Let's made a PC-base metal detector with usb interface !!!

[Tinkerer]

Hi all,

I think you have to chew on this brilliant "invention" a bit longer. I don't want to tease you.
I want you:
- train you learning by doing & experimenting
- train your natural "pattern recognition system" (brain)
- train you thinking out of the box

You have to make dozens of variations of the parameters I have mentioned earlier to fully understand the design (as I did).

I'll tell you the missing link to the latest VLF Oscillator Detector Controller project too, to make it wide-band TX (multi-frequency):
Connect a capacitor between Tp1 or Tp2 to V- (V- being the choke connection point or the source of the mosfets). The capacitor value can be 2x - 10x of the value of the CTX.

All the magic is made by using the magic choke. Long live the magic choke!

Cheers,
Aziz

Thank you for posting the interesting TX variation.
Good you tell us about the need of trying many variations, my first try as not impressing. Obviously it needs the right timing with the right proportions between the C's and L's.

I will give it another run.

[ODM]

Careful about self quenching chirps. Those can be temperature sensitive if semiconductor parasitics are noticeable. Then again with TX current monitoring it'd be less of an issue. No measurement of tx current though? With an instrumentation amp setup like that it'd be easy. Always Itx -> Urx, why different now?

[Aziz]

Thank you for posting the interesting TX variation.
Good you tell us about the need of trying many variations, my first try as not impressing. Obviously it needs the right timing with the right proportions between the C's and L's.

I will give it another run.

Hi Tinkerer,

that was probably the reason, why it hasn't been "discovered" yet.
Don't give up! Try dozens of variations. Believe me, I have made hundreds of variations.
Important rules:
- mosfet may not avalance
- zero voltage switch (ZVS) rules may apply
- design is critical and can cause harm to the electronic parts due to high currents and improper use of the design
- boost-mode happens, when the voltage after the choke (with buffer capacitor) increases (above the input voltage)
- high current spikes at the capacitors may not happen (ZVS rule)

I'll post at a later time the full working variation spice files.
The transmitters are flexible in use and can stimulate large targets better.

Cheers,
Aziz

[Aziz]

I forgot one important tip:

Make the spice simulation for a longer time until the circuit operation stabilizes!!!
(50-100 ms or so)

Good luck in finding the magic.

Aziz

[Aziz]

Hi all,

I recommend to start with the standard TEM transmitter and move the big elcos (>1000µF) to the choke input connection and put additionally an unipolar (high voltage) capacitor after the choke output. Change the value of the unipolar high voltage capacitor value (larger than tuning capacitor and much much lower than the big elcos) until you get the following coil currents (see below).

Have you recognised, that the circuit resonates at two different frequencies? Flyback period (mosfet off) is a half cosine current wave form and the pulse-on period sine waves.

Number of radiant sine-waves (during mosfet-on):
0.5, 1.5, 2.5, .., etc: no boost-mode (power efficient)
0.5-1, 1.5-2, .., etc: boost-mode (for sine and half-cosines)
1-1.5, 2-2.5, .., etc: boost-mode (for sine waves)

Now some coil currents:










After you have understood the principle, you can continue with the CLASS-E TEM version. You get same coil currents. The optional capacitor reduces the mosfet voltage so it should not break-down. ZVS-rule with optional capacitor is important here.

Cheers,
Aziz

[Aziz]

Did anyone have time to look at the new transmitters?
Comeon guys, you are lazy.
Aziz

[PiTec]

Hi Aziz,

nice idea to use Class-E The first part of your riddle is not too difficult (post #1640), but I had not enough time yet to try the harder part (and that looks more difficult).

All this reminds me of some CPT (contactless power transfer) experiments that I did a few years ago. I used a series resonant tank circuit for TX and a parallel tank circuit for RX and was able to transfer a few watts over a distance of a few mm using two small spiral coils. For some more fun, I added a bidirectional data transfer by modulating the TX power and the load of the RX circuit (similar to what RFID tags do). I have to dig up the still existing bread board and the documentation. At least I have a photo of the test setup. There are three 100 Ohms resistors in parallel on the RX side (changing color already ) and the voltage after rectifying is 15.9 volts DC, i.e. 7.7 watts. The coil distance is maybe 5mm.



And the two coils:



I already posted a link on CPT in the Round/Square coils thread. Here it is again and some more:

Wireless Power Transfer in Loosely Coupled Links: Coil Misalignment Model
http://thesciencedude.com/projects/F...nt%20model.pdf

A Study of Loosely Coupled Coils for Wireless Power Transfer
http://thesciencedude.com/projects/F...eric%20wpt.pdf

Primary Current Generation for a Contactless Power Transfer System Using Free Oscillation and Energy Injection Control
http://manuscript.jpe.or.kr/ltkPSWeb...aspx?ppseq=506

Inductively Coupled Wireless Power Transfer With Class-DE Power Amplifier
http://www.s-lab.nd.chiba-u.jp/achie...APCCAS2012.pdf

This one sounds interesting, but looks like it‘s not freely available:
IEEE Xplore - Design and Optimization of a Class-E Amplifier for a Loosely Coupled Planar Wireless Power System
http://ieeexplore.ieee.org/xpl/login...number=5299023

And here is a well done document on class E amps, however, from Switzerland and in German:
Klasse-E-Endverstärker
http://web.fhnw.ch/technik/projekte/...unctional.html
http://web.fhnw.ch/technik/projekte/...bericht_P6.pdf

I hope the links are helpful. Now we need some more time to figure out how to use a Class-E amp for a detector TX …

Thomas

[Aziz]

Hi Thomas,

glad to see, that someone is interested in this magic & free design.
The shown basic schematics above are already working and one don't need a modification. Only the parts value and frequency parameters need a bit tweaking.

