开云体育

Hi Larry,
You can also add solder points to the SMA sockets. On the black NanoVNA they are welded on both sides, but only one PCB side is soldered on the white NanoVNA.
David, F4HTQ.

I have been hacking my Gecko unit to add shields and I also decided to add an SD card slot I removed from an old Blackberry cellphone.
As I mentioned in a previous post, old 2-way radios and cellphones have a treasure trove of re-usable parts.
The first photo shows a pile of shields from several old Motorola HT1250 portables. Cellphone shields are similar.
I got the TF card slot from an old Blackberry.
Use a heat gun to remove them as well as switches.
TIP: Heat the opposite side of the PCB when removing parts to prevent heat damage.

The second photo shows a closeup of the installed shields - big difference in noise levels!

The third and fourth photos show the TF card slot where I had to scrape some ground plane solder mask for securing it.
In the upper left corner, you can see the push button I installed on the BOOT solder pads.
I know the firmware doesn't support the SD card - yet - but that is what hacking is all about!
Now, I'm waiting to receive the Black (Original design) version with the USB-C connector .
BTW, that old Motorola cellphone battery lasts well over 4 hours.

Hi Ernst,

I have updated with nanoVNA_800_aa_20190705.dfu and no more spikes are present.
Thanks.
Georges.

I have not taken enough 1 port data between the two to state their measurement delta if any. As far as I can tell, the white unit is no different electrically other than the firmware, 2 track display vs. 4 track. Hence the pixel size array on the 4 track firmware vs. 2 track is somewhat finer. My colleague and I are going to update the firmware in the white one so 2 channels are engaged; S11 and S21. Currently, I think I will do some 1 port measurements, the worse (best) test case is measure a 1 dB pad shorted to provide a 2 dB return loss and compare their 1 port responses. That stresses the Gamma value and puts it out at 0.86 reflection coefficient.

________________________________
From: [email protected] <[email protected]> on behalf of Greg Ashley <Gregory@...>
Sent: Tuesday, July 9, 2019 10:51 AM
To: [email protected]
Subject: Re: [nanovna-users] New nanoVNA received

Is this black one, better than the white one? If so, do you have a link to buy it off the ebay ?

-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of alan victor via Groups.Io
Sent: 08 July 2019 19:16
To: [email protected]
Subject: [nanovna-users] New nanoVNA received

Just received my 2nd nanoVNA. This is the black case, no Gecko white which works current to date quite well with 2 track display. This black case arrived ON TIME and well packed. Apparently held it's charge all the way through shipment and booted with a 4 track display.

These instruments are little like a box of cracker jacks or is it chocolate, your not quite sure what's in the box. So far, so good. This unit came in its own plastic protective case with all the accessories enclosed. The construction looks very good and robust. The ON-OFF slide switch and USB-C connector are more robust. There is a screen protector film over the display and no scratches on the display from a stylus being used on the soft screen for pretest. The style of the 50 ohm load is DIFFERENT. The short and open appear the same as with prior cal kit. Is this 50 ohm termination superior? Don't know. The prior was ~ 18 dB best case at 900 MHz. Very important, IT IS CLEAR THAT THE ASSEMBLY OF THIS UNIT HAS THE RF SHIELDS OVER THE BRIDGE AREA AND MIXER AREA AS SUGGESTED IN THE EARLIER POST. ALSO, THESE SHIELDS LOOK LIKE DIE CUT BRASS WITH PRECISION FIT. THESE WERE NOT CUT OUT WITH TIN SNIPS OR A SCISSOR AND HAND SOLDERED IN. THIS LOOKS LIKE A PRO JOB. Hurrah!

There is a green film strip sticking out of the left side of the display. Looks like the kinda of thing you would pull to engage the battery contacts on a new toy to protect the battery during shipment. However, I have not touched it and the unit turns on out of the box. The display has the 4 color traces and the pixel appearance is smaller and crisper than the white gecko unit.

