Flawless PCB Design: Get The Most From Your LC filter. Part 7

[@franzliszt3195]

Really like the black background on your charts! So far your material and explanations are the clearest I've seen.

[@HansRosenberg74]

Thanks a lot, my first few videos did not have this black background. Black is much more pleasant to watch.

[@Fooballium]

@@HansRosenberg74 You chose the dark side? Jedis will find you guys. :))

[@franzliszt3195]

@@HansRosenberg74 also…..you could tone down your own room background a bit.

[@HansRosenberg74]

lighting is a really nasty thing when making videos. I cannot build a controlled studio in my room when it comes to light unfortunately. And the light changes throughout the day that I'm recording, clouds, sunlight..... its very tricky.... Not much I can do there . My camera also has limitation unfortunately.

[@Fooballium]

@@HansRosenberg74 I really like the video also with black color. It is also very easy to watch at night. My eyes are taking rest with this way. The content is in a good quality. I think this is the most crucial part. Thanks for your efforts Hans.

[@KostyanichW]

The VNA can directly measure capacitors and inductors. Turn on the Smith diagram, calibrate and connect the component under study to the VNA connector. Set the marker to display R and X. X will show you the value of the inductance or capacitance. R will show parasitic resistance. After that, you can easily calculate Q using the formula Q = X / R

[@KostyanichW]

Also, by moving the marker, you can explore the behavior of your component at different frequencies. Since the various parasitic parameters of the component depend non-linearly on the frequency. And finally, with the vna, you have access to very high measurement accuracy up to nH units for inductors and tens of fF (yes femto) for capacitances

[@HansRosenberg74]

Thanks for the tip. I used that indeed to measure the value of my inductors. You can see the test structure on my green transmission line pcb. I've made options there for open/short/load on the pcb to measure components. I've always learned that a smith chart measurement are not very accurate near the edges and a good inductor is always on the edge. I never tried to challenge that thoroughly though. This is why I relied on the resonator method here. I am curious to try it once though. Calibrate very accurately on the pcb with the mentioned structures and then indeed, you can calculate Q vs F in one go. Would be great if that was accurate.

[@febobartoli]

Thanks I always look forward to these videos. It's clear it's taking you a lot of time to put all this information together so clearly.

[@HansRosenberg74]

Thanks!

[@marais75012]

you're a good teacher...and peace and quiet is the way to go

[@HansRosenberg74]

Thanks!

[@johnwest7993]

Thank you for this video. Some comments on coils: I think you will find the best coil Q geometry is closer to one with a wider diameter than height, and varies with the diameter of the wire and the spacing between the windings, due to Skin Effect, inter-winding capacitance losses, and dielectric losses in the coil forming material. Tall, single layer coil forms are sub optimal, but are used to allow for inexpensive ferrite tuning slugs to be used in them and to minimize footprint size. Also, a coil dielectric like air, if possible, or polystyrene if coil support is required provides lower dielectric loss than other plastics. I'm sure all of the associated parameters are included in modern coil winding programs for single layer solenoids, but I learned them years ago the hard way. Also, a good way to quickly test unknown dielectric materials for comparative suitability as a coil form is to put them in a microwave and see which material stays coolest. There are significant differences in core materials, as you touched upon. I think of adding a dielectric form material to an air coil as adding capacitors between the coil windings, just like the interwinding capacitance that already exists between the turns. That helps to more easily picture the optimum mechanical design for a high Q form, and suggests an 'imitation' basket-weave form could be 3D printed which would allow for an exterior wind for convenience but still provide the wire shape and reduced interwinding capacitance common to basket weave posts.

[@Qsderto]

If he had used silver-plated litz, the quality factor would have become even higher. Yes, the further the turns are from each other, the smaller the interturn capacitance(that's why flat wires make sense). Experienced radio amateurs have demonstrated that the optimal quality coil has a length-to-diameter ratio L/D ≈ 1

[@HansRosenberg74]

Thank you both for your info, it is very helpful. I knew the microwave trick, I used it frequently when I was designing bluetooth antennas for headsets. I used to make really high q inductors for the 100Khz range using epcos ep cores and litz wire. The litz wire made a big difference. I also found these: www.coilcraft.com/getmedia/55a4b40a-2e02-4bf5-b0af-2ea5db75b6cf/1010vs.pdf . These things must have a pretty big interwinding capacitance, however, the Q is insanely high. Wouldn't that capacitance just reduce the resonance frequency? What is the relationship between that interwinding capacitance and Q factor?

[@HansRosenberg74]

I found this page: www.robkalmeijer.nl/techniek/electronica/radiotechniek/hambladen/qex/2001/09_10/page33/index.html Wow, didn't know there was so much behind calculating the q of an air core inductor. Fascinating. With this knowledge in hand, it would be really cool to build a VCO with an insane resonator and a really low phase noise.....

[@Qsderto]

@@HansRosenberg74 www.g3ynh.info/zdocs/magnetics/appendix/self_res/self-res.pdf "The self-resonance and self-capacitance of solenoid coils Dr. David W Knight 2013" I still haven't read it but it seems like a worthy effort. Capacitance is the same reactivity as inductance. At certain frequencies this matters. If frequencies are ten times or less below self-resonance, capacitance is unlikely to matter. I simply proceed from the well-known formula for calculating the quality factor.

