¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io
  1. Nanovna-Users
  2. Messages
Expanded Topics Messages
Date
Re: ferrite identification

 

Re: Windows 10 H4 Driver

 

If I understand correctly, The Windows 8 Driver is available on the Windows 10 System to access the virtual com port. The DeFuse_Demo uses a Windows 10 Driver. Where do you get the Windows 8 driver for it and how do you get the DfuSe+Demo to use it?

John KE0ZUW
Re: Glue broken FT 249-31 ??

 

--- On Saturday, October 23, 2021, 10:18:48 AM EDT, Manfred Mornhinweg <manfred@...> wrote:
Assuming that the material characteristics weren't affected in the fire
This is probably a good assumption, because ferrites are made of "already burned-up" material (Fe2O3).

73,
Ken N8KH
Re: Glue broken FT 249-31 ??

 

Terry,

What sort of power were you running. At 100 W... these get RED hot! Fine for QRP. Unless these cores are operated in a true balanced configuration, temp rise is a real challenge and if you get near the Curie temp... watch out.
Re: Glue broken FT 249-31 ??

 

I don't think that this was an EMP, nor lightning. The coax seems intact from the balun down, but the balun burned, and also the coax above it. That looks like massive overload of the cores, probably due to a combination of high power with a rather inconvenient combination of the balun's impedance with those of the antenna/feedline/earth system.

Assuming that the material characteristics weren't affected in the fire, the only effect of gluing the cores back together is a slight reduction in effective permeability. It depends on the thickness of the glue layers relative to the total path length of the core, and the ratio between the material's permeability at a given frequency to that of the glue or air (1).

With low permeability materials, a good, precise fit, and watery thin glue, the effect of the glue joint can be negligible. With thicker glue, a poor fit, and high permeability materials the effect gets larger.

A few weeks ago I had an "aha" moment when I was doing loss tests on junkbox EE cores. I just apply an adjustable drive level, at adjustable frequency, from an oscillator/driver circuit, and monitor the temperature rise of the core. I noticed that in many of my EE cores one side leg got much hotter than the other. Changing the clamping pressure changed this. Despite the hardness of ferrite, the small deformation was enough to change the airgaps in those joints, so that most magnetic flux went through the side having the smaller gap, thus overheating it while the other side stayed cool!

Among those cores there was one that had a broken leg, which I had glued back in place with cyanoacrilate glue (instant bonder). No matter what I tried, the glued-back leg stayed stone cold while the other leg got very hot! Clearly the glue layer, and the resulting less perfect mating of the lapped surface between the two E's, drove the effective permeability of that side much higher than the one on the healthy side of the core.

Those ferrites have a permeability of 2000. Material 31 has a permeability of 1500 at low frequencies, so it's not much different, but in the higher RF range it's much lower. So, I would expect that glueing back together your toroids is likely to cause a noticeable reduction of impedance on the lowest bands, but little effect on the highest ones.

Try it, and measure. Nothing to lose.
Re: ferrite identification

 

On 10/20/21 8:44 PM, Dragan Milivojevic wrote:
If you can, I have a full bin of unknown ferrites waiting for
classification ;)

I have 2 bins. One for non-conductive cores (ferrite) and the other for conductive (iron).
Re: Glue broken FT 249-31 ??

 

Yep. If the fit between the broken ends is very good, and the glue line kept very thin, the performance will be essentially the same. Why not use your nanoVNA to compare the repaired broken cores to one that isn't broken, assuming you have one. You can check them just holding/taping the broken pieces tightly together. If you don't have an unbroken one, compare the broken performance the specs for the core.
Re: Glue broken FT 249-31 ??

 

Probably a magnetic pulse.? ?Hi current sort of made the choke into a linear electric motor.
Yes, you can glue them back together with virtually no difference in inductance.Back in my telephone equipment days, we would assemble complex cores with epoxy.Far easier than a complex winding operation.? Kent

On Friday, October 22, 2021, 09:21:15 AM CDT, Terry Wassell <k3jt.wv@...> wrote:

Several of my cores cracked and I am wondering if I can still use them if I superglue or epoxy them back together.? I thought it was lightning blew up my 160m tuning box and choke, but it might have been choke core heating.? ??
Glue broken FT 249-31 ??

 

Several of my cores cracked and I am wondering if I can still use them if I superglue or epoxy them back together. I thought it was lightning blew up my 160m tuning box and choke, but it might have been choke core heating. ??

