Their datasheets are pretty terrible on electrical specifications and characteristics compared to other vendors. The documentation is highly targeted towards software engineers.
Such a wonderful breakdown. Your questions for clarification turned this into an exceptional class. You should do a full class with experts with your questions and commentary.
Robert, first of all, thanks for your videos! They are very helpful. As for the feed line design I think there is a mistake, the GND layer must be cleared under the series inductor used in matching circuits. Leaving GND under the matching component influences the feed line impedance. As for distances between vias along the feed line it should be short enough comparable to the wavelength to reduce losses. In my practice I use a 2mm pitch, which is good enough in most of the cases.
Regarding the suggested ground clearance below the inductor: where do you have this Information from? From several reference designs I could not see this …
Antenna design is a very tricky thing. The old masters did it all via experimentation and using intuition. You can get an amazing amount of gain with the proper antenna. When i was doing Ham radio a kid there was a rich man who built stacked rhombic antennas on his estate. that is the kind of antenna they used in WW2 for maximum range. It uses a lot of room, and is fixed in direction, but boy does it work!
Great tutorial, actually you don't need an expensive VNA to get experience with matching circuits nowadays. cheap nano vna or libre vna will do the same job. We use Optenni Lab for RF matching, it has a large database of known manufacturers like Johansson, Murata, TDK etc. Unfortunately it is not free, but there is a trial version as far as I know.
I am curious as to the default inductor brand Nordic uses. TDK and Murata both seem to be popular, but from a naive perspective, I don't consider them interchangeable. I suspect that even at 2.4GHz, the PCB stackup matters to inductance values. As you asked in another video, "when is a capacitor not a capacitor?" when it's above it's SRF, it acts like an inductor. That was a big insight to DC blocking capacitor values for me.
Thank you for your explanations! However, I don‘t see that you have adjusted the impedance of the transmission line to 50 Ohms. Having high power output while seeing low harmonics is definitely a good sign for having reached 50 Ohms, but could you also measure this?
There is one question I have. In this case that the Bluetooth IC has a 50ohm output you only need to concern about the matching circuit of pcb trace. If the IC output has a 35ohm impedance, how to match the antenna? Thanks
I can see that there is a 0.7pF capacitor on the reference design. Why not just make the trace a bit wider at the point so the parasitic capacitance with the ground layer underneath produces the correct value without the actual component?
Thank you so much for your effort but I think here exist lots of trick in the video that are not well explained at least I felt everything is mute! For example what happens when decrease or increase antenna to it's antenna impedance?
SWF port for antenna testing that is on the Development kit. She shows the part number so you can look at the component. The part works in a way that it disconnects the PCB antenna if you plug in a cable to measure
*I have one very important question:* How she estimated the quarterwavelength is 22mm? ?if we consider the effective dielectric constant of the microstrip trace as 3.3 then using the formula, Quater wavelength = c/(4*(sqrt(effective dielectric constant)*Frequency) This gives us 16.85mm not 22mm... 22mm is only possible when effective dielectric constant is 1.77 which cannot be the case in a microstrip trace!!! Please Explain!!1
Those calculations apply to dimensions of a microstrip line, a transmission line that has a line on top and a reference plane beneath a given dielectric material. The electrical length of the signal propagating in said dielectric is therefore affected by the permittivity as you well said. The antenna has no reference ground, it's just a metallic strip in the air (and a bit of the PCB dielectric, but no other reference beneath), so the length is more approximate to the free-space length - but still a little affected by the presence of a higher dielectric material in the vicinity of the conductor plus the fact that a monopole (as well as a dipole) should be slightly smaller than a free-space quarter/half of the wavelength, plus the fact that it is bent which adds inductance and thus further reduces the effective length to have the same resonance. There's no calculation that gives a precise value for it, the way to go is to make it the free-space quarter wavelength and then "tune it", by slicing at the antenna, arriving at the correct length for the desired frequency. That's why, unless you have simulation software, you can either build some prototypes with copper tape (and a lot of cutting skills), or otherwise follow the recommended designs. Hope that helped.
can anyone explain to me, how the orientation of the antenna is?, if it is placed on a nordic nrd52833dk as shows in the video, how is the orientation of the axes x,y,z?
Overall a great video! I've been looking for AGES and haven't come across impedance matching of microstrip antennas. I've also realised, standard capacitor and inductor books WILL NOT YIELD ANY GOOD results - the High Q caps Kaja used might be the winner - definitely worth the $80 AUD for the set. One thing I didn't quite understand was why she decided to shorten the antenna to reach 50 Ohm... why does that work rather than adding more capacitance to put the impedance into the capacitive zone and then a series inductor to put it into 50Ohm + 0J? Would that have worked instead? Are there any links or information about this? I'm a bit stumped on what to search.
