开云体育

Pretty much the only way around it is to root your phone and install a 7.0 or
higher rom from XDA or one of the other sites that offer un-official upgrades.

.. or rebuild the app, specifying support for older Androids.
Build tools warns about most dependency issues.

Hi Larry -

Well, I have seen equipment where there is an enclosure covering electronics
on one side and a full metal shield on the underside of the PCB which is
screwed to the top enclosure through the board.

I >>suspect<< there might be worthwhile improvements on shielded nanoVNAs
by lining somewhat taller shields with this material.

My hope is for eventual smaller Bluetooth instruments
that delegate control and display to e.g. mobile apps.
Apple supports generic virtual COM port devices by Bluetooth, but not USB.

Given problematic power inverter noise, briefly shutting those down during each ADC sample
and running from capacitor storage would be more feasible if not powering a display.

1) Repository itself does not includes ChibiOS( git pull -f -t -p --all --recurse-submodules --progress )
2) U need ARM cross compiler( arm-none-eabi-gcc )
then, "make clean && make"

I did misread the cloned files - the C code is NOT in Python folder.

I like to modify my post

Is there a tutorial on how to recompile the C code after cloning it from repository?
I am getting many "missing references" so far.
I found "required " libraries and will download them using standard apt install ,
however SOME libraries are of type "library_dev" which would be nice to know beforehand.

On Fri, Nov 1, 2019 at 03:04 AM, Raymond Domp Frank wrote:
Please see images. 0V at bottom graticule, Vertical scale 1V/div.

Raymond

On Fri, Nov 1, 2019 at 02:55 AM, QRP RX wrote:

But this is actual for VERSION screen only,

You're right, of course. I showed the battery voltage display screen at the same time.

Raymond

On Fri, Nov 1, 2019 at 02:47 AM, QRP RX wrote:

vbat_offset command is implemented

I read about that. Nice but totally superfluous for my NanoVNA, as you may have read. In my NanoVNA, deviation is about +30 mV for all voltages between 2.90V and 4.20V battery voltage.
New battery icon will look quite sophisticated!

I just added LED ON before VBAT measurement and LED OFF after battery measurement in the code and measured pulse duration with oscilloscope

I looked at the duration of the Vbat loading. I'll place my images ASAP. Never done that in this group before and it's almost 3AM where I live so other priorities...

Raymond

Ah... I think I know where you got 100 ms :D

When you open CONFIG => VERSION it performs VBAT sample in a loop with 100 ms delay :D

But this is actual for VERSION screen only, when it works in usual mode, it sample VBAT once at the end of each sweep (once per second).

On Fri, Nov 1, 2019 at 03:39 AM, Raymond Domp Frank wrote:

Did you look at the actual waveform or derived it from the source code?

I just added LED ON before VBAT measurement and LED OFF after battery measurement in the code and measured pulse duration with oscilloscope :)

On 10/31/19 6:41 AM, Yaya via Groups.Io wrote:

was wondering if the nanoVNA can somehow be used to perform IP3 and IMD measurements of a transceiver?

???? Maybe.? For measuring IMD of a receiver, you need at least 2 relatively clean signal generators, and a combiner that isolates the generators from each other so that they don't pull each other off frequency, generate intermod inside the generators, etc. Signals generated by the nanoVNA might be clean enough on some frequencies, or you might need to do some filtering to clean them up a bit.? And some IMD measurements require a modulated signal with control over the modulating frequency and modulation depth, which the nanoVNA doesn't provide.? For things like measuring the IMD of a SSB transmitter, you need a good two-tone generator, a power attenuator on the output, and once you have that you could probably measure the level of all the tones and the generated IMD signals, and do your calculations from there.? At every step you would have to do things to convince yourself that whatever intermod products you are seeing are being generated in the equipment under test, and NOT in the test equipment.? So, you MIGHT be able to use the nanoVNA for part of the equipment to do your tests, but it would probably be a fair bit of work.

On Fri, Nov 1, 2019 at 01:18 AM, Raymond Domp Frank wrote:

Offset as fixed in the firmware, version 0.4.2-bf9c4ba, Oct. 28, 2019

vbat_offset command is implemented and available in source code of NanoVNA-Q

But I didn't released firmware yet. It will be available in NanoVNA-Q 0.4.3.

In addition it will include a new and colored battery icon with higher resolution (it has 8 levels, while current firmware battery icon has just 3 levels).

On Fri, Nov 1, 2019 at 02:31 AM, QRP RX wrote:

I have no idea where 100 ms was taken, just tested on NanoVNA-Q firmware.
VBAT is sampled every 1.00 second during about 1.2 ms. So, the duty cycle is
0.12 %

Did you look at the actual waveform or derived it from the source code?
My measurements definitely show a 100 ms period for the pull-down on Vbat. I'm sure, just checked my photographs. I'll try and post them tomorrow.