20211013_082035.jpg
20211013_081802.jpg

Re: Using NanoVNA only with USB

 

You should not have any (big) issues running off of USB as that is how most of the PC applications talk to the device.
You may want to wrap the USB cable through a torroid though to reduce any RFI from the PC.

On Friday, October 22, 2021, 06:28:37 a.m. EDT, Francesco <realfran@...> wrote:

Hello, using the NanoVNA only from the USB supply without battery some problem?
Using NanoVNA only with USB

 

Hello, using the NanoVNA only from the USB supply without battery some problem?
Re: Broken S-A-A v2 ?

 

Had exactly the same problem, tried Eugen's suggestion and it worked! Thanks!
Re: "Q", Coils, toroids, and guesswork?

 

Well done, including making all the contacts and improvements! Thanks.

Op do 21 okt. 2021 om 23:46 schreef Roger Need via groups.io <sailtamarack=
[email protected]>:

Today I saw that Owen Duffy posted a new blog entry about Miguel Vaca's
Ferrite calculator.



He noticed a discrepancy between my measurements of an FT50-43 inductor,
his calculator and Miguel's calculator. In this post I can explain the
reason why.

I also noticed an inductor calculation discrepancy and contacted Miguel by
email a few days ago. He is very responsive to user input and suggestions
and quickly discovered that it was a coding error and released version 0.8
withing a few hours. During our conversation I mentioned that Owen Duffy
had reported on his blog that Fair-Rite changed the complex permeability
graph and CSV data for Mix 43 in 2020. This is significant because it
makes a considerable difference in the calculated inductance. Miguel added
an option to use old or new Mix 43 data for the calculation.

When I made my measurement a few years ago I used the old Mix 43 data
because my ferrites have been in the parts bin for years. Attached is what
I measured on a RigExpert AA-55 Zoom one port VNA and what I got with the
Duffy calculator. Very close agreement to 4 uH.

Attached is a screenshot from Miguel Vaca's version 0.8 calculator using
pre-2020 Mix 43 data. It shows 4.09 uH at 3.63 MHz. So the two calculator
measurements (Duffy and Vaca), for practical purposes, are the same.

Also attached are screenshots using the latest Fair-Rite Mix 43 data. The
Miguel calculator gives 6 uH at 3.6 Mhz and the Duffy calculator gives 6.05
uH

In summary there was an inductor calculation error in version 0.7.
However it was a user input error on Owen Duffy's part that led him to
believe there was a discrepancy in the latest version 0.8. He is a
reputable technical writer and I am sure he will correct this error so his
readers will not be misled.

If anyone finds bugs, errors or has suggestions I suggest they contact
Miguel directly by email (email link at top of app). The app is currently
in beta release and he appreciates constructive feedback.

Roger
Re: "Q", Coils, toroids, and guesswork?

 

Today I saw that Owen Duffy posted a new blog entry about Miguel Vaca's Ferrite calculator.



He noticed a discrepancy between my measurements of an FT50-43 inductor, his calculator and Miguel's calculator. In this post I can explain the reason why.

I also noticed an inductor calculation discrepancy and contacted Miguel by email a few days ago. He is very responsive to user input and suggestions and quickly discovered that it was a coding error and released version 0.8 withing a few hours. During our conversation I mentioned that Owen Duffy had reported on his blog that Fair-Rite changed the complex permeability graph and CSV data for Mix 43 in 2020. This is significant because it makes a considerable difference in the calculated inductance. Miguel added an option to use old or new Mix 43 data for the calculation.

When I made my measurement a few years ago I used the old Mix 43 data because my ferrites have been in the parts bin for years. Attached is what I measured on a RigExpert AA-55 Zoom one port VNA and what I got with the Duffy calculator. Very close agreement to 4 uH.

Attached is a screenshot from Miguel Vaca's version 0.8 calculator using pre-2020 Mix 43 data. It shows 4.09 uH at 3.63 MHz. So the two calculator measurements (Duffy and Vaca), for practical purposes, are the same.

Also attached are screenshots using the latest Fair-Rite Mix 43 data. The Miguel calculator gives 6 uH at 3.6 Mhz and the Duffy calculator gives 6.05 uH

In summary there was an inductor calculation error in version 0.7. However it was a user input error on Owen Duffy's part that led him to believe there was a discrepancy in the latest version 0.8. He is a reputable technical writer and I am sure he will correct this error so his readers will not be misled.

