly. The sounds and VDI you get are just the results of those calculations. It's plenty fast and powerful to do the job it needs to do. Ignore the cost bit as inexpensive doesn't mean it isn't powerful. It's more a concern when designing electronics for manufacture. Modern microcontrollers are incredibly powerful for their size and cost. The microcontrollers they use can process all those signals with ease. It really comes down to how the detector was designed. They specify how fast it needs to process and fix those numbers in the firmware that are then loaded onto the microcontroller. It's like a governor on a car. They may have a part that can go super fast but they limit it.

We even have control over the speed that it processes a new batch of signals. That's what the reactivity is. You can have it processing really quickly which will unmask things in iron, or cycle slower to get more depth but have things masked as it's going to display the predominant signal. Many detectors work on those fixed cycle times rather than as fast as possible because the detector coil takes a bit of time to react. By adjusting the time window that it can react you can tune the detector's performance. The actual workings gets pretty complex and my understanding is basic in the detector realm. I do understand signal processing though. Dealing with 5, 10, or a hundred frequencies is an easy problem compared to things like the signals coming into modems for your internet.

It's easy to react to everything you said with appropriate microcontroller choices, but that's not the only factor. There can be limits in software as well based on how they do things. Things like reactivity settings are one of them.

The limits in detectors is more how good we are at writing the software to analyze those signals rather than any speed or processing constraints that aren't imposed by the people writing the software than the ability of hardware to do them.


Simple answer: There are no hardware limitations to do what you're asking about. In fact it's trivially easy. It's the software that often limits things.

 

[hound]

So it’s the micro controller you stated low powered, so considering iron, ground mineralization, pull tabs and everything else in that close proximity of a coil sweep , you would or wouldn’t admit it’s relatively impossible to react to all that and then some in that second or two of a coil sweep over such a small area correct, I’m trying to understand from your prospective. Is the micro processor, chip or controller capable of responding accurately in just a short notice or you saying also the equinox being more expensive Isn’t using a high end part either to justify the cost and accuracy, thankyou sir please understand from my end not being a engineer either, thankyou

Not low powered as in slow, low powered as in optimized for power efficiency. The microcontroller they used is going to be more than sufficient for the task it needs to do.

 

[Charlie P. (NY)]

Which shines further? A 75 watt laser or a 75 watt incandescent bulb? On the other hand, which would you rather try and read with in a dark room? Each has an advantage depending on the use.

Not quite a proper analogy for detectors; but I believe a good single frequency detector is better than a poor multi-frequency detector. ALL the wavelengths a handheld detector use are 6 MILES or more long wavelengths (10km to 30km), so for the first 18 inches of that 380,000 foot long wave it probably makes less difference than some would have us believe.

When it might matter is if one frequency has nearby interference. Then a detector that allows you to step up or down in frequency is worth having. Some single frequency detectors do this (I have used it on my F-75).

 

[jmc24]

I have owned a Whites DFX, Whites V3i, Minelab Explorer and have owned both the Equinox 600 and the Equinox 800. I still own two 800s, an XP Deus, a Teknetics G2+ and a Minelab SDC2300.

Simultaneous multi frequency operation has been greatly improved in the signal processing, target separation and recovery speed departments since the early Minelab FBS1 detectors.

Minelab has never said that all five frequencies are being transmitted, received and processed simultaneously when using any one of the eight search modes on the Equinox 800. That is part of the hype problem that Minelab marketing apparently likes to engage in. I would guess two or three are being used at once in each of the search modes and from what I understand they are being processed using nano technology and nano sized measurements which can take two or three different frequencies, calculate an optimal ground balance point between them and calculate the ferrous/non-ferrous ratios of targets faster than instantly. So, the brilliance of the software engineers and their ability to implement their concepts functionally is what has really improved in my opinion.

I have hunted an Equinox side by side with the DFX, V3i and Explorer. They are simply put ---- SLOOOOOOW---- compared to the Equinox. They still detect very well and the Explorer can really go deep in a sparse target environment. Put them in thick iron or aluminum trash for very long...................why? The Equinox is right up there near the best detectors in the world for that kind of hunting. Is it the best--no. Is it very good however in thick iron or modern trash---yes.

The problem I have with the Equinox Multi IQ is keeping up with it myself. The target information it provides and the speed at which it does it is pretty amazing.

