NanoVNA-F V3 Review/Teardown, Compared to V2

[@KerryWongBlog]

Please turn off "Stable Volume" in the video playback setting (the first menu item after clicking the gear icon) as for some reason, this new KZbin playback feature made the video much noisier than usual.

[@jspencerg]

I appreciate your consistently thorough reviews. I'm considering getting vna with a higher frequency range than my v2plus4. I see I should consider this model in addition to the litevna. Could you do more comparison between this and the litevna?

[@dondonaldson1684]

Thanks Kerry, excellent presentation.

[@wiwingmargahayu6831]

your telco engineering skill is amazing Sir

[@rickfletcher4297]

Excellent review and presentation! I've placed an order for one.

[@fernandorosa9850]

Excelent video!. Could you made a comparision with liteVNA 64?.

[@craigrotay3732]

Hi Kerry! Thank you for getting the V3 so fast and producing another informative comparison. For the spectrum analysis, the output power for the V3 seemed to be set about 6dB higher than the V2 at 1GHz, and close to 10dB at 4.1. If you run the V3 at lower power, does it look a little cleaner? Or is the analyzer getting hit too hard (probably not...) Kind of surprised about the V3 spectrum, but still it's amazing for a low cost VNA.

[@KerryWongBlog]

I am a little surprised by the V3 spectrum as well. The spectrum looks similar at different power settings. It is slight better towards lower frequencies but in general the phase noise is more pronounced than V2.

[@craigrotay3732]

@KerryWongBlog Great information Kerry. Thank you. I knew you probably checked this aspect thoroughly since it's so unexpected. I suppose one good thing is that its output power seems to be a bit higher than the V2. Do you suppose there is a fix?

[@KerryWongBlog]

​@@craigrotay3732I was a little puzzled by the phase noise as the phase noise of MAX2780 is pretty low. Unfortunately, the software used in the NanoVNA-F is not open source so it would be hard to pinpoint where the issue lies.

[@joseppuig925]

Probably they didn't implement the MAX2780 correctly (loop filter component values, for instance), or they are not sending the correct configuration for its internal registers, hence, the bad phase noise.

[@se2702]

Return loss (RL) measurements are positive (+) for passive devices. If a negative sign is indicated (-) the measurement is really reflection coefficient characterized in dB.

[@n1vca]

Brilliant review - thank you very much - it answeres every question I have had

[@ernestb.2377]

Hi, would you know how these V2 and V3 roughly compare to H4? There are a lot lot of types, versions and producers these days...

[@Chris-hy6jy]

Hopefully they'll release a v4 with a better signal generator. A sine wave generator would be nice 😉

[@gonzoruiz7154]

Hello, thank you for the video which is very interesting. I have the F V3 broken and I'm looking for the schematic and I can't find it, if you know where could I find it? Thanks a lot.

[@jankromka9584]

Thank you Kerry.👍

[@no-ee]

It seems it v3 has lower noise floor and therefore better dynamic range. Right?

[@worroSfOretsevraH]

I'm wondering for how long will the center pin last of those connectors with all the screwing-unscrewing done.

[@madmax2069]

Yeah, SMAs are pretty fragile, you would have though they would have used a more robust connector.

[@vincei4252]

I just checked quality TE female SMA on DigKey and they list 500 cycles. Some are much, much worse i.e. fewer cycles. Better to have some sacrificial cables attached I suppose. I never understand instrumentation that could be used heavily having SMA connectors TBH.

[@madmax2069]

@@vincei4252 I'm more partial to using a solid adapter instead of an adapter cable that way you still keep the plugging and unplugging cycles to a minimum on the connector on the unit. With a solid adapter you don't have to worry about the cable being just another point of failure, but there again the cable can also keep forces from directly effecting the connector on the unit.

[@Gengh13]

It is pretty easy to replace when they fail.

[@madmax2069]

@@Gengh13 to some it's easy, to some they see using chewing gum to stick two spoons together as hard.

[@techr9186]

is your noise floor at S21 port still remains around -70db? mine went up to -50db, (-40db at 6ghz) in one week, no idea why

[@KerryWongBlog]

Well, mine is not that uniform. towards the low frequency end and high frequency end (up to 6.3GHz) the S21 noise floor is roughly at -70, but between 1.2 and 4 GHz there are some peaks around -60 dB. The worst is around 2 to 2.5 GHz which was around -50 dB.