"Inductively Coupled Wireless Power Transfer (to targets *LOL*) With Class-E Power TEM." sounds good.

Well, I personally prefer the power efficient mode (0.5, 1.5, 2.5, 3.5, .. etc. radiant sine waves). This will stimulate the large and deep targets much better than a PI mode. Note, that during the sine wave swingings, the mosfet is switched fully on.

"Einfach, aber genial!"

Cheers,
Aziz

[PiTec]

Yep, I know you mentioned that the circuit is already working, so I should have said 'we need some more time to figure out the correct part values'.

Finding suitable values for only one cap in the standard TEM circuit is no problem, but tweaking 4 parameters (pulse frequency, duty cycle and 2 caps) is much harder. I guess it’s possible to calculate the values - maybe there is something in one of the CPT and Class-E amp documents.

Thomas

[mikebg]

TOOTHBRUSH DESIGN

Thomas, I have several toothbrushes containing wireless charged Akku. They are made in China much before papers cited in your post. Their design is made in Germany by TCM team before more than 10 years.
What you and Aziz wish to design? Unlike you, I will give other answer:

You and Aziz are trying to design an EFFICIENT TRANSMITTER FOR WIDEBAND METAL DETECTOR.

But you guys do not use the FREQUENCY DOMAIN - THE MOST POWERFUL TOOL FOR ANALYSIS AND DESIGN.

Why?

Here is the principle of TCM wireless toothbrush:

[Aziz]

Thomas, I have several toothbrushes containing wireless charged Akku. They are made in China much before papers cited in your post. Their design is made in Germany by TCM team before more than 10 years.
What you and Aziz wish to design? Unlike you, I will give other answer:

You and Aziz are trying to design an EFFICIENT TRANSMITTER FOR WIDEBAND METAL DETECTOR.

But you guys do not use the FREQUENCY DOMAIN - THE MOST POWERFUL TOOL FOR ANALYSIS AND DESIGN.

Why?

Here is the principle of TCM wireless toothbrush:

*LOL*

Of course, I'm using the frequency domain stuff. This is the reason, why I can use the simple transmitters above.

These two little "inventions" have been overlooked.
*LOL*

Aziz

[mikebg]

*LOL*

Of course, I'm using the frequency domain stuff. This is the reason, why I can use the simple transmitters above.

These two little "inventions" have been overlooked.
*LOL*

Aziz

Mein Herr, If you use FD, then explain the frequency characteristics of your targets, metallic junk and ground signal. What frequency spectrum should radiate your TX and how your RX should identify targets. If you have this information, the Frequency domain can show you in what Radiofrequency band should work your TX, what is the most effective form of TX current, how RX should calculate color coefficients of target.

Why "Schwachstrom Elektro Ingeneure" trained in Germany use ELF radiofrequency band for target identification at end of 19-th century? What waveform generates their TX ?

http://www.geotech1.com/forums/showt...704#post164704

[PiTec]

Mike, you can’t compare the wireless toothbrush charging with the CPT designs from the links. The CPT designs have much more power because of the poor coil coupling (several cm and more apart) – comparable with the problem that we have to excite a target below our coils.

The toothbrush chargers use a split transformer with quite a good coupling, driven by a sine oscillator of a few kHz (round iron core in the charging station plus air coil in the toothbrush that fits around the core).

Thomas

[Aziz]

Mein Herr, If you use FD, then explain the frequency characteristics of your targets, metallic junk and ground signal. What frequency spectrum should radiate your TX and how your RX should identify targets. If you have this information, the Frequency domain can show you in what Radiofrequency band should work your TX, what is the most effective form of TX current, how RX should calculate color coefficients of target.

Why "Schwachstrom Elektro Ingeneure" trained in Germany use ELF radiofrequency band for target identification at end of 19-th century? What waveform generates their TX ?

http://www.geotech1.com/forums/showt...704#post164704

That is very very trivial.
The little magnetic domain guys - ooops - the walls - aren't that fast (-> frequency dependent magnetic susceptibility).
The eddy current response is dominating the high frequency region. The magnetic response the low frequency region.
An easy task to do the GB finally.

Just transmit in the frequency region you are actually sampling. Any current wave form is suitable.

Aziz

[PiTec]

Hi Aziz,

I finally found some more time to play with your simulations today. Looks really interesting

Concerning class-E, there is a method to calculate the caps in one of the CPT links I posted a few days ago. This works well as long as the load impedance is very accurate and the switch resistance is almost zero. After I replaced the MOSFET with a 2 mOhm type, the calculation results were within 100nF of the values needed for the simulation to work. You have probably noticed that the circuit is very sensitive to small variations of the cap values, and even worse, also to variations of the load impedance. Just a few tens of mOhms more or less and it completely stops working. So probably not useful for us.

Your TEM 2.0 is far less sensitive to variations of cap values and coil resistance. I got the same simulation results as you posted in the Wideband thread:
http://www.geotech1.com/forums/showthread.php?20163-WIDEBAND-%28BROADBAND%29-TECHNOLOGY&p=167464#post167464

Very impressive Thanks for sharing this - I will definitely take a closer look within the next days.

Thomas