开云体育

Made new topic for more easy search.

On this weekend can upload my sources an work under it

Update firmware to 0.8.4.3
Remove frequency from properties data (rebuild it on load)
Optimize config and properties struct (made data align, and made more compact size)
New battery icon
Fix calibration interpolate function near harmonic frequency change (you can see 10 Segment measure from 290MHz to 305MHz on screenshots before and after), interpolate now not need src frequency table for build. For freq in middle point calibration data from different harmonic now made extrapolate from 2 points before or 2 points after, on all other cases made interpolate.

For H4 version remove frequency from properties data allow reduce calibration slot size for 201 point from 12kB to 8kB (and allow save 7 slots)
For H version slot size still > 4kB (4152 byte, need reduce size on 56 bytes to 4096 for more compact save slot size 4kB)

WARNING, all config and calibration will be reset to default after update
Default sweep_points = 101 for both
Default bandwidth = 1k
Default threshold = 300MHz (on some chips possibly unstable work, in some cases need reduce it, for example my H version need 299.8MHz, H4 version can stable at 301MHz, you can see it on scan range from 295 to 305MHz if set 10 segments as bad data recive in freq near 300MHz)

You can change it and save

Hi DiSlord,

Thanks vy much for the new firmware, I flashed it into my NanoVNA-H4.
I still have a small bump at 300MHz as you can see in the attached
picture, so not so good as your picture, sorry. I used the same
settings in NanoVNA Saver (non Calibrated).

I recalibrated the VNA-H4 standalone at 10k to 2.7G before.

The picture is taken mat 300MHz threshold, I tried a few different
settings, but no improvement.

Cheers,

Jos

Op 21-4-2020 om 22:15 schreef DiSlord:picture

Made new topic for more easy search.

On this weekend can upload my sources an work under it

Update firmware to 0.8.4.3
Remove frequency from properties data (rebuild it on load) Optimize
config and properties struct (made data align, and made more

compact size)

New battery icon
Fix calibration interpolate function near harmonic frequency change
(you can

see 10 Segment measure from 290MHz to 305MHz on screenshots before and
after), interpolate now not need src frequency table for build. For
freq in middle point calibration data from different harmonic now made
extrapolate from 2 points before or 2 points after, on all other cases made interpolate.

For H4 version remove frequency from properties data allow reduce

calibration slot size for 201 point from 12kB to 8kB (and allow save 7
slots)

For H version slot size still > 4kB (4152 byte, need reduce size on
56 bytes

to 4096 for more compact save slot size 4kB)

WARNING, all config and calibration will be reset to default after
update Default sweep_points = 101 for both Default bandwidth = 1k
Default threshold = 300MHz (on some chips possibly unstable work, in
some

cases need reduce it, for example my H version need 299.8MHz, H4
version can stable at 301MHz, you can see it on scan range from 295 to
305MHz if set 10 segments as bad data recive in freq near 300MHz)

You can change it and save

?usp=sharing

On Wed, Apr 22, 2020 at 12:53 PM, <erik@...> wrote:

A stepwise transition when changing harmonics could be caused by non linearity
in the SA612 producing internally a different level of harmonics that disturbs
the real signal. As it is nonlinear it can not be calibrated out.
A way to test is a measurement that depends on a linear scale.
An example is measuring a SMA tee with two 50 ohm loads connected so you get a
25 ohm load with some delay.
On a calibrated nanoVNA this should look like a arc starting at pure 25ohm and
circling upwards with 50ohm as centerpoint.
Any non linearity will be clearly visible as the arc moves away from 25ohm
circle around pure 50 ohm.

--
NanoVNA Wiki: /g/nanovna-users/wiki/home
NanoVNA Files: /g/nanovna-users/files
Erik, PD0EK

On screenshot i measure my test 2.0 SWR calibration load (i test it up to 4GHz on Anritsu S820d and it show ~2 in all range)

I try compensate this by add L0 coefficient for Short calibrate standart in NanoVNA Saver. But need different values for every harmonic (and on smith pure SHORT calibrate standart look like arc).

I not have good knowelege in this, and don`t wont change this part of code while not understand it.

Hi DiSlord,

I did some more testing on my NanoVNA-H4 with your? 0.8.4.3 firmware using NanoVNA-Saver and found a small square in the S11 Returnloss display at 300 MHz..... After setting the threshold from 300 to 300.9MHz it was completely gone (see my pictures attached ) ............After calibration in NanoVNA-Saver I took these pictures . The 25 Ohms pictures were taken with SMA Tee + a Female - Female adaptor (not taken into the calibration).

All measurements were done with NanoVNA-Saver set to 10 Segs and the -H4 bandwidth set to 1kHz. My 50 Ohms standard measures (4 point) 49.85 Ohms at 10kHz , so I used this value for "load calibration, the parallel R measures 50.69 Ohms (4points 10kHz).

My ten cents.

Regards

Jos

? Op 22-4-2020 om 22:34 schreef DiSlord:

On Wed, Apr 22, 2020 at 12:53 PM, <erik@...> wrote:

A stepwise transition when changing harmonics could be caused by non linearity
in the SA612 producing internally a different level of harmonics that disturbs
the real signal. As it is nonlinear it can not be calibrated out.
A way to test is a measurement that depends on a linear scale.
An example is measuring a SMA tee with two 50 ohm loads connected so you get a
25 ohm load with some delay.
On a calibrated nanoVNA this should look like a arc starting at pure 25ohm and
circling upwards with 50ohm as centerpoint.
Any non linearity will be clearly visible as the arc moves away from 25ohm
circle around pure 50 ohm.

--
NanoVNA Wiki: /g/nanovna-users/wiki/home
NanoVNA Files: /g/nanovna-users/files
Erik, PD0EK

On screenshot i measure my test 2.0 SWR calibration load (i test it up to 4GHz on Anritsu S820d and it show ~2 in all range)

I try compensate this by add L0 coefficient for Short calibrate standart in NanoVNA Saver. But need different values for every harmonic (and on smith? pure SHORT calibrate standart look like arc).

I not have good knowelege in this, and don`t wont change this part of code while not understand it.

On Thu, Apr 23, 2020 at 11:38 AM, DiSlord wrote:

... Also i think need set default 'threshold' to 301MHz for H4 (on my H4 i don`t see any problems in any threshold range from 250 to 304MHz) it give better result for si5351 PLL change (all my test then i align band freq range, i made on this value) ...
===================================

Dislord,
I also needed to set the default "threshold" to 301 MHz for my H4. With a default threshold of 300 MHz my display looked like the attached when sweeping from 290 - 320 MHz. After changing the threshold setting to 301 MHz, the display became stable again with a slight discontinuity at 301 MHz.

An inexperienced user would not understand what was happening if their display looked like the attached while sweeping a narrow band centered at 300 MHz. They also might not be computer literate enough to use the "threshold" command from a computer terminal to fix it.

Would it be possible to add a CONFIG-THRESHOLD menu to allow changing the threshold frequency with using a computer terminal?
Exciting to hear you are considering supporting the microSD card option on the H4 for screenshot and waveform storage.
- Herb