开云体育

Thanks for the reply, Vladimir.

Yes, I was able to fumble my way around and did in fact find that at the
time of writing my first note I didn't have the firmware installed right. I
was able to get that accomplished and the rest came together well with the
directions. Boy oh boy does that new firmware make a huge difference. this
little H model has a whole new lease on life.

Thank you.

--
*Andy Sayler, Head Monkey | Keyed Up/Oil Can Garage.*
Down The Street | Lynden, WA 98264 | mobile (360) 319-7417

For measuring larger components below 60 MHz I use a couple of other jigs which work quite well. The first one is built with off-the-shelf cables and a BNC to banana jack adapter. I built a version in a box without the bulky adapter that works slightly better and is easier to use. The cal loads were built from double-sided scrap PCB, copper foil and a 49.9 ohm SMD resistor.

Photos attached

Roger

On 9/6/21 10:55 AM, Maynard Wright, P. E., W6PAP wrote:

Hi, Jim,

Schelkunoff and Friis define "end effect" as being due to the non-zero current flow into the capacitance of the end cap of a finite radius conductor (Section 8.23).? The ARRL Antenna Books and Handbooks define end effect as being due to the increased capacitance of the loop of wire through the end insulator.? Some of the Antenna Books mention the dielectric constant of the end insulator as being a factor.? The 1952 Handbook (my first brand new one) makes a clear distinction between the shortening of the resonant length as a function of the length to diameter ratio and the end loop/insulator effect, which it terms "end effect."

yeah, but the "shortening for resonance" exists even if there is no "end insulator" (example is a rigid rod).

Schelkunoff is talking about a *model* for the wire antenna that would be simple to calculate (along with a bicone feed, right?). That is, he and Friis published a simplified model that would allow calculation of various things by hand/slide-rule.? The actual calculation to solve the integral involves summing infinite series that do not rapidly converge, so they aren't analytically convenient.

A lot of people from the late 1800s on developed a whole raft of approximations to make computation simple.? Sort of intermediate, for instance, is King's 3 term approximation.? They all start by assuming a particular current distribution, and go from there.

In discussing current expansion on wires, the NEC-2 description acknowledges that current may flow into the end cap when the free end is a wire of finite radius.? Separate expressions are provided for a free end and a junction, although the term "end effect" is not used.

That's a slightly different thing - NEC makes a thin wire assumption, and doesn't model currents going "around" the circumference of a thick conductor, nor does it model currents on the end plate of a "wire" - imagine a solid end tube.?? The large diameter "wire" is modeled, essentially, as a bunch of infinitely thin wires in parallel at a radius from the center (like a cage dipole)? (although wrapped into the basis function, they're not separately modeled).? It also uses the diameter when computing the skin resistance effect for a lossy wire.

These seem to be the same effect, but with the ARRL description describing the larger contribution of the end loop and insulator that Schelkunoff and Friis and the NEC-2 description ignore.

Chapters 24 and 25 of Orfanidis seem to assume zero current at a free end, but I'll have to go back and read them in more detail. Fortunately, I can run a lot of Matlab code under GNU Octave, which is easier on my budget.

Virtually all of Orfanidis runs under Octave - many hours spent on airplanes doing modeling with his codes.? There might be a couple routines you have to rename because they overlap with something Octave provides.

73,

Maynard
W6PAP


On 9/3/21 9:58 AM, Jim Lux wrote:

On 9/3/21 9:23 AM, Maynard Wright, P. E., W6PAP wrote:

Texts and references on antennas such as Schelkunoff and Friis, "Antennas - Theory and Practice," Bell Telephone Laboratories, 1952 (printed by Wiley) also discuss "end effect" and "fringing effect" as being characteristic of an antenna element that has a finite radius.

yeah, but later works (Kraus, Balanis) abandon that terminology - because it implies that there's some sort of lumped phenomenon going on, and there isn't.

What it really comes from is if you integrate the Electric Field Integral Equation (EFIE) the solution is a combination of the cosine integral and the sine integral (not the integral of cosine and sin). The real part of the feedpoint impedance is related to the Cosine integral, and the imaginary part is the Sine integral. it doesn't go through zero at lambda/2, but slightly different.

Check out Kraus, chapter 5 (where the cosine and sine integrals are defined) or chapter 10 section 3 (where the self impedance is rigorously developed).? Chapter 9 on the Moment Method is also good (esp since it's what NEC is based on) - (Chapter in 2nd ed, probably the same).

