Induction Balance Stuff - Single/Multi Frequency Response, GB, Disc, Measurements, Ideas, Fun, etc.

[Aziz]

Hi all,

this is very very interesting now. Watch out, we are entering a novel break-through technology now.

I have checked the simple parallel LC resonance (TX-coil + capacitor) on the sound card output/input lines with a high series resistor to get into the off-resonance effects. On a single frequency (low and high). Well, it shows the off-resonance effects for both frequencies. But with both X and R responses, phase shifts, etc.
So it's not the same.

The way I'm using the Off-Resonance TEM2 method seems to cancel the X-Response at appropriate high series resistor (The Heater resistor) on the TXref channel.
So what's going on there?

I love my heating resistor now. My "Hero Resistor".
My RX channel is still offline.
Cheers,
Aziz

[Aziz]

This is so nice. Metal detecting on a mono coil with no X response. A big bunch of Australian hotrocks won't even be detected.
I have not connected the RX coil yet. This will give us the turbo boost.

[Aziz]

Hi all,

there are many nice logic level n-ch mosfets available. I will try different ones. A third off-resonance TEM2 transmitter test board for mosfet version will be done, if I get the mosfets.
The gate source threshold voltage Vgs(th) is specified for 1 .. 2 V. It should be enough to get them switched on with 2-cells (2.4 - 3 V). Even they don't fully switch-on. It doesn't matter.
My BJT version with BC337-40 has approx. 1 Ohm switch-on resistance at approx. 5-6 mA base current.

The mosfet version will reduce the power consumption further.
I have at the moment 3 mA for standby (no transmit signal) and 15.3 mA for full operation with two Ni-MH cells. Most of the energy is lost for switching the transistors and the heater resistor. Hater resistor? Yep, I hate it.

Let's make the low voltage, ultra low power, ground breaking and rocket-science gizmo.
Cheers

[Aziz]

If the TXref channel is out of X-response, we can fully identify the R and X response on the RX channel now. We can solve the simple linear equation.
Without losing any target response. And we have additional TXref channel for target signals.
Win-win.

I will look for logic level mosfets. And I have a bunch of missing parts. I'm also short of soldering wire. Time to spend some money for the parts.

[Aziz]

Bingo! Time to celebrate!



I have increased the R2 upto 1 kOhm so the transmitter is starving of energy. The TXref channel got pure R-response channel now. X-response not detectable. No phase change either.
This is very good news.
Cheers

[Aziz]

Hi all,

the low power, low voltage off-resonance TEM2 is finished and is working as specified.

I had to change the diode D2 1N4148 into a Shottky diode 1N5817. This lowers the voltage drop and the transmit driver comes close to the power supply rails. This ensures operating voltages down to 2 V (I have tested the circuit with flat batteries).

Next step:
Starving the transmitter a bit of its power by increasing the R2 (The Heater). Looking at the R and X response change. R response is the resistive response caused by targets eddy current. Targets "steal" and "burn" energy out of the transmitter. R response should cause the TXref signal go down therefore. Going towards off-resonance.
X response is the ground caused response (iron oxide, ferrite, etc. ) . R response change should be much bigger than X response on the TXref channel.

Then:
Connecting a receive coil (RX) in IB configuration. The RX coil will see the X and R response changes too. But it detects more X response because the induction balance will get out of balance by so called "reactive" materials nearby the coil (ferrite, iron oxides, iron, .. ) .

We can eliminate these X responses in the RX-channel by processing both TXref and RX signal channels. And this process is called ground balance. But we don't want to lose much target signal by ground balancing. This is the reason for having the TXref channel in this off-resonance configuration.

Aziz

[Aziz]

Oh man!,

forgot, that the NiMH battery cells can have more than 1.2 V voltage, if they have been freshly charged. I also want to use the standard 1.5 V cell batteries.
I have to limit the flyback voltage to max. 39 V (I have found the zener diode for). The heater resistor rised up to 330 Ohms to limit the transmitter current further. Or I may increase the driving resistor to lower the base current of the transmitter transistor.

So the finaly specification for a two cell battery: operating voltage down to 1 V each cell (or min 2 V total) and max 2*1.5 V = 3 V (total).
I'm going to heat up my soldering iron today. I don't want to wait for parts to get them.

