开云体育

The output power of the SI5351 is automatic set depending the requested sweep frequency, 2mA below 300MHz, 8mA above 300MHz
Similar for a manually set fixed output frequency using the freq command

You can manually control the power through console by issuing

power [0-3]

The power is in small steps controlling the SI5351 current into the output attenuator
0->2mA
1->4mA
2->6mA
3->8mA

Issuing
power -1
sets the nanoVNA again in automatic power mode

If you report output power please indicate at which power setting the output is generated.

Hi, Andy,

we had the same idea, I measured the output and the frequency accuracy on the weekend. It is very bad if there are 5 kHz difference, e.g. on LF...the band is < 2kHz wide.
I have set the measurement results under Photos, Output to frequency.

Measuring equipment Advantest R3131 with Rubidium-10MHz reference.

73 Joe

Why not try the following console commands through a script:

pause (the scanning - still generates output)

now loop the following via a data driven stream...
freq 1000000
(wait)
freq 1000005

I don't know how fast these commands can be sent - talk to Rune (nanovna-saver author)

73
Larry

I just checked my NanaoVNA output and frequency accuracy, utterly superb.

Good enough to generate WSPR signals, or simple FSK/CW QRSS signals.

The latter are normally send with a 5Hz shift, with a dot length of 6 seconds.

Is there a simple way to use the VNA to generate such signals via USB ?

73 de Andy

I just measure 100mv peak into 50R at 50Khz to 50Mhz (-10dbm)
At 70Mhz it was 70mv (convenient to remember).
I was amazed at how accurate the out put frequency was according to my GPS calibrated freq counter too.

73 de Andy

I think the -10 dBm number is based on conversion of the unterminated voltage. My unit puts out -17 dBm at 100 MHz and -19 dBm at 900 MHz when properly terminated in a 50 ohm load. This level is consistent with other Si5351 devices I own.

One of those devices is the RFzero board which is a GPS referenced, Arduino based synthesizer. It comes with some very nice beacon software that allows it to run WSPR up to 1.5 GHz using harmonics of the Si5351. I have mine operating on 23 cm and get good WSPR decodes using the fifth harmonic.

Warren Allgyer
WA8TOD

Hi Rune:

Keep up the great work - you're an indispensable asset!

You might try thinking a little bit laterally. Other similar python programs I use (like the qspectrum sdr data analyzer) split the window into two sections; one contains the graphs - it scales within limits to track screen size but doesn't scroll, and the second is a sidebar which contains all data and settings and scrolls as desired - with less essential things at the bottom.

I think some of the QT screen designer programs generate platform adaptation code that helps with these problems.

In any case, thanks for working on this problem. I know it's hard to keep a balance at this stage of development between getting the core functionality stable and making the program pretty!

M

Hi hwalker
Ok with your commercial kit, if I am guessing right you do not have the cal kit data ?
The HP male load I have a couple of these and are far from impressed. I have also a BNC special made cal kit characterized against a Rosenberger professional kit with full documentation so I have a made a couple of BNC kit like on the photo with calibration plane data. However if used as ideal kit the calibration is super but of course not at the calibration plane.
The male adaptors are special made as quick insert which I do not think is in the trade so lucky me ?
Kind regards
Kurt

-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af hwalker
Sendt: 19. september 2019 14:02
Til: [email protected]
Emne: Re: [nanovna-users] Electrical Delay_Port Extension

Kurt,
Thanks for the comments. I am fortunate to have a commercial BNC VSWR kit that came with OSL standards plus various mismatches. I picked it up from an engineer's estate sell and was able to verify the kit's performance up to 1 GHz using a HP8753C VNA. The kit is my gold standard against which I verified my everyday use home made BNC OSL. The home made BNC OSL (see attached photo) pretty much follow your construction suggestions, except the 50 ohm load is an HP11593A (picked up two at the same estate sell). I was told the commercial VSWR kit was manufactured by Berkeley labs but have not been able to verify that info.

