Discussion in 'Antennas, Feedlines, Towers & Rotors' started by SP3L, Jun 14, 2021.
Yes. Just ignore it. Everything will work correctly.
I can measure this if you like. I have a NanoVNA and a RigExpert AA-230 Zoom.
Let me know if you prefer spot frequencies or frequency scan.
I am also uploading an existing excel file with some measurements for the validation of your tool.
Also, an idea for the user interface of such a function:
1) An additional entry for the "Supplier" selection of Option C will set the relevant mode, reading: "Options A&B combined".
2) The remaining three selections will not be any more available for the user to select.
3) "Material OD/Height" should display something of the sort: " Opt. A and Opt. B" .
3) "Number of cores" will be the sum of the two options. Min value 1+1=2, max 3+3=6
4) "Number of turns", after a check that the numbers in Opt. A and Opt. B are equal, it will display this number as well.
In case the numbers are different, an error message should be displayed and calculation paused until user syncs the two numbers.
thanks for your files.
If it is not too much work, it would be great if you could fill out a table like the one below:
In all cases the same number of turns, say, 10.
If the measurement process is very laborious, omit every second frequency or so.
We can discuss your GUI proposal a bit later when we are sure how to calculate hybrid cores correctly.
I am attaching the empty table as a zipped Excel file.
Looking forward to your measurement data.
Will be glad to do it but believe it will be more accurate with a single turn. The reason for this is twofold:
1) RigExpert is not that accurate in high impedance scenarios.
2) The measuring harness (piece of wire plus N-type connector to 2 wire lead adapter) has by itself an impedance that must be measured and compensated for.
The shape of the wire plays an important role into this. A short length easily maintains its geometry with or without cores. A multi turn coil doesn't.
This is, in pictures, what I mean and OSL calibration.
OK, I understand. So, please do it for a single turn and we shall see what difference we are getting with a hybrid core.
For users of Excel 2000 through 2003:
I have created "TFCI Calculator (XL2000-2003) v1.1.xls" which is a replacement for what Jacek attached in post #7. If you would like a copy please contact me by direct private email. My address can be found on my home page or on my QRZ page.
Please note that this workbook is intended for use only with Excel 2000 through 2003 and that there is no guarantee that it will be updated with any future enhancements. If you are using Excel 2007 or newer, including Office 365, please get the current workbook (and Help) as found in the zip file on Jacek's download page.
The spikes between 32 and 40 Mhz of the naked cal. loop must be resonances that were of too high a value to be compensated by the meter.
Linearity wise it looks ok. At 50 MHz one core was measured as 115 nH, ditto the other, 231 the combo.
Antenna files also attached, ".txt" extension needs to be changed to ".antdata"
And now the 43 and 61 combo.
Linearity wise it looks ok. At 50 MHz the 43 core measures 115 nH, the 61 core 72nH and the combo 187nH.
Since, at a first glance, linearity looks ok, I don't think there will be any major problem including the combo into your tool.
#12 (43) Z= 15.9 +j36.2
#5 (61) Z =0.00 +j22.8 *
#12 & #5 Z= 15.2 +j58.8
(*) This 0.00 in the real part looks a bit fishy. On the combo it is like a "negative" resistance of 0.7. The imaginary part tracks better 36.2 +22.8 = 59.0 vs 58.8. Some of this difference can be attributed to the rounding error.
I guess, each time I unscrew the banana terminal to insert a new core, or cores, the contact resistance changes a bit from the value that is compensated by OSL, hence over-compensating.
In calculus "negative" resistance can exist, in real life not.
Excel 2003...nothing like using 18 year old software. Are you running a Pentium Pro CPU too?
What a team !!
Jacek and Dan, many thanks for the joint to produce this fine tool.
73, Maximo - EA1DDO
shall we invite you too next time?
thank you very much for the measurement data. I will analyze it during the weekend.
If you do not mind, I will reply to you by email as this subject is too technical to bother all the readers of this thread.
Sure, we can communicate via e-mails. I sent my e-mail to yours, as per your QRZ.com page.
I thought of a way to set-up a "mildly" multi-turn measurement.
First and foremost I need to find a plastic former matching the dimensions of a two core assembly onto which 3 turns will be wound.
This could have been easy to 3D print if I had such a capability, but I don't have.
Three turns seem reasonable for keeping the impedance lower to 1kΩ where the RigExpert has a decent accuracy.
As inductance is proportional to the square of the number of turns, 4 of them on two cores will exceed 1kΩ.
Three measurement samples should be prepared: a) On the dummy former for OSL calibration b) On two same material cores and c) On two different material cores.
Three equal length wires shall be space-wound over the 2/3 of the core's circumference.
But, I will wait to hear from you first as this long week-end I will be away from home.