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

[Aziz]

Hi all,

I will also buy a set of NPO capacitors soon. They could even be much better (temp stable, low ESR, rel low D). And very very small. Perfectly to sit in between the coil bundle wires. Paralleled to form a thin string of pearls.

[Aziz]

Hi all,

we have finally found an elegant way to "hide" all the capacitors even in inside a thick mono coil bundle made of plain standard ribbon cable. And the capacitors will be nicely shielded with the whole coil bundle too. Yes, the capacitors needs to be shielded as well. And the ribbon cable acts like a true HF litz wire (parallel 10 or more stranded individual wires).
I will leave a small stripe of gap between 1/2 to 2/3 of the coil bundle width during winding the mono coil and place all the hard wired capacitors there then.

Two thin and shielded stereo earphone cables going from the detector coil directly into the sound card. They can be really thin and flexible. They won't bear significant power (µPower).

I have to look, how much ribbon cable I can use for the mono coil and I have to solder them together, to get a 30 - 40 m long ribbon cable. Enough for a mono coil.
Aziz

[Aziz]

Yep, this idea is novel.
Leaving a limitted region gap for placing the parts in the coil housing, where the magnetic field strength is low at this gap region and hence produces less eddy currents by the parts placed in the coil.
Hurry up Donald (my best mate), we need to file a patent application to get damn fkn rich!

[Aziz]

Now there is a magic trick, where we can leave a gap region to place the capacitors. This region is limitted to the whole capacitors volume space. In other regions, there won't be any gap.
To show this, I have changed the wires to single line wire model. So you can see, where the region is.
It is approx. 1/3 of the bundle width from the outside of the coil. A limitted gap region of let's say 1.5 cm (radial) x 4 cm (circumference) should be enough.
The coil will be wound around the gap regions, if this reached.

Is this novel?
A limitted gap region within a coil bundle to place the capacitors? Or other parts?

[Aziz]

This is the gap between the coil winding bundles with the ribbon cable mono coil.

[Aziz]

Hi all,

what about my ribbon cable mono coil? Where to place the capacitors?
Ribbon cables can't be wound arbitrary. Best to do it as a spiral wound, which isn't best solution for coil capacitance. This isn't critical if operated in VLF mode however.
We have to divide the mono coil in two spiral coil parts, connected in series.
In my example below, I use a 10 core ribbon cable. All cores are connected parallel to lower the coil resistance.
I have taken 16 turns for the lower coil half and 16 turns for the upper coil half in my coil simulation. The mono coil consists of total 32 turns.


As you can see, between the coil halves, there is a minimum of magnetic field region. At this place (as a ring), the capacitors can be placed.

Of course, we could take two different coil radius halves too. Let's look at it next time.

Aziz

[Aziz]

Hi all,

where to place all the capacitors in the coil housing of the mono coil VLF?
A place, where the magnetic field is quite low would be optimal.

There is one good possible solution. Two solenoids connected in series, as basket weave winding to lower the coil capacitance a bit.
Approximately in the middle of the two solenoids (slightly nearer to the inner solenoid) is a good place, where the magnetic field gets cancelled.
Along the ring of low magnetic field, this is a good place to put all the capacitors.

This picture shows, where it could be:

Aziz

[Aziz]

Hi all,

it is high time, to change the bad clips and wires of the basic passive sound card LCR meter. This has an huge impact for accurate measurements.
I should look for the other LCR meter breadboard. It has active j-fet buffers (dual op-amp voltage followers) for much more accurate and extended range measurements due to the input impedance of the line inputs. It is lost without a trace and I don't know, where it is.

Can you imagine, that the 24 bit, 192 kHz sound card LCR meter can outperform a 200 - 300 buck professional LCR meter?

I can measure ESR and D, Q much more accurate now. My PeakTech LCR meter isn't required anymore. It isn't accurate enough too.
ESR and D must be as low as possible at the operating frequency. Both parameters are critical.
MPK10 capacitors are showing very good measured figures. Particularly for ESR.
Aziz

[Aziz]

Hi all,

I am very delighted about the MKP10 capacitors.
They have lower ESR (half) and very low loss factor D compared to FKP1 and FKP2 capacitors.
I have found much more MKP10 capacitors in my box. But I'll focus to low capacitance and low voltage ratings this time to get the capacitors small enough.

