LibreVNA Overview, Firmware Update, and Calibration
Вставка
- Опубліковано 23 січ 2025
- In this video, we take an introductory look at a newer Vector Network Analyzer, the LibreVNA. This particular model was sent to me free of charge from AURSINC in exchange for this video review.
Support TheSmokinApe Channel on Patreon Here: / thesmokinape
Links to products in the video:
Check out the LibreVNA here: amz.run/5UUz
As an Amazon Associate, I earn from qualifying purchases.
S Parameters explained: • S-Parameters and the N...
Check out my NanoVNA Playlist: • NanoVNA Playlist
Links from the video:
github.com/jan...
Check out my HAM Radio, Scanner, Walkie Talkie Playlist: • HAM Radio, Scanning an...
If you liked this video you can see more like it here:
/ thesmokinape
You can also follow me on Twitter and Instagram:
/ thesmokinape
/ thesmokinape
Nice review, it is always interesting to see what other people think of the device. As the developer of the LibreVNA, I'd like to point out a couple of points which I think might help understand the VNA/GUI a little bit better:
04:28 Harmonic mixing only comes into play above 6GHz (yes, you can actually go to about 8 GHz if you enable harmonic mixing in the preferences). Above 3 GHz the reduced dynamic range is mostly due to internal signal leagage.
11:34 The LibreVNA does not use the virtual communication port (as the NanoVNA does as far as I know), so it doesn't show up as a COM Port. Communication is handled by libusb, you don't need any specific drivers on windows.
20:49 The Line measurement is not needed for a SOLT calibration (and in fact, is not used at all during the calculations for that calibration). You only need that for a TRL calibration. I admit that the dialog can be a bit confusing here. If you try to activate a SOLT calibration before taking any measurements, you will be directed to a list of only the required measurements (no Line measurement).
21:54 Try double-clicking on a chart, it sould switch to displaying only that chart (double-click again to see all charts again). No need to close all the other ones, if you want to keep them around for later.
22:37 Quick tip: you can also drag-and-drop traces from the list on the left onto any graph. Might be faster than right-clicking on the graph and selecting them from the context menu.
If you have any further questions, you can reach me either by mail (see GitHub repo) or ask in the group (groups.io/g/LibreVNA-support). For bug reports, please open issues directly on GitHub.
Wow, thanks for commenting and sharing that information. Much appreciated!
I'm considering this vs the nano vna v2 plus4. Can the librevna do a port offset or electrical length offset? I intend to use this for measuring impedances accurately and perform matching of my circuits through a coax feed on my DUT. If my question is answered, I'm sold on librevna.
The group doesnt exist? :(
Thanks for the info.... 73' de K4WRF
I'm subbing you. been watching you a while. No BS, you explain simply and clearly, without dancing around the subject an talking for an hour without fully explaining...like most ham videos. Thanks
That's awesome, thanks Steve!
A few remarks on your calibration procedure:
* Others already remarked: keep connection to the onboard connectors as few as possible. Best practice: Insert a sacrificial female-male connector. Also, consider getting SMA standards that don't rotate the pin. So, to make it long lasting, have: Port1 := "Port1->SacrificialSMA-FM-Adapter" and likewise for Port 2 ;-)
* You want to calibrate to planes as close as possible to your device under test (DUT). That means you will never attach calibration standards to your VNA directly. You will calibrate to the *ends* of your coax cables. That "thru" you show, is not a thru. It's a line. "Thru" means: Line of zero length (cf. step 2 below for the fineprint ;-) ).
Your calibration should go as follows:
1.) Connect coax cables that are long enough to get to your DUT easily.
2.) To be really picky here: attach a female-female adapter to the end of one of the cables to enable a proper through. I.e.: Port1->Coax-| open end open end |-FF-Adapter->Coax->Port2
3.) Go through all steps *at the end* of your coax cables. I.e. the short goes to the open end of your coax cables and so on.
4.) The through measurement will in the proper case be just the two cable ends attached to each other. (I.e. you'll end up with Port1->Coax->FF-Adapter->Coax->Port2). If you did not follow step 2 (e.g. because your DUT has two female adapters and you're not really interested in super-nice cal) you'll still end up with this, but the FF-adapter won't be properly accounted for.
