Re: Better, Worse, Worst....... baloney.
Rudi
The NanoVNA app display is scalable. The data beneath the display has 0.001 dB resolution. The actual value at that resolution can be viewed by hovering the pointer over the data curve.
The point here is the White Salamander has been represented to ¡°drift¡± 10 dB or more from cold startup to operation. My tests do not support that claim.
WA8TOD
toggle quoted message
Show quoted text
On Sep 2, 2019, at 11:50 PM, reuterr@... wrote:
Hello Warren,
The diagram has an amplitude span of 70 dB.
In order to see a 0.1 dB deviation it is too coarse.
An amplitude span of 5 dB would be adequate in this case,
if you really want to prove something.
73, Rudi DL5FA
|
Re: If repackaging an NanoVNA, consider adding a bias-T
I saw one of russian(probably) measure chinese cheap LNA( I remember 9-12V / +26dB ) S21 with NanoVNA
He does not attach attenuators to CH1... == NanoVNA does not specified input level of CH1 ==
Not proper way to test I think. But it seems there are no compressions :-)
This is not a professional lab gear. I don't want this stuff's price over $50.
If someone wants to measure active antenna, I think buying DC coupler is better choice.
|
Re: How to measure source impedance?
This sounds more like poor engineering than any actual design goal.
Any HT I have used will happily transmit into a large mismatch with no adverse affects.The fact that this is a 10W HT would lead me to believe that it is probably operating on the fringes of capacity anyhow, and probably hard pushed to dissipate all that heat, even in a perfectly matched situation.
I don't think you'll be able to measure the impedance of the transmit path in place. If you had schematics you could calculate it though.
/ Gerry
|
Re: If repackaging an NanoVNA, consider adding a bias-T
David...
You're not wrong in any of what you say, but I think that your expectations of a $50 hobby-level VNA has maybe been skewed by your experiences with high-end lab grade equipment. The Bias-T options alone on your HP VNA would cost more than a handful of Nano VNAs.
My 0.02
/ Gerry
|
How to measure source impedance?
I have 10W U/V transceiver.
Some of same transceiver users reports burnt after change 3rd party antenna.
-'Don't use 50 Ohm matched aftermarket antenna. This transceiver not matched 50 Ohm'
-'Burnt after even genuine Nagoya/Diamond antenna'
Almost RF designed 50 Ohm( Attenuators, Filters, Antennas, Couplers, ... ) But, How about in the wild?
I don't want spending times to trial/error(watching remotely mounted RF monitor & trimming antenna)
Any genius/proper methods of measuring over 30dB powered source impedance?
Objective: Measuring 10W transceiver's SMA port impedance -> Making custom load calibration kit for VNA -> Making good quality(matched) custom antenna for my transceiver.
Items what I have:
- RF Explorer(6G combo)
- SDR(Lime, RTL)
- miniVNA Tiny
- Oscilloscope(2D72)
- Several 50 Ohm load/terminator(1~15W)
- Several 30dB attenuator
- Kinds of directional coupler( , , Chinese cheap others... )
- VSWR meter for U/V band(SureCom)
Knowledge what I know:
- Audio output port's impedance measure with fine resistor & multimeter ( But, My transceiver is 1W even at LOW mode! I don't want let my transceiver burn without antenna )
- Basic skills of using VNA( Measuring load impedance of unknown antenna/filters, measuring reflection/attenuation )
|
Re: PC Software can not connect to NanoVNA
Please ignore my post.
I tried with another computer, there it worked.
Rebooted the original Windows 10 machine and now it works.
No clue why....
|
Re: Early app for the NanoVNA
Hi Jeff,
for running on Windows, you have the option of using the pre-built release,
which contains a .exe - available here:
Otherwise, having installed Python 3.7.4, you need to install the
prerequisite packages:
pip3 install pyqt5
pip3 install pyserial
pip3 install numpy
(numpy might already come with 3.7.4, I'm not certain)
Then you can start the app:
python nanovna-saver.py
I hope this helps :-)
--
Rune / 5Q5R
toggle quoted message
Show quoted text
On Tue, 3 Sep 2019 at 02:42, Jeff Anderson <jca1955@...> wrote: Rune, I've just downloaded Python 3.7.4. Could you briefly outline what
other steps an old neophyte like myself need to take so that I can use your
Python script?
Many thanks!
- Jeff, k6jca
|
Re: Getting started with the nanoVNA guide
Thanks everyone. Particularly Dr Kirby. Lots of great advice there. I need to work out the best way to collate that together and what changes need to be in part 1 Vs part 2 or even 3.