I won't use the FKP2 capacitors in the LC tank anymore.
This is an insiders tip (also from audiophile forums).
The low budget MKP10 capacitors are the next best option. The best capacitors would be mica capacitors (extreme low loss and temperature stable). But they would cost me 5 bucks (!!!) for one 10 nF capacitor. I would require 2 - 3 of them. 15 bucks for the LC-tank only? No way!

Anyway, I have enough 10 nF MKP10 capacitors at home right now. And they are small enough too (with 630 V= rating).

I will test some other capacitor types. Maybe I will find another good types.
Cheers

PS: Typo found, it should be MKP10 of course.

[Aziz]

Big big surprize!

I have found a bunch of WIMA MKP10 capacitors:
Guess what!
Even they are metalized film capacitors (not foil), they have very low D and high Q. Some of them outperforming FKP1 and FKP2.
They are available upto very high voltage ratings.
And they are small enough to place them inside the coil housing for the LC-tank (low voltage ratings).

It seems, I have to order some more capacitors.

[Aziz]

Hi all,

can the AI fix a code, even there is no obvious bug in the code? Yes, it can fix it to have more bugs thereafter. If you persist on saying, that there is a specific bug there.
This happened to me. There was a sign bug in the old Lock-in demodulator. I was feeding the code with wrong phase information.

I have tested the LCR meter code now. It works perfect.

Calibration: 4-stage step-calibration in these order
1. LCRCALIB_REFSHORT : Reference resistor Rref (0.1%) shorted -> input channel L/R calibration (amplitude + phase correction)
2. LCRCALIB_DUTOPEN : DUT open calibration
3. LCRCALIB_DUTSHORT: DUT short calibration
4. LCRCALIB_DUTLOAD: DUT load calibration with a second precision reference resistor (0.1%)

Stage 4 improves the accuracy of the LCR meter radically.
The fixed code for the LCR meter has been placed in the attachment.

All calculations will be done in the complex plane.
I can measure D and Q of the capacitors more accurately now. The LC-tank will be made with the best D and Q capacitors.

Aziz

[Aziz]

Hi all,

well, I can work with Claude Code with minimum settings (low effort mode).
It is better than waiting and not getting any answer.
It fixed the LCR meter code for me finally.
Aziz

[Aziz]

Hi all,

this is the current AI LCR meter implementation with bug fixing corrections or little optimizations.
I know, there are still minor bugs in there.
I get same results with my old implementation using calibration procedure up to Dut short (no Dut load applied).
I still can not trust in the Dut load calibaration/correction part. This part has to be fully checked yet.

Do not get confused about the LCRMETER2 structure or LCRMeter2_.. function names. I am running the old LCR meter with new LCR meter (2) at the same time to compare results.
But it shows the nice implementation using complex math with complex measured voltages Vref and Vdut straight out of the I/Q demodulator outputs.
Aziz

[boilcoil]

.................................................. .......
I have to break down the task prompts into simple portions.

It's really a good idea to divide the task into separate parts.

.................................................. .......
But it consumed all the free AI tokens. Next try in 5 hours.

Hmm, this is really annoying.
I don't know if it would help, if you tried to work "in parallel", i.e. register another profile, necessarily with another IP (another computer, proxy...). The article above hints at this when they trying with another IP. You should have a "double resource" in such a mode of operation. The only problem would be if in the process of work up to now you gave the AI some specific data or information that it used. This "external update" will be lost in your new profile and you will have to conduct the training again.

[Aziz]

Hi boilcoil,

Hey Aziz,
I don't want to spam your topic, but if you want fewer errors and bugs - use Python. It is the most well-known language in AL. If you are interested in why LLMs make errors - read the attached article. It is useful mainly because it gives you guidelines on how to proceed most usefully with AL and what to expect from it.

thanks for the interesting paper.
But I don't use Python in my projects. Only C/C++. C++ option part is even kept very very low (Windows GUI related only).
DSP related stuff only in strict ANSI-C (C89 or C99). Strict procedural programming. No object oriented stuff.
So the LLM's should have a very large learing base with ansi-c code.

Claude Code is really nice. But I am facing with token limitations and penalty waiting periods (I'm using the free version only).
"Fix the code" prompt didn't get any results today. But it consumed all the free AI tokens. Next try in 5 hours.
I have to break down the task prompts into simple portions.
Aziz