Orfanidis, chapter 24 and 25 - Chap 24 gives all the popular ways (and some not so popular) ways to approximately solve the equation with comparisons among them. Chap 25, section 3 looks at self and mutual impedance, and derives the solution using the Ci() and Si() functions.




I find Kraus more "readable" but Orfanidis, since he's updating it all the time, and it's online, is better for more modern approaches. Orfanidis is substantially more math intensive, but on the other hand, he gives you matlab examples for computation.

Hi, Jim,

Schelkunoff and Friis define "end effect" as being due to the non-zero current flow into the capacitance of the end cap of a finite radius conductor (Section 8.23).? The ARRL Antenna Books and Handbooks define end effect as being due to the increased capacitance of the loop of wire through the end insulator.? Some of the Antenna Books mention the dielectric constant of the end insulator as being a factor.? The 1952 Handbook (my first brand new one) makes a clear distinction between the shortening of the resonant length as a function of the length to diameter ratio and the end loop/insulator effect, which it terms "end effect."

In discussing current expansion on wires, the NEC-2 description acknowledges that current may flow into the end cap when the free end is a wire of finite radius.? Separate expressions are provided for a free end and a junction, although the term "end effect" is not used.

These seem to be the same effect, but with the ARRL description describing the larger contribution of the end loop and insulator that Schelkunoff and Friis and the NEC-2 description ignore.

Chapters 24 and 25 of Orfanidis seem to assume zero current at a free end, but I'll have to go back and read them in more detail. Fortunately, I can run a lot of Matlab code under GNU Octave, which is easier on my budget.

73,

Maynard
W6PAP

Hello Hugen, DiSlord and others,

First:

I know that this is a user and not developer forum. So please apologize for getting deeply into the code (for anyone not directly involoved). This is my first post here, if I have further questions, I will probably go to github.


And: many thanks to Hugen, DiSlord and anyone else responsible for the great work they did and still do. Please continue, we all appreciate that !

Update to USB switch off and on boot problem:

The boot hangs in the first DSP_WAIT, waiting for I2S data arriving from the AIC3204.

i2s_end_callback() is never called
(so DSP_WAIT is in an endless loop waiting for wait_count goes down to 0)

Cause for that:

The prior I2C register load for the AIC3204 didn’t work (my assumption)

I wrote some additional tests, and the ChibiOS calls i2cMasterTransmitTimeout() fail with MSG_TIMEOUT

I added some additional timeout, used an additional SW reset only function, added some delays, no chance.

The cause for the register load problem seems to be the following:
I already mentioned that with USB connected and switched off, VDD is at around 0.7V.
The reset pin of the AIC3204 has the usual circuit: C to GND with pullup R.

But: due to VDD is at 0.7 V in off-position, the reset pin is not low enough, so AIC3204 will not get a HW reset on power up. (My guess, did not measure that).

Sadly, R4 is not connected to STM32 , so there is no chance to force a reset with PB14 of the STM32.

And now to the 0.7 V:

My assumption is that this comes from the voltage divider R40/R39 for
detection of USB voltage.

This not used in code.

Ok, but this connection seems to be the cause (for the 0.7 V), due to protection diode internally in the STM32.

For me, It’s not a big deal, I have an easy workaround for that (I will connect a resistor to GND).

Perhaps Hugen can take this into account if there will be ever a new HW revision …

Assuming that this is not a real defect in one of my devices, but a tolerance thing...

And for me, it was a great learning experience trying to understand the code.


73 / best regards,
Rainer, DK2ZR

I've been reading some past posts, interested in measuring small components. I've come across two jigs, Roger's, and Manfred's. I'm wondering what the advantages and disadvantages are of each.

Looking at the jigs, I don't understand the benefit of using Manfred's, since it has large areas of plane on both conductors. Doesn't that introduce more parasitic capacitance than Roger's small jumper pin fixture? There must be a good reason for each. I'm just not sure what they are.

Albert

Photo - NanoVNA-H, pcb version 3.4
left - VDD and BOOT0

Hello Andy,

have You Driver for NanoVNA installed?
If Yes,
see this photo :

/g/nanovna-users/photo/0/34671.23266.0?p=Created,,,20,2,180,0

please short VDD+BOOT0 and turn on NanoVNA ---> DFU mode

Now start DfuSeDemo and upgrade FW.