I have found that trying to write a short beginers guide is actually a pretty difficult balancing act. You don't want to go too deep, but if you keep it too basic, you can actually mislead people.
Regarding the 'high-end¡¯ VNA comments. I think I picked my words poorly. I wasn't really trying to refer to really high end units like the PNA. I was thinking more of the regular professional units that such as Anristsu shockline or R&S ZVL. High end with respect to the nanoVNA. Good enough that it comes with traceable calibration. I will change my wording slightly.
Thanks again
|
The software is able to ignore the odd harmonics of that square wave.
Works something like this:
Two Si5351 outputs are used.
CLK1 is what goes out to the DUT.
CLK0 is always 5khz higher than CLK1, they are mixed
in the three SA612A mixers, the resulting 3 channels of audio
are digitized by a 3 channel 16 bit ADC and the resultant data
fed to the ARM processor. All three channels of audio are
at 5khz, but have different amplitudes and phase.
The 5khz audio is filtered out using DSP techniques (an FFT)
inside the processor, any products resulting from harmonics will
not be 5khz.
For example, assume the signal to the DUT is 100 mhz.
So CLK1 is a 100mhz square wave out to the DUT,
it is mixed with CLK0 at 100.005 mhz resulting in that 5khz audio.
The 3'rd harmonics will be at 300.000 and 300.015,
resulting in 15khz audio which is easily rejected.
When operating at 600mhz, we're beyond the maximum frequency
that the Si5351 can produce. So CLK1 sends out a 200mhz
square wave to the DUT, and we mix that with CLK0 at 200.00166 mhz.
The third harmonic of CLK0 is 600.005 mhz, and it is now
only the third harmonics of CLK1 and CLK0 that produce
a 5khz mixing product.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Mon, Sep 2, 2019 at 08:42 PM, Reginald Beardsley wrote: FWIW Here is a display from my 8560A of the spectrum of the nanoVNA source
producing a 10 MHz CW signal and another showing how it is connected.
The strong odd order harmonic content is probably the major limitation on the
accuracy of the unit. I'm not aware of a way to correct the issue in software.
If someone is, please explain. I'd really like to know about it.
Have Fun!
Reg
|
Stupid cell phone autocorrupt!? My comment should read "hygroscopic", not "gyroscopic".
?Wayne Maynard <wb6bfn@...>?
toggle quoted message
Show quoted text
On Sep 2, 2019, 20:58, at 20:58, Wayne <wb6bfn@...> wrote: Reginald has an excellent point about residue removal.? 91% isopropyl
alcohol works well as does acetone if done quickly.? Just be sure to
remove any left over residue, usually a white powdery stuff.? In
aerospace and military applications we spray the cleaned boards with
polyurethane but that's very hard to remove for repairs.? I've found
that Krylon crystal clear works just as well as a moisture barrier, is
not gyroscopic and is easily removed with a hot soldering iron (don't
breath the fumes).
Never leave anything that absorbs water on a sensitive electronics
circuit board!
?Wayne Maynard <wb6bfn@...>?
On Sep 2, 2019, 20:28, at 20:28, "Reginald Beardsley via Groups.Io"
<pulaskite@...> wrote:
I replaced both SMA connectors with bulkhead units soldered on both
sides of the PCB and while cleaning up was reminded of an important
detail.
The OEM did not clean the flux residue after soldering the SMA
connectors. If you want a stable unit, clean it off with isopropyl and
an old toothbrush. Otherwise you will have time and humidity varying
parasitics. I dry things with a hair dryer on low after cleaning using
91% isopropyl to get rid of the water residue. If you see residue of
*any* kind, rinse it again, blot with a towel and dry. I do *not*
leave any residue behind.
I've repaired a bunch of consumer gear and a 34401A just by cleaning up
the flux residue.
Have Fun!
Reg
|
You can also find 99% isopropyl if you look around, at least in the US. Amazon carries it as well.
toggle quoted message
Show quoted text
On Sep 2, 2019, at 20:58, Wayne <wb6bfn@...> wrote:
Reginald has an excellent point about residue removal. 91% isopropyl alcohol works well as does acetone if done quickly. Just be sure to remove any left over residue, usually a white powdery stuff. In aerospace and military applications we spray the cleaned boards with polyurethane but that's very hard to remove for repairs. I've found that Krylon crystal clear works just as well as a moisture barrier, is not gyroscopic and is easily removed with a hot soldering iron (don't breath the fumes).
Never leave anything that absorbs water on a sensitive electronics circuit board!
?Wayne Maynard <wb6bfn@...>?