She seem to didn't mention it at the same time she did that, but I believe she was shortening the antenna to help move the markers into the centre of the resonant frequency shown on the SWR chart on the right. She was trying to avoid adding another unnecessary component.
Adding excessive C can be malicious from efficiency point of view, even high Q capacitors have their losses. It is better to hand-tune antenna on prototypes and move forward with new design.
I’m making a board design on using the 5340 reference design at the moment. I would love to know if it is possible to use the reference design with a dual layer 1.55mm board? As of now I have no routing on layer 2-4 and just a few traces on the bottom layer, would it be possible to just have a dual layer board then or is there some problems I don’t see doing that?
The short video formats (tiktok, shorts, etc) changed behavior and expectations of people from videos. If it's not interesting from second 1, they just move to a different video. That is what is unfortunately happening. PS: Kaja explains what she does later in the video.
Hello, Robert will you ever make a tutorial where you make your own board with Wifi and Bluetooth chip with 2.4Ghz Antenna (maybe inverted F) in Altium. Because you yourself explain things so much simpler and easier. Make an hour or 2 long video like this idea. It would really help a lot of people.
i would have never imagined knowledge of this calibre would be so accessible, your videos have been helping a lot in my advancing hobby
Perfect subject. She knows her job and your questions completes the tutorial in excellency. Congrats.Thanks.
For me, Nordic is the role model of product documentation and product support. Thanks for inviting Kaja here, great job!
Their datasheets are pretty terrible on electrical specifications and characteristics compared to other vendors. The documentation is highly targeted towards software engineers.
Such a wonderful breakdown. Your questions for clarification turned this into an exceptional class. You should do a full class with experts with your questions and commentary.
Robert, first of all, thanks for your videos! They are very helpful. As for the feed line design I think there is a mistake, the GND layer must be cleared under the series inductor used in matching circuits. Leaving GND under the matching component influences the feed line impedance. As for distances between vias along the feed line it should be short enough comparable to the wavelength to reduce losses. In my practice I use a 2mm pitch, which is good enough in most of the cases.
Regarding the suggested ground clearance below the inductor: where do you have this Information from? From several reference designs I could not see this …
Thank you so much Robert Sir for organizing this valuable and informational session!
Antenna design is a very tricky thing. The old masters did it all via experimentation and using intuition. You can get an amazing amount of gain with the proper antenna. When i was doing Ham radio a kid there was a rich man who built stacked rhombic antennas on his estate. that is the kind of antenna they used in WW2 for maximum range. It uses a lot of room, and is fixed in direction, but boy does it work!
Many big THANK Kaja and Robert for this really interesting video
As usual great video. Actually all in your guests session what you ask seems to be our questions. Nice informative video.
Really nice stuff about antenna design , thank you for sharing your knowledge!
Great content, thank you Robert and Kaja
One of my favorite videos on the channel. Thanks!
Incredible explanation!!! Thank you very much!!!
wow - Great video so glad this DK was covered.
Excellent info! I use nrf chips en my work
Great as usual sir Robert
what a useful video! thanks a lot Robert and Kaja.
Great tutorial, actually you don't need an expensive VNA to get experience with matching circuits nowadays. cheap nano vna or libre vna will do the same job. We use Optenni Lab for RF matching, it has a large database of known manufacturers like Johansson, Murata, TDK etc. Unfortunately it is not free, but there is a trial version as far as I know.
I am curious as to the default inductor brand Nordic uses. TDK and Murata both seem to be popular, but from a naive perspective, I don't consider them interchangeable. I suspect that even at 2.4GHz, the PCB stackup matters to inductance values.
As you asked in another video, "when is a capacitor not a capacitor?" when it's above it's SRF, it acts like an inductor. That was a big insight to DC blocking capacitor values for me.
The sample kit she was using looked like one I have from Johanson P/N C402
Thank you for your explanations! However, I don‘t see that you have adjusted the impedance of the transmission line to 50 Ohms. Having high power output while seeing low harmonics is definitely a good sign for having reached 50 Ohms, but could you also measure this?
The same thing i want to know
Another awesome and delightful video
Brilliant 👏 Thank you both.
Such fantastic content ❤
beauty with brains , a combination to kill for. A rarity in itself to find in electronics !
pretty sweet presentation! thanks! :D
Thank you so much. Really really helpful and useful.....!!!!!!!!!!!
There is one question I have. In this case that the Bluetooth IC has a 50ohm output you only need to concern about the matching circuit of pcb trace. If the IC output has a 35ohm impedance, how to match the antenna? Thanks
I can see that there is a 0.7pF capacitor on the reference design. Why not just make the trace a bit wider at the point so the parasitic capacitance with the ground layer underneath produces the correct value without the actual component?
I don't know, but I guess it's probably easier with a capacitor for tolerances and one can use a cheaper pcb material with less impedance control
Thank you so much for your effort but I think here exist lots of trick in the video that are not well explained at least I felt everything is mute! For example what happens when decrease or increase antenna to it's antenna impedance?
