开云体育

The last was dead on turn on shortly after RX so it went back, refund obtained and ordered a new version.
Not the white 'gecko' version this time.
Hopefully will fair a bit better.

72
Dom
M1KTA

thank you for sharing DFU file, now I can upgrade my nanovna, it would be interesting to put DFU fimware file in the file folder of the group

again thank you

Gilles

Alan,
Any links to where you found this photo?

It looks like what I did to my unit.
Those users with that own a NanoVNA that has only bare FR4 or plastic front can mount and wire-in their own separate buttons.

Now, if the jog switch could be made more responsive.....

73
Larry

Apparently someone is looking at the issue of the device mechanics.
This looks encouraging. Anyone looked at this device in detail?
Thanks,

Alan

Ok, thanks for the update, I will check it out.. I used anolder version..

Hi reuterr, great job !

Someone should update the wiki firmware page.
A summary/note of the various firmware versions available would also be very useful too.

Cheers,
Diego

Hello Pluto,
I have it converted under windows 10 with the DFUFileMgr.exe and added USB Device ID DF11.
program with (MacOS or Linux):
$ cd to file folder
# place jumper on boot0 pins - REBOOT
$ dfu-util -v -d 0483:df11 --alt 0 -D nanoVNA_0.1.0_ch.dfu
# remove jumper - REBOOT

73, Rudi DL5FA

Yes,

That is what I do

Hi All !

A question about the new 0.1.0-20190914 firmware release by edy555 ().
The zip file (nanovna-firmware-0.1.0.zip) includes a .hex release (ch.hex).

Is it possible to use DFU File Managre to convert it to DFU and upgrade the firmware by DFUseDemo ?

Thank you !!
Cheers,
Diego

It would be really nice if NanoVNA firmware had an identifier in the "info"
text that identified the software, not just the build time - for example to
be able to set reasonable start/stop frequency limits, as well as whether
to use the "scan" command or not.

--
Rune / 5Q5R

I did an update yesterday because I made an error in the stack setting. Build time (as you can see in the info screen) 15-9-2019, 10:13
For me the cursor is working now
When did you download?
Inside the config menu there is a dfu command to boot the nanoVNA in boot load mode. No need for jumper setting

David

I too initially questioned the quality of the RF Bridge, primarily due to the price. I did the same with the Nanovna and for the same reason. In both cases, overcoming the learning curve and careful calibration have led to results that put those concerns to rest. I use the devices for hobby work as well as in my business and I do not need 0.1 dB accuracy in either case.

Now I take the Nanovna by itself onto job sites instead of lugging the Rigol/RF Bridge/cable kit. The internal power supply alone makes field setups a snap. The Nanovna is in daily used here........ no time or need for firmware upgrades and research. It has become perhaps my most valued tool.

I would be curious to know what deficiencies you see with the RF Bridge versus the HP bridge but that is probably OT here.

Best regards,
WA8TOD
===========================================

Warren,

I think that at the time I bought the bridge more as a learning tool, and it was less than the price of a reasonable meal out. With now having the nanoVNA I see no use for the bridge. As you say, having the portability of both the nanoVNA and the FA-VA5 is the "deal-breaker", although I still use the Rigol SA/TG and the support and documentation on my DG8SAQ VNWA is second to none - this matters. For higher frequency work I have the AAI N1201SA which covers up to 2.7 GHz, so useful for QO-100.

All hobby use as I'm nominally retired.

I don't have an HP bridge - it was just a generic quality comparison term!

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

Hi Erik,

Thanks for the firmware update.
Have tested it a bit, and seems to be working fine, of course with reduced
accuracy in the high end.. (would be nice with some averaging now, but
guess there might not be enough memory for that)

But I think there is a bug using the jog-switch, when I move the cursor
with it, first time is ok, but second or third or fourth time using it, the
NanoVNA locks up. Only power off/on restores it. Also tried to use
NanoVNA-Saver when it was frozen, and as expected it did not manage to do
anything.


BR,
Askild

David,

As noted in your follow-up topic, you need to 'Normalize' the Rigol with the Open load is applied to the DUT port and the Calibration Load applied to the REF port. This will set Return Loss to zero for the entire range. Then, when a test load is applied to the DUT port the Return Loss can be read from the screen.

Warren Allgyer
WA8TOD

If this duplicates a prior post I apologize but, if so, there are so many new users that it may bear repeating.

The same DfuSe Demo software that is used to upgrade the device firmware can be used to save the state of an existing device setup. If you have your Nanovna completely calibrated with Channel, Format, Marker, and Stimulus setups on each of Save0 - Save4 and you would like to preserve that setup for future use, you can use DfuSe Demo to upload the existing firmware along with all settings to a new, descriptive filename stored on the PC. If you then change firmware, calibration setups, etc and you want to restore the device you simply load the saved file in the same way you load firmware upgrade. All settings and calibrations are restored as well.

I have several files stored with different setups and calibrations and can download them at will to any of my three devices. It is a very handy capability for the lab bench.

Warren Allgyer
WA8TOD

David

I have two NanoVNAs and two 1-3000 MHz RF bridges. I get identical results with all four. The VNAs are calibrated with one set of supplied OSL terminations. When I use the bridge with the NanoVNA I connect Port 1 to the Input of the bridge, Port 2 to the output, and terminate the REF port with the NanoVNA 50 ohm load. I then calibrate the VNA for "Through" only with the DUT terminal open. This sets the CH1 Logmag return loss to zero. From this point forward calibrated NanoVNAs give identical readings to the RF Bridge down to reasonable levels. Once Return Loss hits 30 dB and beyond I pay no attention to discrepancies because they are meaningless for the most part.