Brgds
Vladimir, dl7pga

Good Morning to the group,

I have an old VNA, I think it's what is called an H. I'm not sure if it's an original or a clone either. If possible I would like to take advantage of some of the semi newer features that this unit does not have. I have been working to update the firmware but am not able to get the device in DFU mode or maybe it is but it's not showing up in DfuSe Demo software. I know this is a super noob question but feel that maybe another user could give a nudge in the right direction.

I was able to find some of the reference material in the Wiki and did do some searches but so far no hard info on the older unit as to whether or not they are upgradable.

Thanks in advance.

Andy,

On 9/6/21 1:31 AM, Dirk via groups.io wrote:

Getting this code after running the executable file:

% /Users/vudo/Downloads/nanovna-saver ; exit;
Traceback (most recent call last):
File "nanovna-saver.py", line 21, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "NanoVNASaver/__main__.py", line 36, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "NanoVNASaver/NanoVNASaver.py", line 28, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "NanoVNASaver/Windows/__init__.py", line 2, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "NanoVNASaver/Windows/AnalysisWindow.py", line 23, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "NanoVNASaver/Analysis/__init__.py", line 1, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "NanoVNASaver/Analysis/Analysis.py", line 22, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "scipy/signal/__init__.py", line 302, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "scipy/signal/filter_design.py", line 16, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "scipy/optimize/__init__.py", line 421, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "scipy/optimize/_shgo.py", line 9, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "scipy/spatial/__init__.py", line 107, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "scipy/spatial/transform/__init__.py", line 19, in <module>
File "rotation.pyx", line 5, in init scipy.spatial.transform.rotation
ModuleNotFoundError: No module named 'scipy.spatial.transform._rotation_groups'
[46515] Failed to execute script nanovna-saver
[Prozess beendet]

I also install python3 over brew and chance to interpreter path, but it don't change anything.
python --version > Python 3.9.7

Any ideas how I try to get it running, thank you all very much.

Very odd, why is NanoVNA saver calling

scipy.spatial.transform.rotation

in module

scipy.spatial.transform._rotation_groups

(I checked, and nothing in the source code even has the string 'spatial' in it.

I wonder if it's some dependency thing, or some defaults, and instead of fft (transform)? (but fft is in scipy.fft module)

---
maybe this might help:


although a different solution to a different problem..

--
It appears from googling the specific error message "
|No module named 'scipy.spatial.transform._rotation_groups'" that there might be an issue with how scipy gets installed via pyinstaller. My scipy is version 1.5.2 Python 3.8.6 | packaged by conda-forge | (default, Oct 7 2020, 18:42:56) [Clang 10.0.1 ] on darwin >>> import scipy >>> scipy.__version__ '1.5.2' >>> import numpy >>> numpy.__version__ '1.19.2' >>> import skrf >>> skrf.__version__ '0.15.4'

|

Getting this code after running the executable file:

% /Users/vudo/Downloads/nanovna-saver ; exit;
Traceback (most recent call last):
File "nanovna-saver.py", line 21, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "NanoVNASaver/__main__.py", line 36, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "NanoVNASaver/NanoVNASaver.py", line 28, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "NanoVNASaver/Windows/__init__.py", line 2, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "NanoVNASaver/Windows/AnalysisWindow.py", line 23, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "NanoVNASaver/Analysis/__init__.py", line 1, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "NanoVNASaver/Analysis/Analysis.py", line 22, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "scipy/signal/__init__.py", line 302, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "scipy/signal/filter_design.py", line 16, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "scipy/optimize/__init__.py", line 421, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "scipy/optimize/_shgo.py", line 9, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "scipy/spatial/__init__.py", line 107, in <module>
File "/Users/runner/hostedtoolcache/Python/3.8.10/x64/lib/python3.8/site-packages/PyInstaller/loader/pyimod03_importers.py", line 493, in exec_module
File "scipy/spatial/transform/__init__.py", line 19, in <module>
File "rotation.pyx", line 5, in init scipy.spatial.transform.rotation
ModuleNotFoundError: No module named 'scipy.spatial.transform._rotation_groups'
[46515] Failed to execute script nanovna-saver
[Prozess beendet]

I also install python3 over brew and chance to interpreter path, but it don't change anything.
python --version > Python 3.9.7

Any ideas how I try to get it running, thank you all very much.