[@ingvarandersson9950]

Thank you, for a good video. Also the mimo-antenna yo where using was intresting, do you have more detail of that one, what brand and model?

[@KerryWongBlog]

The mimo-antenna is from Waveform. I've done a few more detailed videos on them, you might wanted to check out: kzbin.info/www/bejne/rJ6XmYlnfrmlppI kzbin.info/www/bejne/qpSQZWx7qqeEb6M kzbin.info/www/bejne/nKbZXoGYfNCZqNE

[@mamatabb4465]

Hello! Thanks for the video, I have same unit. I’m wondering if you can make a video on how to test a coax cable? Thank you

[@amirb715]

@4:20 I dont see any BW on the screen but in all VNA's higher BW >> faster sweep. That's not surprising. That's how it should be. TO compare the sweep speed you need to make sure they have the same IF bandwidth. It is very similar to RBW in SA

[@KerryWongBlog]

I don't think adjusting bandwidth is an option on the NanoVNA-F V2 or V3 I know NanoVNA V2 Plus4 supports it, but didn't see it on -F V2 or V3. I will comb through the settings again when I get a chance, but if you find the settings please let me know.

[@argasasrgetthe9032]

You mentioned at 3:08 there was a firmware update available for the V2 to increase its frequency range to 4.4ghz. Is that an official update anyone can get? And is it even any good at lets say 3.5ghz? Nice review Love your Video's!

[@electronics.unmessed]

Very good! Thx for sharing.

[@robust5615]

do the position of antenna ..horizontal or vertical will effect its performance ?

[@Игорь-щ9б7с]

Hello! How to make a screenshot like it’s done in LitieVNA 64?

[@ernieschatz3783]

I wonder why the signal quality of the SigGen took such a step backwards.

[@AlexanderBukh]

Thank you, so informative.

[@JAKOB1977]

Are you sure that you ain't confusing V2 with the V3 and perhaps mixed them up when you tested the freq output.? You emphasize that your V3 is identical on its output as the V2 and it will max do 4.4G..you sadly don't show revision info on your V3 so difficult to compare what firmware that are on your V3. I also recived one of the V3 a few weeks back, and same maker Sysjoint with SeeSii prouctlabel, and that unit will gladly go to 6.3GHz on its signal gen out.. it will go from 10K to 6.3G if I recall correctly, though since its an MAX2870 synth chip.. the output RF selectable power, are locked below 23MHz. What V3 firmware are you on (says so in the bootup sequence), since you are limited to max 4.4GHz on your V3? 16:04 One thing I missed in your video, where the output power..how well does it match, with the labels 0dBm, -3dBm, 6,dBm, 9dBm.?..on my RF power meter, it was around 2 to -3dBM off, so fx -9dBm setting at fx 1G it was around -11dBm.. did you check it on your spectrum analyzer? Thx for the video.

[@WR3ND]

Great video. Thanks.

[@SpinStar1956]

Appreciate you showing us this but one thing concerns me is losing the lower 50 KHz capability, as I've made comment before. Yeah, you can say that anyone buying a 6-GHz unit doesn't care, but I worry that we'll never see newer and more advanced units that have good low-end. It is nice to have this real low-end for us guys that like to mess with design of VLF, MF and HF gear using low-frequency IF's and filters... 73!

[@hinz1]

NanoVNAs are great for price, but the lower frequency band sucks, compared to like a HP4195A Would have been awesome, if they had used downconverters/mixers within the native frequency range of a ADF4350 to get a nice sine wave stimulus signal down to DC, similar how HP used to do it. Instead of down-divided squarewave from ADF4350 and Si5331....

[@Chris-hy6jy]

Have you seen the SV4401A or SV6301A? Not sure if these are just a larger screen version of the nanovna or whether they have better performance. Would be good to see a review by you!

[@KerryWongBlog]

Good timing! I just got an SV6301A and will review it shortly.

[@Chris-hy6jy]

@@KerryWongBlog oh fantastic! Thanks 😂

[@snakezdewiggle6084]

The reflective display will drive you insane. , ;)

[@anlpereira]

I'm dreaming with just a V2 version to measure PCB antenna of my ESP32 projects. 🙃

[@gino.avanzini]

great video! How does it compare with the liteVNA?