[Aziz]

Look,

the circuit simplifies a bit. This is the simple low voltage BJT version, limitting the flyback voltage below the break-down voltage of the transmit transistor (BC337-40). The TX-coil current will swing between approx. +100 and -100 mA. The TXref signal between 0 and 2.3 V. Enough to feed it into the sound card input.
Operating voltage 2*1.2 V, 43 mW power consumption, 18 mA battery current. Max flyback voltage is 40 V.

This is a nice circuit. Ready to burn out some parts?

Cheers

PS: There is no choke anymore. The heater resistor has taken the job for.

[Aziz]

Hi all,

I have not decided yet about the selection of the transmit transistor: BJT or logic level n-ch mosfet
BJT:
good: simple, fast, wide range of transistors available, operating voltage down to 2 V possible (due to LM393 comparator), I have convenient transistors at home
bad: requires relative high base current to fully switch-on, bad power efficiency, requires high voltage protection zener-diode due to thermal runaway of the BJT

Mosfet:
good: simple, fast, good power efficiency
bad: logic level below 2 Vgs(th) is probably difficult to find and I don't have any convenient at home. Minimum operating voltage could be higher than 2.4 V.

Well, power efficiency is not going to be relevant here. We are burning a lot of battery energy to limit the transmitters current either. That's the reason for, why I like to use low operating voltage.

I will look, which appropriate n-ch mosfets I can buy. Otherwise, I could even use the NPN BC337-40 transistor (max 45V, 500 mA).
Cheers

[Aziz]

BTW, do we have logic level mosfets? Yes!
Well, ok, in this case, I'm going to use only one mosfet. It will greatly improve performance and efficiency.

[Aziz]

Hi all,

I'm thinking of an another KISS test board, which follows the off-resonance principle using the critical energy balanced double tuned TEM2 transmitter.
Let's call it simply Off-Resonance TEM2 transmitter.

We don't need much power to the TX coil. We don't have extremely high flyback voltages (100 V max should be enough). We don't need more battery power or battery voltage.

The LM393 should run according to the data sheet from +2 V to 36V.
How about a 2-cell battery version? This is 2*1.2 V = 2.4 V (NiMH battery) or 2*1.5 V = 3 V (standard battery)
Yes, I would like it. That's a good suggestions. Accepted.

No mosfets? Ok, accepted.
Simplified circuit? Accepted.
Still TX-current up to +100/-100 mA and above? Yes.

Ok, I'll give my best.

[Aziz]

Hi all,

now the promissed pics and proof of work. It is a messy work at the moment however.

Look, I have reached the optimum operating point.


The flyback voltage is approx. 340 V, the Vc node at approx. 14 V. I have operated the TEM2 transmitter with another battery source (6*1.2 V) during searching for the issues.

What a mess!


And the FFT spectrum of the TXref output:

Well, the SNR (signal-to-noise ratio) is approx. 75 dB. Could be better. It is still noisy.
The choke L1 on the bread board is disabled.

Cheers,
Aziz

[Aziz]

Hi all,

I'm making some progress on the dual-tuned TEM2 transmitter. I think, the choke L1 is too close to my mosfet gate signal. If I take the choke far away, it is working fine and stable as specified. It seems, that it is a layout issue.

BTW, I have found what I was looking for:
Regarding post #53, #54: The critical energy balanced dual-tuned TEM2 transmitter.

This solution is quite interesting. There is a big relation of resistive/reactive response and the resistive >> reactive response. The decoded signal on the TXref will solve the target response on the IB-RX coil side with minimum losses regards to ground balance. This critical energy balanced TEM2 transmitter is best operated on high operating frequencies (>10 kHz), where the resistive loss effects are summed up every cycle.

This solution is working on mono coils too with reduced sensitivity. With IB-coil..
Aziz

[Aziz]

Hi all,

no, no! I can't really tame the damn transmitter. The faulty coil hasn't to do anything with the issue. There must be other reasons (layout, EMI, ..). I don't know yet.
I will do some experiments. Maybe I can fix the issue by chance.

[Aziz]

Hi all,

I am just wondering, whether the faulty TX coil was responsible for not reaching the optimal operating point and instability of my initial design.
On higher operating frequency, the TX coil current was low and the flyback voltage wasn't high enough to cause much sparks in the TX winding wires.

LTspice didn't show any unstable condition. This is, what I have to know for sure.
If yes, I'm going to go back again. I don't like using shunt resistors due to efficiency considerations.

Oh man!, how many changes do I need to the test board?
Cheers