If anyone finds bugs, errors or has suggestions I suggest they contact Miguel directly by email (email link at top of app). The app is currently in beta release and he appreciates constructive feedback.

Roger

4 Turns of FT50-43 Duffy latest Mix data.PNG
4 Turns of FT50-43 Duffy old Mix data.png
Miguel Vaca Calculation - New.PNG
Miguel Vaca Calculation -Old.PNG

Re: ferrite identification

 

**Updated post

Fair-Rite has an article on determining the Mix of a ferrite.



All the calculations to determine the initial permeability of a material can be done by measuring the toroid, winding some turns on the toroid, measuring the inductance at 10 to 100 kHz. and entering the data into this calculator.



Roger
Re: "Q", Coils, toroids, and guesswork?

 

To have something to do while locked up in an unforgiving apartment (out of
the High Park Fire here in N. Colorado in 2012), I built what I considered
my "ultimate" crystal radio. However, I should note I cheated just a bit
by using germanium diodes instead of galena and a cat whisker. Also proved
to myself that the good 'ol germanium diodes work a bit better than good
Schottky diodes (from HP).

All that put aside:

1) tapping the coil down for the detector makes a huge improvement in
selectivity - Q of the single tuned circuit - and to a lesser extent, the
recovered audio. It is a trade-off between recovered audio and
selectivity. I'm located in N. Colorado in the county. The single HF
antenna I use is a 450-foot long doublet (implying non-resonance in the ham
bands). It's fed with parallel conductor transmission and a common mode
choke in the shack. It's lowest 1/2-wavelength resonance is roughly 950
kHz. During good propagation in the evenings (once the D-layer has
dissipated and the power reduction at sunset goes into effect for AM
broadcasters), I'm able to easily separate as many as 18 to 22 different
stations with reasonably recovered audio. We have one "blowtorch" in
Boulder at 760 kHz that at times presents a problem as they do not always
follow the FCC rules of power reduction and another in Fort Collins on 600
kHz for the same.

2) I also did something somewhat unorthodox in addressing the antenna
"connection" to the receiver. Instead of tapping down for the antenna
coupling, my antenna input to the receiver is a separate coil wound on the
same PVC form as the main resonant circuit. The "input" inductor is
separate from the main tuned circuit and coupled to the antenna through a
series roughly 200 pF variable capacitor. With that, I can control the
current through the "input" coil and, therefore, the loading of the main
(larger) resonant circuit due to the antenna connection. This also has a
major effect on selectivity. Again, a trade-off between recovered audio
and selectivity.

There are also some good resources online. There is also a Crystal Radio
Society online:

JUST MY OPINION: I believe everyone playing with RF should build at least
one crystal radio *from scratch*. In my opinion, it's the best teaching
tool addressing resonance and Q. Here in Colorado, all we need to do to
procure a chunk of galena is visit one of the many mines in the mountains.
Virtually all of them have galena present on the tailings piles.

NOTE ON GALENA: The fine sparkly crystalline galena does not work for
crystal radios. The fine crystalline structure is caused by up to 10%
silver substituting in the PbS crystal matrix. Too much silver in the
galena spoils its "band gap / rectification" properties. Good old chunky
large crystalline galena works best.

NOTE ON THE APARTMENT RFI: Sure, it was a large complex, but...... The
RFI was so bad in the complex that I could easily detect RFI at
many......many harmonics of 60 Hz with this crystal radio. SMPSs, Plasma
TVs, and PCs in every unit!!! Not to mention the 2.45 GHz emissions around
meal times (detected on a spectrum analyzer and cliplead for an antenna).
I used the downspout as an antenna in that situation as we were, by choice,
on the top, third, floor.

Dave - W?LEV

On Thu, Oct 21, 2021 at 1:34 AM Andrew Kurtz via groups.io <adkurtz=
[email protected]> wrote:

VERY helpful; thanks! I think I am half-way there as I have put a
coupling coil on my tuning inductor so the antenna signal goes through only
that, with about 15% as many turns as the inductor. On the other side, I
have played with tapping the detector off somewhere down the tuning
inductor, but it seemed not to work. I will play with that some more¡­
Andy

On Oct 20, 2021, at 8:22 PM, Manfred Mornhinweg <manfred@...>
wrote:

Andy,

with your crystal radio something similar happens as I described in my
post, about the low pass filter. Your radio's tuned circuit is loaded both
by the antenna and by the detector+headphones. This loading will cause a
loaded Q that's much lower than the natural Q (unloaded Q) of the resonant
circuit. You only need an unloaded Q several times higher than the loaded
Q, and from there on any additional improvement in the Q of the coil won't
make an appreciable difference. The selectivity will be limited and
dominated by the loading by the antenna and the detector.