On Sep 2, 2019, 20:28, at 20:28, "Reginald Beardsley via Groups.Io" <pulaskite@...> wrote:
I replaced both SMA connectors with bulkhead units soldered on both
sides of the PCB and while cleaning up was reminded of an important
detail.
The OEM did not clean the flux residue after soldering the SMA
connectors. If you want a stable unit, clean it off with isopropyl and
an old toothbrush. Otherwise you will have time and humidity varying
parasitics. I dry things with a hair dryer on low after cleaning using
91% isopropyl to get rid of the water residue. If you see residue of
*any* kind, rinse it again, blot with a towel and dry. I do *not*
leave any residue behind.
I've repaired a bunch of consumer gear and a 34401A just by cleaning up
the flux residue.
Have Fun!
Reg
|
Reginald has an excellent point about residue removal.? 91% isopropyl alcohol works well as does acetone if done quickly.? Just be sure to remove any left over residue, usually a white powdery stuff.? In aerospace and military applications we spray the cleaned boards with polyurethane but that's very hard to remove for repairs.? I've found that Krylon crystal clear works just as well as a moisture barrier, is not gyroscopic and is easily removed with a hot soldering iron (don't breath the fumes).
Never leave anything that absorbs water on a sensitive electronics circuit board!
?Wayne Maynard <wb6bfn@...>?
toggle quoted message
Show quoted text
On Sep 2, 2019, 20:28, at 20:28, "Reginald Beardsley via Groups.Io" <pulaskite@...> wrote: I replaced both SMA connectors with bulkhead units soldered on both
sides of the PCB and while cleaning up was reminded of an important
detail.
The OEM did not clean the flux residue after soldering the SMA
connectors. If you want a stable unit, clean it off with isopropyl and
an old toothbrush. Otherwise you will have time and humidity varying
parasitics. I dry things with a hair dryer on low after cleaning using
91% isopropyl to get rid of the water residue. If you see residue of
*any* kind, rinse it again, blot with a towel and dry. I do *not*
leave any residue behind.
I've repaired a bunch of consumer gear and a 34401A just by cleaning up
the flux residue.
Have Fun!
Reg
|
Re: Better, Worse, Worst....... baloney.
Hello Warren,
The diagram has an amplitude span of 70 dB.
In order to see a 0.1 dB deviation it is too coarse.
An amplitude span of 5 dB would be adequate in this case,
if you really want to prove something.
73, Rudi DL5FA
|
I just discovered that the screen is glued in place with adhesive which is sensitive to isopropyl. so masking it off when cleaning is probably a good idea.
|
FWIW Here is a display from my 8560A of the spectrum of the nanoVNA source producing a 10 MHz CW signal and another showing how it is connected.
The strong odd order harmonic content is probably the major limitation on the accuracy of the unit. I'm not aware of a way to correct the issue in software. If someone is, please explain. I'd really like to know about it.
Have Fun!
Reg
|
I replaced both SMA connectors with bulkhead units soldered on both sides of the PCB and while cleaning up was reminded of an important detail.
The OEM did not clean the flux residue after soldering the SMA connectors. If you want a stable unit, clean it off with isopropyl and an old toothbrush. Otherwise you will have time and humidity varying parasitics. I dry things with a hair dryer on low after cleaning using 91% isopropyl to get rid of the water residue. If you see residue of *any* kind, rinse it again, blot with a towel and dry. I do *not* leave any residue behind.
I've repaired a bunch of consumer gear and a 34401A just by cleaning up the flux residue.
Have Fun!
Reg
|
Re: Better, Worse, Worst....... baloney.
Well Warren, by choosing the words that you did, you did an excellent
impersonation of someone who was trying to demean my efforts.
But I am pleased that you are getting results that please you on your
device.
Onwards and upwards.
Roger
toggle quoted message
Show quoted text
On Tue, 3 Sep 2019 at 12:18, Warren Allgyer <allgyer@...> wrote: Roger, I am not trying to demean your efforts. I am just pointing out that
such differences are not significant in the real world not just to me.
There may indeed be real world significant differences but 900 MHz span
noise floor measurements are not going to reveal them.
I have just done a cold, calibrated start of my White Salamander with a
very poor 150 ohm load which is highly reactive at 720 MHz and stored the
results. The return loss is 25 dB. I will recheck in one hour and see what
drift has taken place. Here is the cold startup.
I will publish the other plot in one hour.
WA8TOD
On Sep 2, 2019, at 7:58 PM, Roger Henderson <hendorog@...> wrote:
Hi Warren,
Not much more to say now, as my results were nonsense and meaningless to
you.
And I won't be running around trying to create scenarios which do mean
something to you.