5.) "Line" should be a segment of known length (typically lambda/4 or so). If your standard does not have one, TRL or other methods requiring a line won't work for you. The nice thing about TRL (thru-reflect-line, btw.) is that you don't need to know the line length absolutely precisely. But for best results it should be in the 1/4 lambda (or 3/4 lambda) range.
When you're done, check your cal for example with a third coax cable and check for consistency. You'll maybe have 300mm cable, that should give you a certain phase shift at a given frequency and you can check if that's what you see on the VNA.
Yup, what you mention is the industry standard for calibrating your VNAs accurately. and correctly. Another thing I like to do (if I'm unsure of the cable qualities or they are new) is to terminate each of coax with a 50 load. On the smith chart, I should be able to see a dot in the centre (or really high >50dB RLS) up to calibration freq. I also like to test if the cables are faulty by wiggling in a little to see huge if there are huge changes in return loss. Typically won't be a problem for scalar measurements, but can be an issue when measuring impedances accurately at higher frequencies.
Very informative video, great job
Thanks for watching Thump!
Nice informative review, thank you. TRL is Through (Port 1 directly on port 2), Reflect (open or short on each port), Line (like the through but longer). This calibration method does not require exactly known test standards and is easier to implement on PCB than SOLT. It is most often used to get S-parameters right on the edge of a device soldered on microstrip lines thus removing the test fixture from the results. I does have quirks on its own though and requires care in design and measurement to work properly.
Thanks for that info gac 👍
I don't know if I have ever used TRL cal below 6GHz. Perhaps between 2 - 6 GHz it would be ok. The microstrip lines tend to get a bit long at these frequencies. The type of circuits you are measuring/characterizing would dictate what type of cal to use. Unless I am doing printed circuit type designs, be it antennas, filters, mixers, couplers/splitters, etc., I likely would never use a TRL cal.
Most of the standards I have used (Anritsu, Rohde & HP/Agilent/Keysight) have standards which are good up to about 8GHz. There are more costly cal sets used above these frequencies. For a HAM operator, I should think a TRL cal would likely never be used. Maybe a radio enthusiast who works on higher frequency things like microstrip filters, delay lines, distributed line matching, etc. would be more appropriate for TRL.
There are ways of "de-embedding" your measurements so you are measuring S-parameters right at the board level. It usually involves "dialing" out the electrical length of the test probe at the board. I have had many a discussion with newbie as well as seasoned RF designers about how to do this. Isn't RF fun?!?!
Trl is also good for really wide band calibrations, you just need to use multiple dely lines
Been on the fence for a few months...the freq range and dynamic range got me. Ordered today. I'll use SMA-savers or a jig out a very short lead to a BNC (like my DG8SAQ VNWA 3 does). But what would be killer is for someone to build a good "test set" so that two port, all S parm VNAs like this can auto-switch. Thanks again for a very organized, easy to watch review!
I appear to stand corrected! On the aursinc store on Amazon, this Q was asked about the Test Set. Here’s the vendor’s response:
“Yes the LibreVNA vector analyzer already contains the the equivalent of an S-parameter test set. A source signal is generated and alternately applied to the RF ports. The incoming signal at both RF ports is measured, resulting in the four S-parameters S11 and S21 (when the source signal is routed to port 1) as well as S12 and S22 (when it is routed to port 2).“
Hey Frank, hope you are doing well. Thanks for the info about the instrument 👍
Good video. The software on this thing has a lot of potential, and you can tweak it tremendously. thanks man
Thanks for checking it out, I read somewhere that you can use NanoVNA Saver with the device but I haven't tried that yet.
Cause you need more test equipment...
@@HamRadioClubhouse Well, yeah.
@@TheSmokinApe Interesting thought. I'll have to try that.
Son, please...
Having to use the adaptors for the cal kit means that when I remove them I get 0.5dB to 1dB oscillations from 1Ghz upward
Kurt Poulsen provides a calibration coefficient files and a guide how you should calibrate. Best practice would be to leave the cables connected and use the cal standards with adapters.
Fair enough, thanks for the comment Dragan 👍
Muchas gracias, greetings from Chile
Glad you liked it, thanks for watching 👍
I use a libreVNA for 4 months. It is very good VNA. but you need to add a heatsink or a fan to cool your libreVNA.
Yes, it does get hot 🔥
Hi, Can I use LibreVNA for WR284 waveguide measurements? WR284 cal kits come with waveguide to coax adapters & loads. Can I use SOLT part of software to calibrate LibreVNA before measuring s parameters of waveguide device? Please advise. Thanks.