You should use taps on the coil, to tap down both the antenna and the
detector as much as possible consistent with acceptable volume, to increase
the loaded Q and thus improve the selectivity.

Assuming you are using a typical 365pF variable capacitor, its reactance
is about 800? at the low end of the broadcast band. If you use 2k?
headphones, and assuming that the detector doesn't change the average load
impedance that gives on the coil (I'm not sure of that right now), then the
loaded Q is already limited to roughly 2.5, which will give extremely poor
selectivity. If you tap down the detector to the midpoint of the coil, it
will act like 8k? across the whole coil, assuming good coupling across all
coil turns, and that will improve the loaded Q to 10. That's still too poor
for good selectivity. And an unloaded Q of 120 is still high enough to
cause almost no degradation of the loaded Q. And we also have to factor in
the loading by the antenna, which further worsens the loaded Q. And at the
high end of the frequency range, the impedance of the coil and the tuning
capacitor are 3 times as high, and your loaded Q gets 3 times lower, making
the percentual bandwith 3 times larger, which at that 3 times higher
frequency is a 9 times worse selectivity against adyacent channels!

So it's pretty clear that to get good selectivity with a crystal radio,
you need to tap down the detector and the antenna really a lot, into the
first tenth or so of the coil, before even starting to worry about
improving its unloaded Q!

Or use more modern circuitry, such as RF amplifiers, perhaps in the form
of a "Q multiplier", which is the name that was given to a positive
feedback amplifier circuit that was used to bring the apparent Q of the
tuned circuit up as high as you wanted, even beyond infinity, which would
make the circuit oscillate.

If you connect the









--
*Dave - W?LEV*
*Just Let Darwin Work*
Re: ferrite identification

 

NanoVNA Sorts Unknown Ferrite Suppression Beads by Fair-Rite Products Corp.

On 10/21/2021 12:56 PM, W0LEV wrote:
At one time in the not too distant past, FairRite posted a well done video
on identifying ferrite material. It was excellent and may even be in the
archives of this group. I have lost the reference to that presentation.
I'd suggest poking around on the FairRite site.
Dave - W?LEV
On Thu, Oct 21, 2021 at 3:31 AM Jim Lux <jim@...> wrote:

On 10/20/21 4:20 PM, Dragan Milivojevic wrote:
There are huge variations in ferrites, parameters of the same model
from different batches can vary greatly.
True, but probably not so much that you can't come up with a quick way
to identify them. 20% variation probably doesn't move it to confuse with
a different mix.

When I get some time this weekend, maybe I'll look at the FairRite data
I have and see if this is an easy way.







--
73,
Mike, N1JEZ
"A closed mouth gathers no feet"
Re: ferrite identification

 

At one time in the not too distant past, FairRite posted a well done video
on identifying ferrite material. It was excellent and may even be in the
archives of this group. I have lost the reference to that presentation.
I'd suggest poking around on the FairRite site.

Dave - W?LEV

On Thu, Oct 21, 2021 at 3:31 AM Jim Lux <jim@...> wrote:

On 10/20/21 4:20 PM, Dragan Milivojevic wrote:
There are huge variations in ferrites, parameters of the same model
from different batches can vary greatly.
True, but probably not so much that you can't come up with a quick way
to identify them. 20% variation probably doesn't move it to confuse with
a different mix.

When I get some time this weekend, maybe I'll look at the FairRite data
I have and see if this is an easy way.







--
*Dave - W?LEV*
*Just Let Darwin Work*
Re: ferrite identification

 

On 10/20/21 8:44 PM, Dragan Milivojevic wrote:
If you can, I have a full bin of unknown ferrites waiting for
classification ;)
don't we all
Re: ferrite identification

 

If you can, I have a full bin of unknown ferrites waiting for
classification ;)

On Thu, 21 Oct 2021 at 05:31, Jim Lux <jim@...> wrote:

On 10/20/21 4:20 PM, Dragan Milivojevic wrote:
There are huge variations in ferrites, parameters of the same model
from different batches can vary greatly.
True, but probably not so much that you can't come up with a quick way
to identify them. 20% variation probably doesn't move it to confuse with
a different mix.

When I get some time this weekend, maybe I'll look at the FairRite data
I have and see if this is an easy way.
14981 - 15000 of 39714