I have shown my results as they are and for what they are worth. They
showed a difference and that was interesting to me.
I am not trying to prove you wrong, just presenting the facts as I happen
to have two devices to compare.
I also thought it would be an interesting project to try and improve the
white unit, as fundamentally they are not that different.
Roger
On Tue, 3 Sep 2019 at 11:11, Warren Allgyer <allgyer@...> wrote:
Roger
I have seen the article. Return loss measurements of 40 dB or greater turn
to mush on most instruments that cost less than $10,000. No surprise the
same would apply to the NanoVNA. But a couple of points:
1) Use of the NanoVNA or even calibration at a 900 MHz span is nonsense
and meaningless. No one would use a 900 MHz span for anything.
2) Characterizing the noise floor and the resulting possible dynamic range
based on such a span is also nonsense. The resolution bandwidth at that
span is 9 MHz! Useful for nothing.
3) Reducing the span to something useful and reasonable will put even your
White Salamander into spec or very near to it for the 600-900 MHz range.
The general spec is 50 dB of dynamic range from 600-900, 60 dB from
300-600, and 70 dB from 0-300. My White Salamander easily meets that spec
at 720 MHz when the span is reduced to something useful like 5 or even 50
MHz.
Attached is my White Salamander at the full 900 MHz span and with Center
at 720 MHz with a 50 and then a 5 MHz span. Worst case, allowing your
generous 10 dB of margin from the noise floor, it can easily resolve and
display return losses down to -35 dB and that is plenty good enough for
99.9% of the folks here who would use the instrument.
Comparing instruments based on a 900 MHz span noise floor is not very
informative. And no one else, including you, have published comparisons of
reasonable RL measurements made at reasonable spans and RBW.
WA8TOD
On Sep 2, 2019, at 5:45 PM, Roger Henderson <hendorog@...> wrote:
Hi Warren,
I don't know if you have seen this before. See the chart labelled
"Reflected Power Measurement Errors for Various Load Return Loss Values"
in this PDF:
In reality the trace we are talking about is the effective directivity of
the device. This is the error corrected directivity and can be compared
to
the directivity of a directional coupler attached to a Spec An or a Scalar
Network Analyser.
The Vector part of the VNA allows this error correction to improve upon the
directivity of the bridge which is in the device. Interestingly the real
physical directivity of my white unit is slightly better than the hugen
unit.
It is the stability which is worse.
You can attach an antenna and see a spike down below this 'floor'. There is
an example of this in one of the online nanovna reviews IIRC.
The reviewer even questions it and says something like 'it shouldn't be
able to measure that'. It sounded to me that he didn't realise that the
error can go in both directions either. I can't find that review it at the
moment though.
I agree with you, in that if the effective directivity is 30dB, you can
measure down to 20dB plus or minus about 3dB - so your measurement is
within 6dB of whatever it really is.
If that is all you need, then that is fine. Absolutely no issue.
Anyway, I did a couple more tests - one up to 6MHz to shine a light on the
low freq performance.
The other was full span.
See the bottom of the page:
The full span test shows them about the same.
The low freq test shows an obvious difference between the units. The
hugen
units looks incredibly good here. Maybe I have done something wrong!
Whether it is material to you Warren, or anyone else, is not for me to
decide. I don't know what you want to use it for. For me, they are about
the same price. I know which one I would rather have.
Before these tests, I did wrap some aluminum tape around both units. There
are still plenty of holes to let the heat out - Its pretty rough - I'm not
a surgeon - and it took about 2 minutes to do.
I can't say if the tape made any difference though.
Its not worth me spending any more time on this. Hopefully someone else can
pick this up and let us know if blocking up the sides, or putting it in a
bigger box as mentioned by Jerry, does help or not.
Roger.
|
Re: Building the firmware
Reg,
Speaking if price drops of T&M gear, I still have my Telequipment D67 scope I bought new in '73. It was $1k Cdn back then for a 25MHz device. Look what you can get for that from the likes of Rigol and others now.
...and of course, this wonderful little NanoVNA for $50.
Cheers
Larry
|
Re: Better, Worse, Worst....... baloney.
As promised¡.. one hour later. The White Salamander with is real world DUT is returning exactly the same results. It has not drifted even a tenth of a dB from its cold start.
May we use real world measurements of real world DUTs when we do comparative tests?