That looks the business allright :)

Jeffrey,
I've worked slightly more on the ideas you presented, and just pushed an
update that makes it *nearly* fit on a 768 pixel screen. It does actually
fit, I believe, if you hide the task bar. I think that's as good as it's
going to get for today :-)

It should probably fit with font size 7, I think anyway. :-)

--
Rune / 5Q5R

On Thu, 19 Sep 2019 at 07:07, Jeffrey Vandenbroucke <jeffrey@...>
wrote:

CollapsibleBox(QtWidgets.QWidget) would be handy - you can "hide" controls
you don't need then.

For myself, I patched the charts to be 300px in height max. Ditched the
s11_control_box and s21_control_box (and the setText statements); as I can
read that with markers. Font size set to 7; and removed the about box. Now
it fits snug inside my (small) display.

Said,
I posted this procedure earlier that I used to check the gain of a RF preamp. The device under test was a Mini-Circuits preamp with a handwritten note saying, "27 dB gain, 1-1000 MHz".

Test procedure used:
1. Turn the amplifier on to allow it to warm up and also to guard against turn on transients damaging the nanoVNA during testing.
2. Turn on the nanoVNA, set the display to CH1/LOGMAG only, 10 dB per DIV, reference position at 4 divisions to be able to see 27 dB gain.
3. Connect a 20 dB attenuator to CH1 and a 6 dB attenuator to CH0 to closely match the expected 27 dB gain of the preamp.
4. Connect attenuators together using a cable. The nanoVNA should be displaying a 26 dB loss over the frequency range. If not your attenuators are not flat
5. Change to the calibrate menu and select "reset" and then perform a "through" calibration. The new reference should be 0 dB at 4 divisions.
6. Connect the output of the 6 dB attenuator to the input of the amplifier and output of the amplifier to the 20 dB attenuator.
7. Read the amplifier gain in dB directly off the nanoVNA using the marker display.

The above only checks the S21 performance of the preamp. That was all I was interested in at the time.

Kurt,
Thanks for the comments. I am fortunate to have a commercial BNC VSWR kit that came with OSL standards plus various mismatches. I picked it up from an engineer's estate sell and was able to verify the kit's performance up to 1 GHz using a HP8753C VNA. The kit is my gold standard against which I verified my everyday use home made BNC OSL. The home made BNC OSL (see attached photo) pretty much follow your construction suggestions, except the 50 ohm load is an HP11593A (picked up two at the same estate sell). I was told the commercial VSWR kit was manufactured by Berkeley labs but have not been able to verify that info.

Very thanks for quick answer qrp.ddc. I made the 2nd manipulation with 20db connected to ch0 & ch1 and turned the Orange & blue traces & have had the S21 & S11.I think i was wrong.
The amplifiers are 20-24db 2m band.
I made a calibration with the 2 Attenuatos connected to have a zero instead of 40db.
Very thanks for infos & 73s dr qrp.ddc

in order to measure S11, you're needs to put 50 ohm load on LNA output and connect CH0 to LNA input. But first, make sure that your LNA can handle 0 dBm level! If it cannot, then there is no simple way to measure S11. If your LNA cannot handle 0 dBm on the input, do not connect it directly to NanoVNA, because you can damage LNA input.

in order to measure S21, you're needs to connect LNA input to CH0 through 20 dB attenuator. And LNA output to CH1 through 20 dB attenuator.

It's better to use 30-40 dB attenuator between LNA output and CH1. But 20 dB also may work if your LNA has not so high gain (smaller than 30 dB).

For high gain LNA (more than 30 dB) you will need 40 dB attenuator between LNA output and CH1, otherwise your CH1 input on NanoVNA may burn out. Do not connect high gain LNA output directly to NanoVNA, you can damage NanoVNA!

You're needs to add attenuators sum to the measured S21 values. In your case you will need to add 40 dB to measured S21. For example, if it shows S21=-30 dB, it means that LNA gain is -30+40=+10 dB.

hwalker, you can find adapter delay by connect unmatched terminator and check Group Delay screen. Of course it will also include delay of terminator, but with a series of measurement you can find exact delay for adapter and for terminator.

Thanks Rune,
That makes life much easier.