[@Qsderto]

well..The solution, of course, is to recognize that the coil is a transmission line, with the only difference that the solution to the problem in the case of a curled line turns out to be quite difficult... Trivial measurements confirm that the coil's self-resonance is related to the total length of the conductor. It is therefore strange that the intrinsic capacitance of single-layer coils is still usually related to the static capacitance that supposedly exists between adjacent turns...

[@breedj1]

Great video. Looking forward to the next RF videos.

[@HansRosenberg74]

next one will go live in 2 weeks. I have to work on my course next week.

[@shahriarrudra7495]

Thanks for sharing your knowledge ☺️

[@HansRosenberg74]

you're welcome

[@nullptr472]

I for one am looking forward to the videos that use that pcb full of 'magic' microwave techniques

[@HansRosenberg74]

That will come in two weeks. in the mean time I have to work on my course

[@kabloui]

No way, cannot wait for the next episode with this new board

[@HansRosenberg74]

I just tested the 2 stage wilkinson combiner, it works quite well, it is within 8% of the simulated centerfrequency.

[@MrNerdHair]

Yeah, super stoked as well. I've known some sliver of what he's gone over so far but I don't recognize any of those things he mentioned on that board so it'll be awesome.

[@jonhoyles714]

Thanks hans for another great video :)

[@HansRosenberg74]

You are very welcome

[@M2N_Instrument]

Thank you, I like your content. The explanation is clear and easy to understand. Anyway when the course Will be launched?

[@HansRosenberg74]

It is going to take a while. Making it is a monumental task since I want it to be at the same level as all my videos. I think it will take me at least 3 months. I think february next year is a realistic target.... Maybe it will be quicker. I also have to keep making youtube videos since that is how I will get course participants.

[@y_x2]

Ignoring the resonance of a coil is a bad idea! You should always have an idea of the resonance and try to work far below it. Except if you try to do a choke.

[@HansRosenberg74]

Ah, soo dumb of me not to explicitly mention that. One of the hardest things making videos like this is not forgetting to mention things which are so obvious. Both inductors of course have a resonance way higher than their operation frequency.....

[@zachreyhelmberger894]

Thank you!!

[@HansRosenberg74]

You're welcome!

[@jaro6985]

Rat race incoming, good stuff.

[@HansRosenberg74]

I hope it works :-D

[@JasonRennieWTF]

Where did you get the NanoVNA V2 Plus5?

[@HansRosenberg74]

I got it here: www.tindie.com/stores/hcxqsgroup/? They say this is the official site. I also heard there are a lot of fakes around which are not as good as the real thing.

[@pierfrancescograzioli4080]

Ciao Hans, scusami se ti scrivo in italiano, vorrei chiederti come hai potuto fare le misure di attenuazione con valori prossimi a -90 dB con un NanoVna?? Quando i dati tecnici di questi strumentini, vengono dichiarati nel migliore dei casi a valori di S21 che non scendono sotto i -70dB? Ciao e grazie se vorrai rispondermi👏👏👏

[@HansRosenberg74]

Ciao, meno male che abbiamo ChatGPT, perché il mio italiano fa pena, anzi, non esiste affatto :-) Penso dipenda dal modello, e dovresti avere quello "vero" e non una delle copie economiche. Io ho un NanoVNA V2 Plus5, che se non ricordo male costava circa 400 dollari. Un saluto, Hans

[@IZ2XAF]

ciao, se hai una versione non cosi' "spinta" puoi usare l'opzione "average" che ti fa la media sui punti e quindi vai a ridurre l'impatto del rumore per ogni sweep dello strumento

[@pierfrancescograzioli4080]

La mia era una semplice domanda, in quanto conosco bene i NanoVna, che tra l'altro io non uso, perché ho un VNA professionale! Ciao

[@IZ2XAF]

@@pierfrancescograzioli4080 Ottimo che hai un vna professionale. Chiaro che era una domanda, anche da parte mia era un semplice consiglio ciao ciao

[@pierfrancescograzioli4080]

@@IZ2XAF Io utilizzo da piu di trent'anni VNA HP 8753, Rohde&Schwarz ZVRE, e so bene come si fa ad aumentare la gamma dinamica, si riduce l'IF Bandwith, si attiva la media sugli sweep (averaging factor) ovviamente si ha un notevole aumento del tempo di sweep, ma si può arrivare anche a -110/120 anche 130 dB di dinamica S21, quello che non capisco è come ci si possa arrivare con un NanoVna, perché nonostante io non lo utilizzi, ho un Chelegance SV4401A, ma non riesco ad arrivare a piu di -75dB, la domanda era questa, visto che Rosenberger ha pubblicato misure con valori di circa -90dB. Ciao

[@fpgaguru2102]

VNA is not the best tool to measure Q factor, Q meter handles high Q measurement much better.

[@HansRosenberg74]

I know! I just don't have one :-) The resonator method is very accurate IF your capacitor has way less loss though. When q values get really high that might be a problem....