[@KerryWongBlog]

It's a lot faster for sure. I haven't done any deeper analysis yet but the noise floor of V3 is lower especially in higher frequency ranges.

[@gino.avanzini]

@@KerryWongBlog thank you!

[@ТарасКорж-г4т]

So if its built-in signal generator has harmonics at -35dB from the signal, then this means the accuracy of network measurements below -35dB cannot be trusted. This would affect all S-measurements.

[@uploadJ]

Not if they do a hetereodyne conversion before detection. A simple broadband detector diode of course "sees all". But a detector diode is a scalar analyzer, not a vector (phase information) analyzer.

[@marekchmura-hanik5152]

@@uploadJ That is true. The erors start at ca -30 dB and reach -17 dB at -60 dB. Some improvement in 0.5.5 firmware but still at least 10 db error.

[@fushizhang6864]

Does the interface allow the lowest freq from 9khz?

[@KerryWongBlog]

No, the V3 starts from 1 MHz.

[@JAKOB1977]

​@@KerryWongBlog (Will dual post the question here also, as its name directed and it would be a shame if people get wrong info from video) Are you sure that you ain't confusing V2 with the V3 when you tested the freq output.? You emphasize that your V3 is identical on its output as the V2 and max 4.4G..you sadly don't show revision info on your V3..I also got one of the V3 a few weeks back Sysjoint maker with SeeSii label branding, and that unit will gladly go to 6.3GHz on its signal gen out.. it will go from 10K to 6.3G, though since its an MAX2870 synth chip.. you will first get output power control above 23MHz. What V3 firmware are you on, since your limited to max 4.4GHz on V3? 16:04 One thing I missed in your video, where the output power..how well does it match, with the labels 0dBm, -3dBm, 6,dBm, 9dBm.?..on my RF power meter, it was around 2 to -3dBM off, so fx -9dBm setting at fx 1G it was around -11dBm.. did you check it on your spectrum analyzer?

[@KerryWongBlog]

​@@JAKOB1977 The firmware on mine is 0.5.0, I haven't updated. Sounds like there were some improvements in later firmware versions. If you watch some of my later videos using the V3 (e.g. testing Werbel microwave combiner) you will see that the power measurement was quite a bit off (off by around 5dBm), which is a pity as the V2 worked perfectly.

[@JAKOB1977]

@@KerryWongBlog I will take a look Kerry.. my unit is rev B fw 0.5.3 and here you do have the full output from 10K to 6.3GHz.. I believe you can find the firmware on sysjoint..if you decide to update, and check the RF power again, and if things has changed.. pls report back... again thx for the vid..

[@Aimsport-video]

“Better spec’d and more capable components” that deliver inferior signal quality. Hmmmm.

[@ТарасКорж-г4т]

Yeah, big concerns with the accuracy of measurements because of a noisy incident signal gen.

[@marekchmura-hanik5152]

@@ТарасКорж-г4т The accuracy i worse than in F "V1". 60 dB attenuator displayed as 43 dB. With new 0.5.5 firmware bit better - as 50dB.

[@pchunter168]

👍

[@uni-byte]

Wow!!! that SA is noisy! Nice review.

[@marekchmura-hanik5152]

It is. I wonder if the high noise floor impacts calibration. Resulting in false readouts with errors of ca 15 dB at -60 db S21 readout in my piece.

[@JuulCPH]

You really shouldn't be promoting products like this. This is using the NanoVNA name but they almost certainly do not have permission from ttrftech who developed and marketed the original NanoVNA. Even worse (in my opinion) the firmware source code seems to not be available even though it seems very likely that it is based on the original NanoVNA firrmware by ttrftech which is licensed under the GPLv3. So not just a trademark violation but a GPL license violation as well. Copying and improving open designs is good but pretending to be the original creator by using the same name and then not sharing back the improvements made to the open source firmware is not just unethical. It's illegal. You're promoting what is almost certainly an illegal product.

[@robust5615]

it is open source or no ?

[@eggmangoogle3146]

​@robust5615 if you think that makes it legal to violate copyright or trademark, you don't know how open source works.

[@robust5615]

@@eggmangoogle3146 open source or no..Answer this ..not put your word into my mouth ...

[@louisho2522]

EEVblog Hi lol