I use a similar setup when using the RF Bridge with my Rigol Spectrum Analyzer. Tracking generator output to RF Bridge Input, RF bridge output to spectrum analyzer input, NanoVNA reference load on the REF port and the DUT port open. I then normalize the spectrum analyzer to zero, connect the DUT, and read the return loss directly from the screen.

Up to 30 dB or so of return loss all four devices produce reasonably identical readings through 750 MHz plus or minus one dB.

WA8TOD
===============================

Thanks, Warren, your explanation helped a lot.

I think there were two issues.

- First I wasn't including the bridge in the SA normalise - hence the dB loss I was seeing even for open or short. Beginner's error.

- Secondly I guess I was expecting higher quality than this bridge actually is. It's not HP quality (or whatever they're called today), but it's not HP price! Fine for indicating resonances.

For resonance out in the field taking the Rigol is not the most convenient (!), and something like the FA-VA5 which covers up to 600 MHz is a robust and well-boxed unit.



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

On Mon, 16 Sep 2019 at 06:30, Rune Broberg <mihtjel@...> wrote:

It sounds like the kind of thing that would be patented. Maybe US6643597B1
is related?

Expired patents can be potential sources for how to do some of these
things.

--
Rune / 5Q5R

Yes, I could see a method like Quickcal could be patented. I can see an
obvious need to do that it you are producing hardware which is easily
copied. I am no expert, but it might be better not to disclose software
ideas, as you can bet that the Chinese will just ignore such patents.

I must admit to being confused by the US patent US6643597B1. I will need to
read it 100 times over to make any sense of it. At first glance it seems to
suggest that one doesn’t needs a calibration kit.


I don’t know what the position is regarding patents and non-commercial
products, or products that are sold commercially, but for which others have
developed GPL software. *Either way, I think the Chinese will ignore any
patent. *

The FieldFox instruments (perhaps with the exception of the brain dead
N9912A) don’t need a calibration kit to make measurements on the ports as
they are calibrated there. This was the post that got me thinking Quickcal
was implemented by a time domain transfer but looking at it again, I
don’t see why I assumed that



Quickcal relies on one having a good knowledge of the connector, as you
need to enter it - both the type and gender.

Dave
--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales.
Company number 08914892.
Registered office:
Stokes Hall Lodge,
Burnham Rd,
Althorne,
Chelmsford,
Essex,
CM3 6DT,
United Kingdom

I also have an identical blue-board bridge except my one does NOT have the two small tracks missing. I guess that only some batches of PC board were in error. Maybe they intended to fit these with two 49.9 ohm resistors rather than the 4 x 100 ohm?

The results from this board are disappointing. I wonder if it is the ferrite that is used? Often one sees short pieces of wire bent in the direction across the bridge to add a few fF of capacitance. This board does not use this trick, maybe it should!

Brian 5B4AHW

It sounds like the kind of thing that would be patented. Maybe US6643597B1
is related?

Expired patents can be potential sources for how to do some of these things.

--
Rune / 5Q5R

On Mon, 16 Sep 2019, 00:18 Dr. David Kirkby from Kirkby Microwave Ltd, <
drkirkby@...> wrote:

The Agilent/Keysight FieldFox range are in theory great analyzers, although
they are pricey. The N9912A is very poor though.

The FirldFox range have an interesting calibration method called
“Quickcal”, which requires no calibration kit - you just leave the cable
open. It is not as good as using a calibration kit, but it has the
advantage of simplicity if used in the field.

You need to enter the type and gender of the connector, then it does its
magic. For return losses of > 14 dB, it is beneficial, but not essential to
use a load too, but that load doesn’t need to be very good.

The method used is proprietary, but a bug reported to the Keysight forum,
and the answer supplied by Keysight, gives one a clue how it works.

Clearly the method had a model of the open connector, the mere fact that
one needs to enter the type and gender of the connector. The bug reported,
and the answer received, suggest that the method makes use of time domain
data. I am guessing that the FieldFox measures the time to return from the
discontinuity in impedance.

The method works for N, 7-16 and female SMA or 3.5 mm. It is not
recommended for male SMA. It doesn’t work for APC7, but I suspect that is
simply because Keysight do not believe people will want to use a rough
calibration method on a precision laboratory connector. An APC7 connector
would be very easy to model.

Long term, it would be a useful addition to the NanoVNA firmware, but it
would be well down the priority list.

Dave.


--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales.
Company number 08914892.
Registered office:
Stokes Hall Lodge,
Burnham Rd,
Althorne,
Chelmsford,
Essex,
CM3 6DT,
United Kingdom

On Mon, 16 Sep 2019 at 02:27, alan victor <avictor73@...> wrote:

Hi Tom,

Very good.

The paper published by Copper Mountain applications was irrespective of
frequency.

The paper did however assume all parameters of the VNA are independent too:

Directivity 50 dB
Source match 46 dB

etc. Those parameters are *very* good, and most unlikely to be achieved
with the NanoVNA, but will *not* be independent of frequency. One can
expect the performance of the load used in the calibration to be poorer at
higher frequencies. One can expect the directivity of the bridge to be be
poorer at high frequencies. So the fact Tom’s results deteriorate at high
frequencies is not surprising to me.

Whilst it is possible to purchase very close tolerance resistors, the same
is not going to be true of inductors or capacitors. I use some 0.005%
resistors to check my HP 4284A LCR meter. They are readably available.
They are 10 times better than the basic accuracy of the LCR meter (0.05%).
But it’s impossible to buy capacitors of that tolerance, and for inductors
the situation is even worse

Alan

Dave

--

Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales.
Company number 08914892.
Registered office:
Stokes Hall Lodge,
Burnham Rd,
Althorne,
Chelmsford,
Essex,
CM3 6DT,
United Kingdom