David

--
David G Hopkins (VK4ZF)


---
This email has been checked for viruses by AVG.

My $300 laptop has 1920 x 1080.
Great for video etc but awful for running normal software on, what a crazy situation.
The result is everything looks so tiny on the menus on 90% of the programs in use.
I often have to go into the desktop settings and tell my machine to use a lower resolution :-(

I think the manufacturers are forgetting that PC's are work tools, and are NOT ll about
games and movies !

73 de Andy
======================================

I think that a great majority of the devices sold today (at least the lower-cost ones) have 1920 x 1080 displays, so it's a sensible standard for which to design. I still have some 1280 x 1024, one 1280 x 768, and two 1600 x 1200 (I prefer the greater number of lines for programming and spreadsheets).

My Linx tablet with its 1920 x 1080 x 250 mm display is "challenging" to read, but does run some SDR software well.

(Next challenge is 4/5K displays...)

73,
David GM8ARV
--
SatSignal Software - Quality software for you
Web:
Email: david-taylor@...
Twitter: @gm8arv

I have the same issue. But use another USB cable is sometimes better. If i push OK in the serialPort window, the port is recognized after a few attempts. But yes,it is not ideal.

The IP5303 chip turns on when the load is connected.
Connect a 20-100 ohm resistor to terminal 8 and the GND, if switching on does not occur, replace IP5303.

Hi hwalker
One comment to you BNC calibration test. Are you cal kit home made with three bulhead adaptors where 2x100 ohm SMD fitted to the centerpin as load and the open as well the load has the center filed down flush with the rear of the adaptors and the short is with a copper or brass disk soldered to the rear then you are "calibrating to the rear" and not to the BNC reference plane then you pretty well calibrated as such a kit is sort of ideal. However if another sort of kit you are trouble by delays in open and short and load may be quit reactive inductive or capacitive you do not encounter for.
The NanoVNA-saver has facilities for entering such delays of short and open.
Kind regards
Kurt

-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af hwalker
Sendt: 19. september 2019 00:47
Til: [email protected]
Emne: [nanovna-users] Electrical Delay_Port Extension

Most of the RF equipment I own have BNC connectors, so one of the first things I did was install SMA-BNC adapters on channels 0 and 1 of my nanoVNA. After recalibrating using a 50 ohm BNC OSL kit, the Smith chart display looked as expected when BNC OSL standards were sequentially connected to CH0. BNC mismatches of 33, 75, 100 and 150 ohms all looked good with expected return losses. I saved the BNC calibration data in "SAVE 1" and retained the original SMA calibration in "SAVE 0" in case I need to remove the SMA-BNC adapters.

qrp.ddc's discussion in group regarding the electrical delay menu option reminded me of how the port extension feature of the HP8753C was used to correct for an adapter attached to the native "N" connector. The HP8753C has a menu option allowing you to enter an electrical delay value, extending the measurement plane out to the end of the adapter and correcting for its additional length. I wondered if this would work with the nanoVNA and so tried the following:

1.Turn on nanoVNA. With nothing connected to the CH0 BNC adapter, the Smith chart displays an open with a tail corresponding to the reactance of the SMA-BNC adapter.

2. On the nanoVNA select DISPLAY_SCALE_ELECTRICAL DELAY.

3. I didn't have a clue as to the electrical delay value for the SMA-BNC adapter, so starting at 100 (ps?) I tried successive values until the nanoVNA displayed a single dot with no tail to the far right of the Smith Chart display. The value needed to achieve this for my SMA-BNC adapter was 180 x 1 as entered on the nanovna screen keypad.

4. With BNC OSL standards sequentially connected to CH0, the results looked almost as good as the original SMA calibration data.

5. BNC mismatches of 33, 75, 100 and 150 ohms looked as good or better than the values obtained during the earlier calibration using my BNC OSL kit.

6. I saved the port extension corrected data in "SAVE 2" .


Thanks qrp.ddc for pointing out the electrical delay menu option and how it could be used for correcting for port extensions. For those users that want to add different adapters to channels 0 and 1 but don't have a suitable OSL kit to correct for the adapters, give the electrical delay menu option a try.