Frequency Domain and Time Domain relationships

[vbeeeks]

To tame the smoke I've got to tame the pulse window to 3rd multiple of L/R so as not to avalance the FETs and detonate the Caps ... effectively to tame the flyback tidal energy which comes rushing in ... and then to break the waves by raising the barrier synchronously in a interfering manner such that wave's energy unpacks or demodulates itself in a smooth fashion ... oh no this is not a poem....

[Aziz]

Wasting power was the day before yesterday.
Recycling energy back was yesterday.
Ultra super duper boost-mode TX is today.

Comeon guys, can ya see the forest for the trees?

Aziz

[Aziz]

BTW guys,

I will publish the new TX when I have the honest promise from all of the patent-trolls not to patent (steal) it.
Deal?

Patent-trolls don't have the integrity to invent it.
Patent-trolls don't have the required high MQ (madness quotient) to invent it.
Patent-trolls don't have the required zero-greed condition to invent it.


Or should I publish it so patent-trolls can't patent it anymore?
Aziz

[mikebg]

There is another advantage: it greatly simplifies power supply design. So is it really an incompetent design, or just another way to do things that works perfectly fine in some applications? Can you name a popular commercial detector that has this kind of search coil?

Carl, many times I use the term "incompetent design" in this forum, which means that the project does not achieve the purpose of my hobby - the highest possible sensitivity of RX and efficiency of TX. If a ham radio designer would design a metal detector, it will seek to achieve this by making at least five things:

1. To maximize the modulation index of TGT signal so that it almost looks like a SSB modulation. This is achieved through suppression of AIR & GND signal in RX input. The easiest method to achieve this is to use induction balanced search head. I showed in the forum that IB is best suited for all types of metal detectors. If I see a project, even if it is BFO, which is not used IB search head, I call it incompetent design because I can easily improve it and get a much better sensitivity with induction balanced coil.
I can not call Howard Rockey (Bugwhiskers) incompetent designer. I like the project QED, but I think it is an incompetent design to the point where someone thought to involve induction balanced search head instead the monocoil.

2. Maximize SNR (Signal to Noise Ratio). At metal detectors, the SNR is not determined by thermal noise and interferences because the synchronous demodulator suppresses them. Furthermore, the noise and interferences are significantly weaker than the parasitic modulation of AIR & GND signal, which is obtained due to the movement of the TX coil relative to ground. If I saw a circuit diagram of TX for CW metal detector, which has no P-I-D controller to suppress parasitic amplitude modulation, for me it is an incompetent design. Minelab and Garrett are using a P-I controller, which gives good results.
Now I remembered a design of Eric Foster. He puts a big TX loop on the ground and seeks with RX coil inside. So there is no modulation induced by the movement of TX coil.

3. Maximal efficiency of TX circuit. This is the main goal of the designers with my hobby, though our TX only occasionally is powered by the car battery. However, metal detectors are powered only by battery. Because energy wastage in the conventional PI detector, The designers with my hobby call it "incompetent designed wideband metal detector". The analysis with Frequency domain shows what wavwform of TX current is most effective for wideband metal detector.

4. AGC of RX preamp. In an incompetent design, the operator should manually adjust maximal possible amplification at actual ground conditions.

5. Carrier recovery for phase reference. The RX should operate as wireless device. A narrow band RX for metal detector is incompetent designed when uses a wire connected to TX to obtain phase reference. That needs impossible phase stability for correct discrimination.

[Davor]

Mike, I mostly concur with your conclusions for most things, yet here I think your radio bias got better of you at points 2, 4 and 5.
2. Invariance of S/N is an illusion due to the sub optimal Rx coil design of most detectors. Physical reality is that Cu wire has only a limited conductivity, and given the opportunity to operate in a non-resonant mode with impedance matching preamp input noise would prove you that there is a thermal noise, and that a synchronous demodulator has very little influence on it. Semi-resonant configurations certainly may qualify as incompetent designs as they screw phase response.
BTW, AM at Tx is a bit less of a problem compared to the variation of harmonic content. You are not wrong, but the culprit you are pointing at is a bit different.
4. AGC is very useful for maintaining modulation properties of a received signal, except for FM where limiters are used to remove the AM component from the signal. Surprisingly metal detectors have only a little use for amplitude, hence a soft limiter is the best option. Not AGC. Designs lacking a soft limiter or compression of any kind are certainly incompetent designs.
5. Independent Rx design would require some sort of Costas loop, but even then such a system would not be much different from a current solution with phases referenced to Tx. Furthermore, if we decide to employ some sort of frequency hopping (which is an intentional exaggeration) a system coupled to a Tx would be much better off due to the problems of loop filters' lag. That's why I consider incompetent all the designs that employ PLLs for phase shifting.

[Qiaozhi]

Carl, many times I use the term "incompetent design" in this forum, which means that the project does not achieve the purpose of my hobby - the highest possible sensitivity of RX and efficiency of TX.

"Incompetent" is the wrong word to use. It means that someone is unable or not legally qualified to perform specified acts or to be held legally responsible for such acts. You cannot say that a circuit or idea is incompetent, but you can say that it is incompetently designed. However, this implies that the designer of the circuit is not fit to do his job. I'm sure that this is not what you actually meant to say, so you should be using the word "sub-optimal" instead.

Have you proved your points 1 to 5 by actually building a detector based on these principles?

