Using the original NanoVNA to measure PDNs

[@ruhnet]

Great stuff, Joe, as always!

[@therfnoob7697]

wow! So many useful informations! Thanks a lot man. Keep up this great work!

[@PY4SR]

Very informative, Joe. Thank you for sharing

[@Wtf95]

Awesome summary faq! Thanks.

[@joesmith-je3tq]

Glad you enjoyed it.

[@irpheus]

Outstanding work, Joe ... and so many thanks for sharing!! Jesper M

[@jspencerg]

Found and watched this video after my question re:cmt at the 61e+video. This was great. I have same curiosity as other commenter, HH, on the equipment. I'll read manual as next step. I appreciate how you're creating experiments that demonstrate effects of component selection and circuit construction on performance. I'm studying and doing labs to learn EE fundamentals. No great goal right now but your nanovna work encourages me to study antenna theory. Thanks! Maybe I'll be ready for a nanovna in 2022 after I understand things a bit more. I'm understanding details more each time I go through video and manual(text). Building foundation.

[@joesmith-je3tq]

>2. What Z characteristics did you build your common mode transformer to? These are designed for a 50 ohm system. This video shows some of the measurements I had made with this core and how it compares. If someone shows me a black toroid with some coax, it has no value. I can use Google images if that's all I am after. I want the material selection. The rest, I can figure out. A good reference book is the ARRL handbook. Good luck with your new learning adventure.

[@dzidmail]

34:16 Why don't we ever see resistors in series with bypass capacitors on a typical smt PCB - with say a microcontroller? (Or do we?) It also goes against intuition as we want to have unrestricted flow from the capacitor to meet a current demand of a digital circuitry.

[@joesmith-je3tq]

You pick cap types with the ESR you are looking for.

[@jstro-hobbytech]

Hey Joe I apologize in advance for the stupid question. I'm not into rf so I thought that pdn with the nano vna was for testing antenna and such. It reads like its really used for testing any digital network like on a pcb in development to make sure power delivery is behaving properly or to troubleshoot the same. Like finding problems of unexpected inductance and stray capacitance? Once again I apologize. You are on a level that I'm sure even your peers are not at.

[@joesmith-je3tq]

The idea is to design the power distribution network for a flat impedance. What is being showing is an attempt to measure it with the NanoVNA.

[@IZ4HDM_Chris]

Hello, I'm interested in the cores you used. What kind of material? How many? Have you added discrete components for flattening the response? Thanks as always for your great value content.

[@zktec1]

I've noticed you use only male SMA connectors on your test fixture PCBs. Was there a special reason for that? Perhaps a reliability concern over the female connector?

[@joesmith-je3tq]

No reason really outside of the standards are male and I use a stainless F/F or coax to F connector. In this video, you see me using coax to M as well to connect to the breadboard.

[@kennarnett8220]

Hey, great video!I have a lot to learn. Question: I didn't get a good look at the SMA to breadboard connectors. Did you make those or purchase? They look handy. Best Wishes!

[@joesmith-je3tq]

Those are just sections of semi ridged coax. Center conductor is solid. Magnet wire was wrapped around the shield and plugged into the board.

[@kennarnett8220]

@@joesmith-je3tq Thank you! Im learning a lot!

[@cwbh10]

Have you looked into the LibreVNA ?

[@joesmith-je3tq]

Not physically but I had looked at their blog.

[@marcin8132]

Hey, thanks for so many useful information, could you provide us more information about how to create own DC block using Teflon spacers?

[@joesmith-je3tq]

These blocks are not something you would typically want. For example, you can't purchase one from Mini-Circuits. For PDN, they needed to work at very low frequencies. This makes the VNA much easier to damage. These are a stacked ceramic. Of course, their value will change with voltage but I have set the low cutoff far below what I needed and typically I would never use them above 5VDC. In the description, there was a link to a thread on attenuators. www.eevblog.com/forum/projects/20db-rf-attenuator-seeking-feedback-to-improve/msg2965430/#msg2965430

[@irpheus]

Hi again, Joe ... With reference to this video and the beginning of your software manual I notice that in order to make sure that everything works I should find an original NanoVNA. However, searching e.g. ebay gives hundreds of suggestions none of which appear to be guaranteed original nanoVNAs. Any chance you know where to buy those original nanoVNAs? I will mainly be using my nanoVNA for PDN measurements on DACs and ADCs, and for characterizing passive components ... Cheers & thanks for any input on this, Jesper M

[@joesmith-je3tq]

Sorry I can''t be of much help. You can try asking in the email group groups.io . They are cheap enough that you could just buy one and try it.

[@irpheus]

@@joesmith-je3tq Hi Joe & thanks for your prompt reply ... I have now posted in the groups.io nanovna-users group and hope they may know about this. And, yes, I realize these nanoVNAs are not costly - it is more the time that may go into trying to figure out why something does not appear to work. I think your software looks really great - and I indeed appreciate the opportunity to be able to use it - but not being a programmer I know that I could spend hours upon hours trying to find out what could be amiss if something does not work. And, admittedly, I'd rather discover electronics, circuitry and components (using your software) in this time ... Cheers, Jesper

[@joesmith-je3tq]

@@irpheus Most of the common problems people run into are a result of them not reading the documentation rather than the product. That said, we are looking at one now that seems to have known hardware problem with the USB supply voltage. Most of the problems I have personally had with the products has been poor firmware. A few people have written me about using the H4 with my software. I never had any luck getting that combination to work but it's possible the newer firmware has improved.

