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

[Aziz]

Hi all,

for a perfect low-loss High-Q mono coil (VLF), I have to buy somewhere a true HF-Litz wire.
I have found one: 200 m roll, 45 x 0,07 0,60mm for 32 bucks.
As it is relative thin, I have to divide the 200 m into 4 - 5 strands and twist them. This should be enough for one mono coil.

I'll think of it.

Let's finish the ribbon mono coil with tiny SMD NP0/C0G LC-tank capacitors first.
Aziz

[Aziz]

Hi all,

I am not happy with my ribbon mono coil yet. Too big and too much weight. The PVC insulation adds a lot of weight to the coil. The coil bundle might produce microphonic effects (it is not a tight coil bundle yet). And it isn't easy to tame the coil bundle. Finalising the coil into coil housing would add more weight. This coil isn't convenient for real field operation. It was a bad idea to use the ribbon cable.

I am looking for other leight weight, simple & easy solutions.

This solution should be easy possible:
The total mono coil is split into two coil halves (each N/2 turns count) with each tight coil bundles, set co-planar, offset center placed in a polystyrene foam groove and fixed.
Coil part 1 is the smaller inner coil half. Coil part 2 is the larger outer coil half.
Wire: thick enamelled copper wire or true HF-litz wire. There is no additional thick or PVC insulation anymore.

Here is the sketch of the mono coil and the zero field drawing:


The zero-field section is easily accessable for placing the NP0/C0G SMD capacitors and soldering the connectors.
We would have less parasitic coil capacitance and less proximity effect too.

A light weight coil design is easily possible. And simple to build too.
Aziz

[Aziz]

Top down view:

[Aziz]

Hi all,

I have been chatting with the AI.

It says, that proximity effect is the most degrading factor in my mono TX coil. The skin effect is generally not much relevant (only 1 - 2%, 1.01 - 1.02 times) and can be neglected. Whereas the proximity effect can cause 3 - 5 times (not percent!) degradation to the ESR of the TX coil.

So the most enemy is the proximity effect in a bundled TX coil. It is even worse, if the bundle is very compact.

We have therefore absolutely no other choice - we have to use the basket weaved TX coil. I could even use the cheap power litz cables (not HF litz) in this case.
I am not going to use my ribbon mono coil. Now focussing to the basket weave mono coil. I have to prepare a rectangular coil former for a two row rectangular basket weave coil. Rectangular coil former is easier to build with wooden strips.

Cheers,
Aziz

[JoyJo]

To reduce the effect of proximity, you need to use a litz wire.

[Aziz]

To reduce the effect of proximity, you need to use a litz wire.

Yep, this is the last step. But I want to use a cheap tinned copper litz wire first. I have a 100 m roll. I can experiment with it before I use a true HF litz wire.

[Aziz]

Well ok,

lets get much smarter than the AI.

Based on the idea of post #365 (the magic cheat):
https://www.geotech1.com/forums/foru...060#post450060

Let's double the magic in there.
This preserves the high Q in the LC-tank by inductive impedance transformer principle.

Everything gets simplified. Even the coil cabling.
We need only the LC-tank capacitor Ct. No more capacitors.
We need two resistors (can be placed in the plug).
We need 1-2 turn primary coil - our low impedance feed-in and feed-out coil.
Our secondary coil is our former VLF mono coil.

Let's assume we have 30 turns on the secondary mono coil (LC-tank). And we have 2 turns for the primary coil. This is N = 30/2 = 15 times relation.
The impedance at the primary coil will be transformed by the factor of N*N = 15² = 225 to the secondary coil.
The source impedance of 2x470 Ohm will be transformed into 211.5 k Ohm (2*470*225). So it does not degrade the Q of the LC tank much.
Our line-input impedance is 10 kOhm. The primary coil is connected differential so it increases to 20 kOhm across the primary coil. This translates (seen by the secondary coil) to 20 kOhm*225 = 4.5 MegOhm. So it does not practically load the LC tank.
And we have a direct low impedance path to the line inputs.

The two series resistors R1 and R2 can be used to adjust/limit the maximum current in the coil (LC coil voltage limitation).

We only need one thin stereo earphone cable to the coil. The resistors can be placed in the output plug jack and split by another stereo cable to the line input jack (short distance).
The LC tank is completely decoupled. The capacitor Ct (or SMD array) is placed in the zero-field region and has the shortest distance to the secondary coil (the main mono search coil).

This is what I mean:


Win-Win-Win.

Aziz

[Aziz]

Hi all,

I am going to test the latest version of the dual frequency mono coil VLF soon.
The very low impedance path to the differential signal is very good to improve the SNR.
A few turns for the additional primary coil will reduce the total parts count and gives more benefits.

BTW, one coil end of the secondary coil can be attached directly to the ground potential of course.

Cheers,
Aziz

[Aziz]

Hi all,

I have tested the last idea. No good.
The sensitivity suffers compared to the simple capacitive drive and capacitive output.
The spice simulation does also says the same.

Now lets focus to the better method.
Aziz

[Aziz]

Hi all,

ok, the resistors R1 and R2 in post #397 are causing too much energy losses. I am going to change them into equivalent impedance Z with two tiny SMD capacitors.
After comparing three different equivalent circuit simulation configurations, they all are producing same target response now. No free lunch. Physics works.

There is only one issue with the solution in post #397:
I am limitted to the maximum line-input voltage level (the sound card input is clipping against voltage supply range +/- 5 V ) . So I am limitted with transmitter coil current and thus maximum transmitter voltage. The idea behind pushing more current through the transmitter coil is simple: Improving the SNR.

There is enough room to maximize the TX coil voltage up to 50 Vp. This is the max. voltage limit of the SMD capacitors.
There are more options:
1. old style: capacitive drive TX, capacitive out TX (with flexible inherent voltage division)
2. 1 x magic: capacitive drive TX, secondary RX coil (with transformer voltage division)
3. 2 x magic: now capacitive primary coil drive, secondary TX (LC-tank), (with transformer voltage division), #397 style
4. 1 x magic: capacitive primary coil drive, capacitive out secondary TX (with flexible inherent voltage division)

To push more current through the transmitter, option 4 is more convenient. The Q of LC-tank won't suffer much in this configuration. And we can set the voltage division factor flexible. Option 3 limits our maximum coil voltage as described.

I'll test option 3 first.
Aziz

[Aziz]

Hi all,

I have replaced the resistors R1 and R2 by 2 x 10 nF SMD capacitors now (option 3). It is working much better now.
I have even used the 0805 SMD types. A copper tape with adhesive film put onto a small paper or cardboard stripe it is quite manageable.

My primary coil: 2 turns,
Secondary coil: 28 turns (the LC-tank mono coil)
L = 465 µH, LR = 0.3 Ohm, Ct = 8 x 3.3 nF = 26.4 nF NP0/C0G SMD, f1=45100 Hz, f2=45200 Hz
N = 28/2 = 14, N² = 196
The two 10 nF capacitors will have an equivalent parallel resistance Rp to the LC-coil:
Rp = 2 * 354 Ohm * N² = 138.8 kOhm. The line-inputs of 2x10 kOhm would add parallel impedance to Rp 3.9 MegOhm and can be neglected. So it does not degrade the Q of the LC-tank much.

BTW, my coil losses are much much more. So the Rp will get quite low. I have used the bad wire for the mono coil (1.5 mm² copper litz with PVC insulation).

I will have not much free time for the next 10 days.
Cheers