A list of common reason for certification failure and the way they are usually solved would be great.
22:13-22:18 What were you guys talking about? LOL, Srr. I dont get it!! Anyway, thank you both for great content!
SWF port for antenna testing that is on the Development kit. She shows the part number so you can look at the component. The part works in a way that it disconnects the PCB antenna if you plug in a cable to measure
Great information!
Why 50ohm meeting point? Can you tune the antenna to the radio directly?
Wow, great video.
*I have one very important question:*
How she estimated the quarterwavelength is 22mm?
?if we consider the effective dielectric constant of the microstrip trace as 3.3 then using the formula,
Quater wavelength = c/(4*(sqrt(effective dielectric constant)*Frequency)
This gives us 16.85mm not 22mm...
22mm is only possible when effective dielectric constant is 1.77 which cannot be the case in a microstrip trace!!! Please Explain!!1
Those calculations apply to dimensions of a microstrip line, a transmission line that has a line on top and a reference plane beneath a given dielectric material. The electrical length of the signal propagating in said dielectric is therefore affected by the permittivity as you well said. The antenna has no reference ground, it's just a metallic strip in the air (and a bit of the PCB dielectric, but no other reference beneath), so the length is more approximate to the free-space length - but still a little affected by the presence of a higher dielectric material in the vicinity of the conductor plus the fact that a monopole (as well as a dipole) should be slightly smaller than a free-space quarter/half of the wavelength, plus the fact that it is bent which adds inductance and thus further reduces the effective length to have the same resonance. There's no calculation that gives a precise value for it, the way to go is to make it the free-space quarter wavelength and then "tune it", by slicing at the antenna, arriving at the correct length for the desired frequency. That's why, unless you have simulation software, you can either build some prototypes with copper tape (and a lot of cutting skills), or otherwise follow the recommended designs. Hope that helped.
@@magictwister8439 Ohk, Thanks for the explanation! I got it now!!
can anyone explain to me, how the orientation of the antenna is?, if it is placed on a nordic nrd52833dk as shows in the video, how is the orientation of the axes x,y,z?
Overall a great video! I've been looking for AGES and haven't come across impedance matching of microstrip antennas. I've also realised, standard capacitor and inductor books WILL NOT YIELD ANY GOOD results - the High Q caps Kaja used might be the winner - definitely worth the $80 AUD for the set.
One thing I didn't quite understand was why she decided to shorten the antenna to reach 50 Ohm... why does that work rather than adding more capacitance to put the impedance into the capacitive zone and then a series inductor to put it into 50Ohm + 0J? Would that have worked instead?
Are there any links or information about this? I'm a bit stumped on what to search.
She seem to didn't mention it at the same time she did that, but I believe she was shortening the antenna to help move the markers into the centre of the resonant frequency shown on the SWR chart on the right. She was trying to avoid adding another unnecessary component.
Adding excessive C can be malicious from efficiency point of view, even high Q capacitors have their losses. It is better to hand-tune antenna on prototypes and move forward with new design.
I’m making a board design on using the 5340 reference design at the moment. I would love to know if it is possible to use the reference design with a dual layer 1.55mm board? As of now I have no routing on layer 2-4 and just a few traces on the bottom layer, would it be possible to just have a dual layer board then or is there some problems I don’t see doing that?
but anyone's PCB stack up is not the same as the reference design. Correct? so why bother to follow the reference design?
Thank you!!
You should have introduced your guest! What her position, experience and so on...
I used to do it, but ... I am curious, what do you think why I don't do that anymore?
@@RobertFeranec To be honest, I don't know the reason why! But I would like to know for sure?
The short video formats (tiktok, shorts, etc) changed behavior and expectations of people from videos. If it's not interesting from second 1, they just move to a different video. That is what is unfortunately happening. PS: Kaja explains what she does later in the video.
@@RobertFeranec thanks for the explanation, Robert!
Tbh, it's never relevant. If the presenter is competent, the audience will know.
Hi, if we want to design a 35 Ohm transmition line is it possible? Thanks
Yes, depends on your trace dimensions, the distance to reference and the used materials.
Hello, Robert will you ever make a tutorial where you make your own board with Wifi and Bluetooth chip with 2.4Ghz Antenna (maybe inverted F) in Altium. Because you yourself explain things so much simpler and easier. Make an hour or 2 long video like this idea. It would really help a lot of people.
it is on my to do list
@@RobertFeranec Cant wait ❤️
mia malkova @8:27
Why send our design to tune? You can introduce guidelines to let us tune on our own, doesn't it?
it's interesting and hardware intensive, better to uses pre-certified modules.
custom design wont fit pre-certified module(s)
idk why these chips do not support wifi while they do support many protocols in 2.4Ghz anyway. Makes no sense.
Their new ones do
Interesting topic! Thanks for sharing!