When replacing a screen on my tablet, the instructions said that using a
heat gun would soften that double-sided adhesive and make it so you could
carefully separate it.

Joe,

I believe you are building the pi network for an RF amplifier. Once you have measured and installed the coil and capacitors in the RF Amplifier you can tweak the settings to compensate for stray capacitance and inductance by measuring the 50 ohm output impedance in the amplifier by simulating the RF Plate resistance.

BEFORE YOU DO THIS MEASUREMENT VERIFY THE AMPLIFIER IS POWERED DOWN AND PLACE A SHORT ACROSS THE POWER FILTER CAPACITORS TO DISCHARGE THE CAPACITORS!

- Place the RF Tube in it's socket. This will simulate the tube capacitance.

The ARRL handbook gives the following approximation formula for calculating the load
resistance of a vacuum tube power amplifier:

Vp
Rl = ----
KIp

where K is 1.5 for class AB, 1.57 for class B, and 2 for class C.

Calculate and install a non-inductive resistor from tube late to ground to simulate the Plate load.

- Install the NanoVNA on the RF Output connector

- Set the PI Network Capacitors and Coil for each band and adjust the capacitors (and coil tap if necessary) to get the output impedance close to 50 +j0 ohms.

The plate formula is an approximation but these measurements will get you into the ballpark.

- Remove the resistor before operating the amplifier.

BEFORE YOU DO THIS MEASUREMENT VERIFY THE AMPLIFIER IS POWERED DOWN AND PLACE A SHORT ACROSS THE POWER FILTER CAPACITORS TO DISCHARGE THE CAPACITORS!

I used to repair and install Marine MF (2 to 4 Mhz) radios on boats when I was in college 45 years ago. I had to adjust the output series coil to resonate the 20 foot whip antenna. I was I had the NanoVNA back in those days!

73 Mike N2MS

Thanks all I was doing it right.

This group is awesome!

Joe

On Sun, Sep 5, 2021 at 10:31 AM, Andrew Kurtz wrote:

I even found C values more stable as frequency changes, which I have convinced
myself is to be expected because parasitic L in a capacitor is weaker than
parasitic C in an inductor, and also the effect of it (being combined in
series rather than in parallel) has a much smaller impact.

It all depends... If the capacitor is a small value lead length has a considerable effect on the measurement. See graph below of a ceramic cap marked as 10 pF.

Measuring caps was discussed in the link I posted above and many interesting comments were made by knowledgeable members of this group on this subject. I think you will find it well worth reading.

Roger

On Sep 5, 2021, at 1:18 PM, Joe WB9SBD <nss@...> wrote:

I have down right now how to accurately measure a coils inductance.

Now what is the most accurate way to measure a capacitor?

Joe

I measured capacitors exactly the same way as inductors: S11 (reflection) with one side of CH0 attached to one end and the other side to the other end. I even found C values more stable as frequency changes, which I have convinced myself is to be expected because parasitic L in a capacitor is weaker than parasitic C in an inductor, and also the effect of it (being combined in series rather than in parallel) has a much smaller impact.

Andy

On Sun, Sep 5, 2021 at 10:18 AM, Joe WB9SBD wrote:

I have down right now how to accurately measure a coils inductance.

Now what is the most accurate way to measure a capacitor?

An extensive discussion of measurement methods, pitfalls and how to test capacitors and inductors in a previous post last February.

/g/nanovna-users/topic/80744049#20809

Also see how to's of the wiki - /g/nanovna-users/wiki#How-to27s

Roger

I have down right now how to accurately measure a coils inductance.

Now what is the most accurate way to measure a capacitor?

Joe

If you use a wallpaper scraper or maybe a “bluntish” Stanley knife to gently prise the old screen up, you should not damage components on the other side of the board.
Before finally committing to lifting it with possible damage to the flexible connector which needs unsoldering from the board, you should check that the ribbon on the new board is identical.

Steve L

Larry:

Todah Rabah.
May you and all our colleagues have a sweet and healthy year!
73
Ray

From the home of
Prof. Emeritus Raymond (Reuven) Boxman
School of Electrical Engineering
Tel Aviv University
Cell: ???? +972 544 634 217

CEO Clear Wave Ltd.? ?????????????
Scientific Writing Courses: ???????