So I will wait for a charge cycle to complete and see how this one plays. Hope this info is of assistance. Alan

Oh well, I guess I spent money for naught?? Maybe I'll save them for another project. I'll check the date code, they still could be old stock
Thanks
Frank

I looked at the part, and they are 602. I'll replace them and add shield cans (there are none) and I paid $75 for this one.
Oh well live and learn

Hi Georges,
those peaks are from the ocsillator circuit. Hugen - great thanks to him for immediate action - has released another firmware nanoVNA_800_aa_20190705.dfu and nanoVNA_800_ch_20190705.dfu which solve the problem,
Firmware, Software and manuals can be found here:

I strongle recommend to update to the latest firmware.

Regards Ernst

Hi Ernst,

I have the exact same 300MHz, 600MHz peaks. I'm actually trying to full charge the battery before making any measurement. I suspect the LiPo switching battery charging circuit to generate this harmonic. I will see if I have the same signal signature right after a full battery charge.
I have the white Gecko version on which I added 3 cavities over the mixers circuits. (see my other post with pictures).

Regards.

Georges.

My white version. I added the RF shield cavities using kapton tape to isolate the electronic and coper foil as RF shield, soldered in each corners. No more signal variation on the mixers when approaching or touching theses sensitives regions.
Mixers used SA602A.

Georges.
VA2GH

Is this black one, better than the white one? If so, do you have a link to buy it off the ebay ?

Thanks, so much. I got my nanoVNA working on my computer now.

Who knows who... could, perhaps take a paid commission to make one of these nanoVNA systems but with more data points ??

Greg

This one sounds exactly like the one I bought from Bangoods

I 'thought' the factory making SA602 had burned down ages ago so all chips manufactured were actually SA612. So will be the same chip (especially if the SMT version).
Info here although pretty sure spoke to Paul Harden about it at the time

Anyway the datasheet is here

72

Dom
M1KTA

I looked at my white Gecko, and they are 602's, so I just ordered 4 to replace my 3 (1 spare). They were $1.41 each,, bus ground was $9.26 from?? Newark

I have a white Gecko unit - and I have ordered another unit from Hugen via
taobao which is waiting to ship.
I agree there are issues with the bridge.
Above 300MHz return loss measurements become quite noisy at only around
15dB RL.

I think using an external bridge connected between port 1 and port 2 to do
the RL measurement will work much better.
In that case the calibration would need to be done on a PC instead of on
the unit (without firmware mods)

Roger

Thanks for sharing. Quite interesting.

On Mon, Jul 8, 2019 at 11:02 PM, <f4htq@...> wrote:

Dear Hugen,
I own the two versions ( your designed version, black, with battery and
shielded entries, and the "clone" white version ).
I noticed that the mixers used in white version are SA602A instead of SA612A.
Do you think that will be a issue ? have you put SA612A mixers on your model ?
I have a huge performance differencies betwen the black and the white version.
at 900MHz little more than 50 dB of dynamic range subsist with the black
version but no more than 30dB with the white version.
Do you think that can be only explained by the missing shielding ?
(However when i made some test on HF and VHF frequencies, the results seems
quite identicals).
regards,
David, F4HTQ.

Hi,David

NE602 = SA602 = NE612 = SA612


Adding shields should improve performance, but some bad modifications on the PCB can also reduce performance.

Thank you！

hugen

Nicely done, thanks for sharing.
Stuart K6YAZ

Here is a 9 MHz crystal filter which is matched to 50 ohms. The response is in good agreement despite not re calibrating over the proper passband range; say 8-10 MHz. Instead I used the 50 kHz - 900 MHz cal and reset the stimulus range to cover a narrower passband. It would be best if you cal with prior knowledge the filter center and span and not task the interpolation routine. The update rate between the PC and the COM port on the nanoVNA is fast enough to permit real time tuning of the Z match networks. I set the refresh rate to 800 mS. Alan

Ahhhh... Great and thanks!

________________________________
From: [email protected] <[email protected]> on behalf of dk1vi <dk1vi@...>
Sent: Monday, July 8, 2019 11:12 PM
To: [email protected]
Subject: Re: [nanovna-users] New nanoVNA received

Alan,
the green strip you mentioned belongs to the protector film. It should help you to to pick it off, after removing the front plate.

Ernst