WA8TOD
toggle quoted message
Show quoted text
On Sep 2, 2019, at 8:18 PM, Warren Allgyer via Groups.Io <allgyer@...> wrote: Roger, I am not trying to demean your efforts. I am just pointing out that such differences are not significant in the real world not just to me. There may indeed be real world significant differences but 900 MHz span noise floor measurements are not going to reveal them. I have just done a cold, calibrated start of my White Salamander with a very poor 150 ohm load which is highly reactive at 720 MHz and stored the results. The return loss is 25 dB. I will recheck in one hour and see what drift has taken place. Here is the cold startup. I will publish the other plot in one hour. WA8TOD On Sep 2, 2019, at 7:58 PM, Roger Henderson <hendorog@...> wrote: Hi Warren, Not much more to say now, as my results were nonsense and meaningless to you. And I won't be running around trying to create scenarios which do mean something to you. I have shown my results as they are and for what they are worth. They showed a difference and that was interesting to me. I am not trying to prove you wrong, just presenting the facts as I happen to have two devices to compare. I also thought it would be an interesting project to try and improve the white unit, as fundamentally they are not that different. Roger On Tue, 3 Sep 2019 at 11:11, Warren Allgyer <allgyer@...> wrote: Roger
I have seen the article. Return loss measurements of 40 dB or greater turn
to mush on most instruments that cost less than $10,000. No surprise the
same would apply to the NanoVNA. But a couple of points:
1) Use of the NanoVNA or even calibration at a 900 MHz span is nonsense
and meaningless. No one would use a 900 MHz span for anything.
2) Characterizing the noise floor and the resulting possible dynamic range
based on such a span is also nonsense. The resolution bandwidth at that
span is 9 MHz! Useful for nothing.
3) Reducing the span to something useful and reasonable will put even your
White Salamander into spec or very near to it for the 600-900 MHz range.
The general spec is 50 dB of dynamic range from 600-900, 60 dB from
300-600, and 70 dB from 0-300. My White Salamander easily meets that spec
at 720 MHz when the span is reduced to something useful like 5 or even 50
MHz.
Attached is my White Salamander at the full 900 MHz span and with Center
at 720 MHz with a 50 and then a 5 MHz span. Worst case, allowing your
generous 10 dB of margin from the noise floor, it can easily resolve and
display return losses down to -35 dB and that is plenty good enough for
99.9% of the folks here who would use the instrument.
Comparing instruments based on a 900 MHz span noise floor is not very
informative. And no one else, including you, have published comparisons of
reasonable RL measurements made at reasonable spans and RBW.
WA8TOD
On Sep 2, 2019, at 5:45 PM, Roger Henderson <hendorog@...> wrote:
Hi Warren,
I don't know if you have seen this before. See the chart labelled
"Reflected Power Measurement Errors for Various Load Return Loss Values"
in this PDF:
In reality the trace we are talking about is the effective directivity of
the device. This is the error corrected directivity and can be compared to
the directivity of a directional coupler attached to a Spec An or a Scalar
Network Analyser.
The Vector part of the VNA allows this error correction to improve upon the
directivity of the bridge which is in the device. Interestingly the real
physical directivity of my white unit is slightly better than the hugen
unit.
It is the stability which is worse.
You can attach an antenna and see a spike down below this 'floor'. There is
an example of this in one of the online nanovna reviews IIRC.
The reviewer even questions it and says something like 'it shouldn't be
able to measure that'. It sounded to me that he didn't realise that the
error can go in both directions either. I can't find that review it at the
moment though.
I agree with you, in that if the effective directivity is 30dB, you can
measure down to 20dB plus or minus about 3dB - so your measurement is
within 6dB of whatever it really is.
If that is all you need, then that is fine. Absolutely no issue.
Anyway, I did a couple more tests - one up to 6MHz to shine a light on the
low freq performance.
The other was full span.
See the bottom of the page:
The full span test shows them about the same.
The low freq test shows an obvious difference between the units. The hugen
units looks incredibly good here. Maybe I have done something wrong!
Whether it is material to you Warren, or anyone else, is not for me to
decide. I don't know what you want to use it for. For me, they are about
the same price. I know which one I would rather have.
Before these tests, I did wrap some aluminum tape around both units. There
are still plenty of holes to let the heat out - Its pretty rough - I'm not
a surgeon - and it took about 2 minutes to do.
I can't say if the tape made any difference though.
Its not worth me spending any more time on this. Hopefully someone else can
pick this up and let us know if blocking up the sides, or putting it in a
bigger box as mentioned by Jerry, does help or not.
Roger.
<PastedGraphic-1.png>
|
Re: Early app for the NanoVNA
Rune, I've just downloaded Python 3.7.4. Could you briefly outline what other steps an old neophyte like myself need to take so that I can use your Python script?
Many thanks!
- Jeff, k6jca
|
Warren Allgyer
HexAndFlex
