Balanced VLF Tx

[Aziz]

Hi all,

I'll let the cat out of the bag!
It's difficult to run the simulation due to convergence problems. But you can try, whether you get approximately the same results. Where is the magic or the bug?

Anyway, here it is:


(BTW, this is the raw first cut.)

Cheers,
Aziz

[WM6]

Aziz, can you explain a word more about your magic?

I do not understand your coils configuration. Are there 4 coils?

[Davor]

Yes! Those are chokes just like in switching PSU. I generally try to avoid those in my designs so I did not even think of this layout. One of my goals was establishing amplitude stability, and this design nicely places one side of a coil voltage neatly between the rails. I thought of a switching capacitor voltage stabilised thing instead to retain amplitude stability, but obviously it works this way too

[Aziz]

Aziz, can you explain a word more about your magic?

I do not understand your coils configuration. Are there 4 coils?

Hi WM6,

the coil configuration is simple. Induction balance TX/RX coil (DD type or concentric co-planar IB coil without the bucking coil shown there or any other IB coil type).

The inductors L1 and L2 (chokes) are doing the very big magic here. Look at the voltage difference between nodes (V+) and (V-). Look at the collector currents of the transistors in the transmitter (that's a perfect switching transition).
The transmitter gets its own power supply (half sine wave) driven by the oscillator itself.
Look at the voltage difference between nodes tp1 and tp2 (over the TX coil).
Look at Vcc and Vee, very low ripple noise.

Oh man!, this is a big invention, if no one has invented it already.

Aziz

[ODM]

A constant current driver is usually done in semiconductors for differential oscillators, but why not chokes? There's plenty of selection freedom, it conserves power, and it's the low noise way to go if you have the space, money (well relative to IC process) and weight for it.

[Davor]

Oh man!, this is a big invention, if no one has invented it already.

I tried as I might but all that I found of all cross coupled oscillators were re-chewed VCO creations based on the above mentioned references that have a SINGLE current source, not resistors on each side, let alone chokes, and especially not in a voltage pumping action.

Let me guess, this is a field day for any respectable ( ) patent troll. Just imagine, 20 years after the original design, and each year a few patents copied from it are lodged as original ideas. Patent system is downright ridiculous.

[Aziz]

BTW guys,

regarding the magic circuit (cct) with the chokes and zener diodes:
The zener diodes are just protecting the circuit (limitting the supply rails for the transmitter). If you drive the cct with more input voltage, pay attention, that the V(+) and V(-) voltage doesn't get too much and is within the operating specification limits of the transistors (in the example above I think it is max. 45 V Uce). During the startup process, the peak voltage can be large and the zener diodes are conducting and limitting it.
Pay attention, that they never breakdown in normal operation (sucking battery power -> inefficient).

And don't put any bypass capacitors to smooth out the V(+) and V(-) voltage. They must be half sine waves (fully intended).

To get a stable oscillator voltage, just use a voltage regulator for Vcc, Vee. It should work for low input voltages too.

Cheers,
Aziz

PS: Anyone has seen the merit of the magic?

[Aziz]

I tried as I might but all that I found of all cross coupled oscillators were re-chewed VCO creations based on the above mentioned references that have a SINGLE current source, not resistors on each side, let alone chokes, and especially not in a voltage pumping action.

Let me guess, this is a field day for any respectable ( ) patent troll. Just imagine, 20 years after the original design, and each year a few patents copied from it are lodged as original ideas. Patent system is downright ridiculous.

Hi Davor,

I agree with you. This is a great day for a patent troll.
(Hurry up!, run to the patent office... *LOL*)

Well, this is a good example of collaboration and sharing of ideas. We can learn a lot and improve things. Patent's follow the opposite direction.

Aziz

[Aziz]

BTW guys,

you can put at the V(+) and V(-) node in the transmitter oscillator a constant current source/sink (depending on the unipolar/bipolar drive mode) and can controll the transmit power (regulated oscillator).
This should give you the stability you are searching for.

Aziz

[WM6]

Let me guess, this is a field day for any respectable ( ) patent troll. Just imagine, 20 years after the original design, and each year a few patents copied from it are lodged as original ideas. Patent system is downright ridiculous.

Hi Davor,
I agree with you. This is a great day for a patent troll.
(Hurry up!, run to the patent office... *LOL*)
Well, this is a good example of collaboration and sharing of ideas. We can learn a lot and improve things. Patent's follow the opposite direction.
Aziz

Worth to send as prior art lecture to couple of patent office points. Bravo guy's!

[Aziz]

BTW guys,

if you drive the oscillator with unipolar voltage source, you only need one choke (and optionally one current source/sink depending on the positive or negative supply rail) of course. You need the two zener diodes to protect the circuit.

I have driven the oscillator with bipolar voltage in my example above, just to have a bipolar voltage for the low noise amplifier design and a common ground for the TX reference and RX signal. (NE5532 will be replaced by a single low noise op-amp later. )

The oscillator design allows even much higher coil currents with current source/sink implemented (the super bang mode *LOL*). One only need high voltage transistors and the TX-reference voltage needs lowered too.
But the current source/sink implementation reduces the power efficiency of course.

Cheers,
Aziz

[WM6]

. If you drive the cct with more input voltage, pay attention, that the V(+) and V(-) voltage doesn't get too much and is within the operating specification limits of the transistors (in the example above I think it is max. 45 V Uce).

Aziz, what would be the best transistor operating regime for this purposes?

[Aziz]

Aziz, what would be the best transistor operating regime for this purposes?

That's not critical. Any complementary pair of n/p transistors with enough headroom to the specified parameters.
The transistors I have taken above are doing a good job. And they shouldn't get hot too (very low losses at the transistors).
Aziz

[Davor]

BTW, this whole balanced Tx thing is a mere consequence of my quest to perfect FKK coils solution. Tx coil is in a balanced regime, and it means it produces no net common mode signal, and any capacitive couplings will produce no voltage at all, so this coil requires NO SHIELDING to behave just nicely.

Aziz, your PC rig idea has one component too many: a Tx shield

[Aziz]

BTW, this whole balanced Tx thing is a mere consequence of my quest to perfect FKK coils solution. Tx coil is in a balanced regime, and it means it produces no net common mode signal, and any capacitive couplings will produce no voltage at all, so this coil requires NO SHIELDING to behave just nicely.

Aziz, your PC rig idea has one component too many: a Tx shield

No, I'm storing some energy in the parasitic coil shielding capacitance referenced to ground. But don't tell it to the others or to the patent trolls.
*LOL*

I wouldn't leave the TX coil shielding. Sure is sure. God s(h)ave the queen.

Aziz