[@irpheus]

@@joesmith-je3tq Hi Joe ... just saw your message. Hmmm ... having thought about things I reckon it might be feasible to mention what I would actually like to measure (noticing in your videos that you are interested in digital circuitries and your demonstration of how an "unstable" amplifier did not work well with the cheaper VNAs): My VNA use is to measure PDN impedances on DACs and ADCs - WHILE - those components are powered up, yet - not - active (no clock or data signals). Thus, the intention is to also get an idea of the intrinsic "hidden" inductance/capacitance PSU pin characteristics of the DACs and ADCs in question - without the DAC/ADC actually sampling. Additionally to characterize components - including larger capacitors (up to several thousand microfarads). I am, however unsure if the square wave output signal of these cheaper VNAs will mess up such measurements? Might you have an idea about this? And, if you reckon it wouldn't work, might you then suggest a not too costly VNA that outputs sine waves (and it would just be a gift if it would also work with your software) and would work for these purposes? I would appreciate your feedback! ... Cheers, Jesper

[@bobkozlarekwa2sqq59]

Is this software available to download?

[@joesmith-je3tq]

For about 6 months now.

[@heinrichhein2605]

would be great to see the inside of your transformer boxes, otherwise it is quite impossible to reproduce them without all of your equipment and testing you have done, image says more than words ;), but I can understand if you want to keep it secret, otherwise nice video

[@joesmith-je3tq]

I assumed most people would have seen common mode transformer and one time or another and of course, I had shown one of the early ones in this video. I also had shown a few of the later ones on EEVBLOG. Even the core types were listed on the boxes of the two I made up. Hardly seems like much of a secret. There was a person on EEVBLOG that claimed to have a fair bit of experience in this area. I was hoping they would come up with an improved design that would be lower cost. More to your point, you are very much correct that without the tools and more importantly, the knowledge to use them, you are pretty much on a difficult path. That said, the original NanoVNA combined with the V2Plus4 should be good enough to get that job done.

[@heinrichhein2605]

@@joesmith-je3tq We are not as smart as you im afraid. You know how much small things about the implementation changes the performance. Like how much you twisted the wire etc. Thanks again!

[@joesmith-je3tq]

@@heinrichhein2605 I wouldn't underestimate the viewers here. During the video, I mentioned this is a 50 ohm coaxial transformer. There's no twisting. However, you are very much correct that as we move up in frequency small details become more important. I expect viewers are are interested in learning which means digging in and getting your hands dirty. Of course there are certainly systems out there today that we use to make these measurements if that is what you are after. I mentioned some of them during the video.

[@heinrichhein2605]

@@joesmith-je3tq True as well, a lot of amazing people here, im maybe always curious if i see a box with electronics even if it is really mundane i just want to look inside you never know what you did not before. But yes, 50Ohm coax around a core is not exiting. But you always seem to have something to learn about even the simplest things. Thanks for putting such info out there for us!

[@joesmith-je3tq]

@@heinrichhein2605 I agree that seeing coax wrapped around some cores isn't very exciting but I added a picture of one to the manual anyway for curious people like yourself to have a look at. You're not going to learn much from it. What is useful is having the S parameter data. From that you can calculate most of what you need to know. I wasn't going to publish the materials I selected as viewers pretty much have all they need given to them on a silver platter.

[@zktec1]

Hi Joe, thank you for sharing your knowledge! I've learned a lot from you over the last couple weeks as I descend down this RF rabbit hole. I'm interested in laying out my own co-planar waveguide PCB test fixtures for 0805 size parts. I'd like to pick your brain about your waveguide PCB fixtures you talked about at 43:00 in your video. So I take it that If you were to redesign those fixtures, you'd place a cut in the ground plane (instead of adding braid) to make a better 4-point connection? Steve Sandler also mentioned the importance of a 4-point connection on these boards in this video: kzbin.info/www/bejne/fne3lWqmfZVpqZY I'd be super curious what your opinion is on his PCB layout he shows in the video...but looks like its for an odd size package. Also, he mentioned that a better way to break the ground loop (especially at super low frequencies) is to use a differential amplifier like his J2113A feeding the receiver port instead of a transformer. I'm not finding a ton of information about these fixtures on the internet. Appreciate any feedback. Thanks again!

[@joesmith-je3tq]

Yes, I would not use a solid ground like I show. Better to use 4-wire at these lower values. Still, results were good enough for this demo and limitation was the VNA, not the test fixture. It may be difficult to break the ground loop at 6GHz with an active circuit. So it is going to depend what you are after. The low frequency limitation of what I show is once again, the VNA.

[@serggorod1423]

34:34 каy момеnt. Ключевой момент!