Sorry, I have no idea.
Biggest problem I see is the calibration standards. I would throw them in the trash. They are going to destroy your SMA connectors on the VNA. The center pin on any connector should never be rotated. It wears out the female and male center contact and it only takes a couple of times to have a noticeable effect at high frequencies. Get a set of standards that have loose/rotatable outer nuts. Your connectors will thank you.
Thanks for the comment Mike. I have been using SMA Male to Female adapters for this exact problem, but your are correct it's an issue.
I haven't purchased a VNA yet. However, I am leaning toward the "N" connector version. It seems it would be durable.
Hi, Can you update the process for the new version V1.5.1, Thanks
Wish I could but I no longer have the device
Hi Ape,
Appears to be a better quality than the NanoVNA. Looks like you can measure and set markers easier. Does have a hefty price tag though. You and the family stay safe. 73 WJ3U
I think you are correct accross the board, thanks for watching Don!
An interesting product. Thank you for the review which was too the point. Is the heat-rise a functional issue or just a concern in your experience? This looks to be a moderately priced alternative for lab work or a more granular detail analysis?
The heat is just something that concerned me, it works great as a VNA 👍
@@TheSmokinApe Many thanks... De K4WRF
For the price class I would have hoped for N connectors. SMA is only rated at 500 cycles if you use the torque wrench.
Very fair feedback
Looking good, a little rich for my blood
First, I would like to say thank you for providing this video. I have been on the fence for some time, and as with many others, I utilize UA-cam and the experiences or demonstrations of other hams to assist in my purchasing decisions. I don't see that this unit is justified at its current price point (which is significantly more expensive as of 11/2023) over a nanoVNA. While it is likely more accurate than a nanoVNA (not totally convinced on that), to what degree? The nanoVNA has proven sufficient to date.
I can understand that perspective, thanks for watching BigJohn 👍
Do you find a vna more useful than a rig expert?
Much, because it has 2 ports and can do through measurements.
@@TheSmokinApe , @rbrazil1000, I'd say it depends. The rig expert is more of a field instrument. Portable and immediate results. VNAs require selecting the frequency ranges and measuring reference components. They provide return and insertion loss data with phase but at the cost of more training and test time. Except for the battery operated nanovna, most are difficult to bring out of the lab.
You have made a huge mstake during calibration.
You calibrated Short, Opn and Load at the VNA connectors but for the Through Cal you added the cables.
For your measurements you used the cables so their error contribution was not included in the calibration. That's why the reflection was so wavy.
The correct calibration should be performed with the cables included. That way, the reference measurement plane would be at where you connect your DUT.
T
Thanks for the feedback Itzhak 👍
Wow! Nice Review! .... Son, there was a USB cable included to hook up to a computer. :P Keep up the good work Ape! 73 - KF6IF
Haha, I have no idea how I missed that. Thanks for Phil 👍
@@TheSmokinApe No worries! Always enjoyable to learn with you!
Is Aursinc an official LibreVNA seller? Can we trust the genuinity of the device?
The Groups.IO group links to Aursinc as a place to purchase them.
A lot of coin for a VNA that gives different readings based on room temperature, humidity and up time. I hope they fix thermal instability issues soon.
Fair feedback
This project is flawed ... so don't dream just go further.
Every vna has these issues. When you look at keysights manuals for they half a million dollar machines they tell you to wait until the vna has warmed up. The lack of active cooling is a concern but every vna is sensitive to these changes. Your calibration is really only valid for about 2 hours as well. Ask lab techs that do precise measurements, they recal multiple times a day
@@Blake-jl8lh Well, it’s not just that it varies but by how much and how soon. Price implies comprehensive design where thermal concerns have been addressed and minimized. That is not so.
It looks powerful and accurate but well beyond my pay grade. lol. I am not advanced enough yet anyway.😄
It is a pricey unit, thanks for watching Don 👍
sdrsharp not the same as your last video
This device has its own software, I do believe that it can use NanoVNASaver as well.
@@TheSmokinApe thanks im just starting , well confused but trying
@@TheSmokinApe i tried to follow but when you downloaded in the new file ,, no bat file to download , could not find it
wish it wasn't so expensive
I am right there with you, check out my NanoVNA playlist for more affordable options.