Raymond

On Fri, Nov 1, 2019 at 02:23 AM, QRP RX wrote:

Vbat is sampled every 100ms during about 1.2 ms

I should modify this (my) statement: "The cathode of D2 is loaded toward GND every 100 ms during about 1.2 ms. This can be seen by Vbat being reduced by a few 100 mV".

Thanks a lot for your comment, QRP RX.

Raymond

I have no idea where 100 ms was taken, just tested on NanoVNA-Q firmware.
VBAT is sampled every 1.00 second during about 1.2 ms. So, the duty cycle is 0.12 %

I used full shields from some scrap HT1250 radios on my first (Gecko) NanoVNA but never really had time to test it. There are photos from the end of June of my work, in the photos area of the forum.?
I like your idea of grounding all unused pins, after tri-stating or setting them to gnd through the firmware.?



On Thu, 31 Oct 2019 at 8:50 PM, Raymond Domp Frank<hewpatek@...> wrote: On Fri, Nov? 1, 2019 at 12:02 AM, Larry Rothman wrote:

how much digital noise reduction do you think would occur by placing bypass
caps on all the unused uP pins to gnd (of course).

Not exactly sure if I understand what you mean. I'm not familiar with these processors but I'd prefer not to bypass them but just short them to ground if firmware could set them appropriately during startup (either low or input). I cannot quantify the improvement but just shorting them to ground is free so no reason not to do that.

I'm certain that PCB layout offers a lot of room for improvement. Full shielding (capping) may not be much better than just placing "walls".

Raymond

On Fri, Nov 1, 2019 at 12:11 AM, Raymond Domp Frank wrote:

Vbat is sampled every 100ms during about 1.2 ms. That's a duty cycle of about
1,2%. That means, only 1.2% of the dropout voltage is visible on average.

The measurement of VBAT happens once at the end of each sweep cycle. So, it happens about once per second.
Before measurement internal divider is activated and after measurement internal divider is deactivated.

@edy555 and @hugen79 firmware takes 2 measurements (1 VBAT + 1 VREF).
NanoVNA-Q takes 64 measurements with average filter (32 VBAT + 32 VREF) with a total measurement time about 1.24 ms (620.5 us + 620.5 us).

The filter was added, because 1 measurement is too noisy.

Well, I have seen equipment where there is an enclosure covering electronics on one side and a full metal shield on the underside of the PCB which is screwed to the top enclosure through the board. Vcos and synthesizers are built like that for high-end radio gear.??You know, it's nice trying to squeeze out as much performance from the Nano but it reaches a point of diminishing returns.??Personally, I'd like to see how the stm32f303 work is progressing.?



On Thu, 31 Oct 2019 at 8:42 PM, Oristo<ormpoa@...> wrote: Hi Larry -

the material you got has about 3db atten around 400mhz.

Thicker 3M material with attenuation at lower frequencies was lots more expensive,
and attenuation seemed mostly proportional to thickness.
I did not expect much to happen at 12kHz..
The material appears to have no appreciable effect from 110kHz - 300MHz,
but does help from 500-1000MHz.
Whacky stuff happens 1000-1100MHz, then material helps until 1200MHz or so.
My unit does not calibrate usefully, with or without EMI material, above 1200MHz.

No, sorry. The latest has the USB-C cc resistors and I think the bridge has changed slightly.



On Thu, 31 Oct 2019 at 8:27 PM, QRP RX<qrp.ddc@...> wrote: On Thu, Oct 31, 2019 at 08:32 PM, Larry Rothman wrote:

Check the attached schematic.

Do you have schematic for V3.1 PCB?

On Fri, Nov 1, 2019 at 12:02 AM, Larry Rothman wrote:

how much digital noise reduction do you think would occur by placing bypass
caps on all the unused uP pins to gnd (of course).

Not exactly sure if I understand what you mean. I'm not familiar with these processors but I'd prefer not to bypass them but just short them to ground if firmware could set them appropriately during startup (either low or input). I cannot quantify the improvement but just shorting them to ground is free so no reason not to do that.

I'm certain that PCB layout offers a lot of room for improvement. Full shielding (capping) may not be much better than just placing "walls".

Raymond

Hi Larry -

the material you got has about 3db atten around 400mhz.

Thicker 3M material with attenuation at lower frequencies was lots more expensive,
and attenuation seemed mostly proportional to thickness.
I did not expect much to happen at 12kHz..
The material appears to have no appreciable effect from 110kHz - 300MHz,
but does help from 500-1000MHz.
Whacky stuff happens 1000-1100MHz, then material helps until 1200MHz or so.
My unit does not calibrate usefully, with or without EMI material, above 1200MHz.