 

[Jason in Enid]

Which shines further? A 75 watt laser or a 75 watt incandescent bulb? On the other hand, which would you rather try and read with in a dark room? Each has an advantage depending on the use.

Not quite a proper analogy for detectors; but I believe a good single frequency detector is better than a poor multi-frequency detector. ALL the wavelengths a handheld detector use are 6 MILES or more long wavelengths (10km to 30km), so for the first 18 inches of that 380,000 foot long wave it probably makes less difference than some would have us believe.

When it might matter is if one frequency has nearby interference. Then a detector that allows you to step up or down in frequency is worth having. Some single frequency detectors do this (I have used it on my F-75).

Wow, I think every part of what you wrote was wrong! lasers and light bulbs are not even close to being a proper analogy. Personal opinion on single freq isnt an argument. 10km wavelength?!!!! Are you high? Metal detectors are using KILOHERZ, aka thousands of cycles per second. You cant transmit a wavelength longer than the length of your antenna. 6 miles? Your last comment is completely negating the entire advantage of what multi-freq can do.

Lastly, although you didnt outright say it, I get them impression you believe that a MF detector must divide the power of a single freq across the number of freqs which is wrong.

 

[Charlie P. (NY)]

According to the calculator I use 7 KHz has a wavelength of 140,571.42857142858 feet. Or 26 miles. 30 KHz is 32,800 feet. Or 6.2 miles.

Frequency Wavelength Calculator

[- removed -]

 

[Jason in Enid]

According to the calculator I use 7 KHz has a wavelength of 140,571.42857142858 feet. Or 26 miles. 30 KHz is 32,800 feet. Or 6.2 miles.

Frequency Wavelength Calculator

Jerk.

First off, I'm not someone afraid to learn new things. I did research the wavelength calculations and you were correct. As incredible as the number seems to someone who doesnt study RF engineering, 7khz does indeed have a peak to peak distance of roughly 26 miles. So on that point I was wrong. Everything else, I stand by. Name calling NEVER improves your argument.

 

[TheGreenBoy]

Maybe we shouldn't bother how the detector works - unless one wants to design or build one- but rather how it performs. Horse sense only could save the buyer from commercial "improvements" that doesn't really make a difference. Do you sincerely belive the multi frequency machines performs so much better than good single frequency detectors. I don't think so, i would appreciate your oppinion and experience.

 

[Charlie P. (NY)]

Neither does throwing someone under the bus - especially for things I "didn't outright say".

Visible light and detector transmissions are both electromagnetic waves. They just occur at different frequencies. (And, yes, light also shows particle properties, but that's not the topic here).

I take back by last comment [I edited it].

Carry on and hunt safe.

 

[Jason in Enid]

Maybe we shouldn't bother how the detector works - unless one wants to design or build one- but rather how it performs. Horse sense only could save the buyer from commercial "improvements" that doesn't really make a difference. Do you sincerely belive the multi frequency machines performs so much better than good single frequency detectors. I don't think so, i would appreciate your oppinion and experience.

The answer, as usual is "It depends". The minelab MFs are doing so much more than any single freq can. Not only is it looking at target ID return across multiple frequencies, its also constantly analyzing the ground response and removing that aspect from the signal to reduce to signal-noise ratio. I've been in this hobby since 1982, and used a lot of different machines. My finds exploded immediately after picking up a ML FBS detector. Now, if you are talking about hunting isolated targets in a very mild soil, then sure, a single freq can holds it own against a MF. When you start adding more and more trash targets and ground minerals, the MF begins to outperform. Minelabs also completely changed HOW signal returns are analyzed going all the way back to the first FBS and no other company has ever been able to duplicate it.

 

[TheGreenBoy]

Interesting observation Jason. I have noticed modern machines bahaves much better on the beach and mineralized soil, though i am not shure this is due to multifrequency or just more modern design in general. As fare as VDI goes, i do not see a significant imrovement in its reliability since CZ3D, neither depth vise, but this is my oppinion. The recovery speed of modern detectors improved significantly, compared to older digital ones, and almost meets the analog machines. The controlers and ADCs are getting faster every day........

Yes, MLs may be doing so much more, but are they performing so much better? Or are they just another VLFs, comparable to Garrett, Nokta, XP....