Balanced VLF Tx

[Sergey_P]

Davor;
oscillator:
-> -> Creates a current in the coil -> current creates a field -> a field creates a response from the environment -> changes the response field of the coil -> oscillator modifies the current -> ->

[Davor]

But it is constantly a reference to the rest of the rig. Do you feel as you are travelling at over 1000km/h at this very moment? Well, you are. So does the Earth beneath your feet which is your reference. Tx frequency is the least of your problems. Target responses will retain their phase angle against a Tx signal in a wide span of Tx frequencies, and even more so if the phase shifters are delay-type ... just like in Verator.

[Sergey_P]

You do not get the phase of the medium, the phase shift oscillator compensates
(Step 1 – bug )

[WM6]

WM6: I was referring to scavenged toroid cores for chokes. Some manufacturers like to use conflicting colors, and then there's some manufacturers that seem to just use whatever paint was available at the time

Thanks. Yes, this is very true, but at least we can done some measurement on unknown toroids:

http://www.worldwidedx.com/home-brew...meability.html

[Davor]

You do not get the phase of the medium, the phase shift oscillator compensates
(Step 1 – bug )

It is obvious that I don't get your point, so please elaborate what is required that the medium phase is compensated correctly. From there we'll see how different approaches deal with this requirement, as there obviously are several different approaches that more or less successfully perform their task. Most probably we'll both end up smarter.

[ODM]

Properly characterizing a core is a good thing. But I do admit sometimes taking the smokey approach and just adding turns by "educated guess" and scoping it to confirm findings, and if it looks dirty after the process I'll re-wind it with fresh wire.

Measuring a core is possible with quite basic lab equipment like a power supply, some resistors, and a frequency source. Even a soundcard will do if there's no actual instruments!

[Aziz]

The Gift Day

Hi guys,

I'm putting the legendary Aziz-Davor* Cross-Coupled Balanced LC Power Oscillator spice files for your convenience to play and experiment with.

*Davor deserves it well as I wouldn't get so far without his valuable contribution.

Note, that the inductors L1 and L2 (chokes) are becoming the emitter inductors/chokes. We need two chokes for best operation. And you shouldn't use low inductivity chokes for this purpose (high inductivity: best efficiency, best sine wave forms, best ripple rejection, etc. ) and they should be at least 1 mH each (Well, I'm using 2 x 2 mH or more) .

Tp1 and Tp2 balances at VBat/2 voltage level and can be used as virtual grounds or references.

Here is the schematics:


Here is the spice files zip archive:
Aziz-Davor-Cross-Coupled-Oscillator.zip

Have fun playing with it.

Cheers,
Aziz

[Sergey_P]

the medium phase is compensated correctly

You make the signal processing of Rx, but it is distorted oscillator

[Davor]

Which one? What would be the criteria of a distorted oscillator that would spoil detection? My view is that the culprit is the 2nd harmonic so all methods of fighting it are valuable contributions, including full wave switchers.

Most probably you have a point, only your posts are too short for me to grasp what they are

[Davor]

Have fun playing with it.

I did. I noted a somewhat odd 2kHz modulation going on, and concluded that it must be related to the chokes interaction with the tank capacitances. It disappears when Rck is raised to 10 so that must be it.

[Aziz]

I did. I noted a somewhat odd 2kHz modulation going on, and concluded that it must be related to the chokes interaction with the tank capacitances. It disappears when Rck is raised to 10 so that must be it.

Yep, it's a typical startup process when the oscillator starts running and a high current is flowing through the inductors first time. Indeed, the inductors resistance (Rck) isn't critical and one can make the inductor with thin wire too (more Ohm). As there is less current flowing, the losses aren't that much.

Aziz

[Aziz]

*LOL*
If you change some parameters, you will get more bang at the expense of more power consumption.
L=50µH
RL=0.1
Cp=2.2µ

TX coil current: 9.3 A pp
Power consumption: 98 mA (1.17 W)

Almost all of the energy is burnt in the TX coils resistance (1.05W, ~90% of the input energy is burnt in the coil). But the transistors should keep cold (15 mW each).
I have found, that low inductivity TX coils (<50µH) have a startup problem (can't start oscillating).

Aziz

[Aziz]

Ok, more precisely: Low Q TX coils have a startup problem. And when the osc. isn't oscillating, high current flow through the transistors can cause overheating and damaging. If the TX coil isn't plugged, this isn't critical (almost no current flowing).

Funny, when playing a bit more with the circuit, it tells you a lot.
Aziz

[Aziz]

Played a bit more.
Result:
One toroid core required for the emitter inductors L1 and L2. Keep attention to the phase dots of the inductors. L1 and L2 can be wound bifilar like.



Cleary an improvement.

Aziz

[Sergey_P]

the culprit is the 2nd harmonic

It is well suppressed in the synchronous detector MD (about nonlinearity of soil - keep silent…)