[vbeeeks]

I don't see anybody contributing to Moodz' bifilar invention which is s promising system and which defines a new day in metal detection as a whole (front end and back end) - the last two schematics show the genius, though I'm waiting for the parts to drop by, most of the people who have successfully duplicated the work may be more enlightened by now to the potential it owes to the md community. The community must understand the signal chain and its optimized benefits as everything is in open --- need to learn and decipher - yes its part radio technology, part car ignition, part radar, part audio, part nuclear, part power electronics but most of it is signal engineering - analog and digital. Moodz on the way to make MRI portable - a wild guess. Imaging is his keyword now since discrimination is his yesterday!!!!. Moodz by the way your 50us sample pulse width and that of BW's was 100us, I presume you sample post Tx starting at 4L/R-rising to 4L/R-decaying, what are the pros and cons of either topology? see the graf on what I mean.

[mikebg]

I also posted incompetent designs

"Incompetent" is the wrong word to use. It means that someone is unable or not legally qualified to perform specified acts or to be held legally responsible for such acts. You cannot say that a circuit or idea is incompetent, but you can say that it is incompetently designed. However, this implies that the designer of the circuit is not fit to do his job. I'm sure that this is not what you actually meant to say, so you should be using the word "sub-optimal" instead.
Have you proved your points 1 to 5 by actually building a detector based on these principles?

George, I use the term "incompetent design" in sense that the project does not achieve the purpose of my hobby - the highest possible sensitivity of RX and efficiency of TX. If a ham radio designer with primitive knowledge like me can improve operation of a project with easy modifidcation, for me that means incompetent design. For example, I think both circuits in my post #26 above are incompetent designed because I can simply place the RX coil in induction balance. Nothing changes for TX coil. It keeps to receive the AIR signal, which you call flyback, and receive large GND signal, which causes parasitic modulation of TX. However the RX coil placed in IB will receive very suppressed AIR&GND signal. Then I can increase gain of preamp and decrease sample delay.

I will give another example for a great company C-scope. My primitive knowledge is that sensitivity of a metal detector depends of possible gain of RX preamp. It should be increased until saturation. When Ivconic showed the circuit diagram of model 1220B, I can't believe that it is not an error. An incompetent designed preamp operates with gain less than 9:
http://www.geotech1.com/forums/showt...560#post135560

Finally, I will give as a bad example my designs. I posted in the forum two powerful CW TX circuits which operate with high efficiency.
http://www.geotech1.com/forums/showt...702#post152702
http://www.geotech1.com/forums/showt...772#post152772
However the use of these circuits in a metal detector is an incompetent design according my hobby, because "QRP amateur radio" means "Use minimal TX power!". No need of powerful TX for metal detectors because this can't increase the SNR. I showed in the forum how seems the output of RX preamp when a wideband metal detector is competent designed.

[Qiaozhi]

George, I use the term "incompetent design" in sense that the project does not achieve the purpose of my hobby - the highest possible sensitivity of RX and efficiency of TX. If a ham radio designer with primitive knowledge like me can improve operation of a project with easy modifidcation, for me that means incompetent design.

Yes - I understand what the word means to you ... but it is incorrect in this context. You should use "sub-optimal", or "could be better", or something similar. Just because a circuit is not absolutely perfect, it does not mean that it was incompetently designed. You are being too harsh.

As your first language is not English, I am trying to guide you into using the right words. Some people could get very upset when you refer to their designs as incompetent. The word "sub-optimal" is much more benign. This says to the reader: "I know this design works, but [IMHO] there is a better way of doing it".

[ODM]

I'd like to point out that several of the designs that are being criticized here are rather aged - often by a decade or more. As such it would be fair to assume they don't represent the "state of the art" any longer.

They're often recycled into hobby designs, particularly the more known and less complex ones, with the intent to improve one particular aspect of the design instead of taking a radically different design path.

Also, several "optimal" solutions are often traded away from the manufacturing point of view. Reducing components, design work and assembly work, and taking component tolerances into account leads into less than mathematically optimal designs

[WM6]

My primitive knowledge is that sensitivity of a metal detector depends of possible gain of RX preamp. It should be increased until saturation. When Ivconic showed the circuit diagram of model 1220B, I can't believe that it is not an error. An incompetent designed preamp operates with gain less than 9:

However the use of these circuits in a metal detector is an incompetent design according my hobby, because "QRP amateur radio" means "Use minimal TX power!".

Mike, what if it shouldn't be increased until saturation?

We do not assemble metal detectors for air test only, but to face with real terrain condition, with coil practically on soil, which can greatly change circuit working conditions (working points).

This is not whole comparable to your QRP amateur radio antenna which is working constantly in air in relative same conditions (or it is out of tuning).

[ODM]

That's a pretty aged design from 1984 or so. What opamp is used, is the reason for low gain broadening the opamp's bandwidth or trying to have smoother phase response for all TX harmonics? The TX oscillator itself doesn't seem amplitude stabilized so TX clipping may have been an issue.

[Gwil]

Reality has a way of disagreeing with theory sometimes - I'd like to see mikebg's answer to the question in the last line of post #51.

[Qiaozhi]

Reality has a way of disagreeing with theory sometimes - I'd like to see mikebg's answer to the question in the last line of post #51.

In theory, there is no difference between theory and practice. But in practice there is.

[Gwil]

Back in 1954 there was a nuclear bomb test codenamed "Castle Bravo". The scientists who designed the device predicted an explosive yield of about five megatons. When they pressed the button there was an enormous bang, a large part of the island where it had been set up disappeared, and a huge cloud of fallout was released. It turned out that the actual yield had been FIFTEEN megatons. That story is enough to make anyone a little skeptical of theory alone.