 Hello and welcome to the Antina Implementation Guide. My name is Miroslav and I will welcome you to the second part of the video. So let's start to learn more. This guideline is going to give you a general insight for implementing PCB antennas into BLE or Gen2.4 GHz design. In this part it will be more about Dalatrix and wearables, so we will focus on impact of plastic enclosure, potting material and human body. In a conclusion I will make a recapitulation of what we have learned in these sessions. In the second part we will deal with what to avoid when designing and implementing an antenna and what Dalatrix objects affect and how they affect the antenna parameters. We will give many examples that show these changes. Based on that we will be more experienced and design it by knowing the consequences. There are special cases but more common ones where the implementation of BLE antenna is more difficult there. We can say that generate a special implementation is a variable product or PCB with antenna is mounted inside metal enclosure or onto metal part or is inside or close to absorbing material or is demanding from a range point of view. A selection of antenna and its placement on board or even out of the board usually depends on the purpose of using. You have different requirements for non-demanding application or for highly absorbent conditions. The antenna and its location must be properly selected and adjusted for such conditions to eventually compensate effects from the nearby objects. Antina serves as a frequency filter and direction filter. It can be designed for operation in presence of human body which reports high attenuation and detunes an antenna or in metallic enclosure for sure not from all sides. These special cases require a bit special design approach with knowledge of consequences. One of the common matters is changing thickness of PCB. Consider relocating the antenna design to a thinner material with the same permittivity. This issue can affect the behavior of the antenna in a certain way. All parameters were preserved except for the thickness of the FR4 material. However, for comparison purposes the size of the coplanar line was adjusted in such a way as to keep its input impedance. Let's see what are the consequences. In this case a resonance frequency has been shifted up, so antenna should be elongated a bit or the matching network adapted. Given the small shift it will certainly be possible. We can see that in principle the antenna has not lost its properties. It is slightly frequency shifted but still usable. Let's look at the other parameters. We can see slightly worse numbers on the VSWR but it's nothing major. It is caused by only a small deviation of the resonance. What we are interested in is the effect to efficiency. I dare to say that the efficiency is shifted towards the maximum due to the adaptation to a different frequency so by adjusting the matching network we will get almost the same numbers as for the original antenna. We can say that the change or thickness of FR4 material caused shift of resonance frequency upwards. If the change is not large it can be adjusted by matching network otherwise the antenna shape must be modified. When looking to radiation pattern nothing happened. It is almost unrecognizable against the default one. We can say that changing the thickness of FR4 material basically preserves the properties of the antenna even on far field side but it is still necessary to check it and adjust the matching network. One of the special case is a small antenna. For devices what implements small antenna we can assume small range and performance. The reason is small efficiency. It is trade off between performance and cost effective or size optimized ocean. The misunderstanding of usage of small ceramic antenna is that people are thinking why not to select a small antenna. After all these are on the market they must work. Furthermore this fits more into my small device. A reason is likely saving space on PCB but there is a hidden impact to performance. Although it is stated that it has a gain it has also certain losses in efficiency and has narrow bandwidth. The smaller the antenna is the more you must care about an accurate impedance adjustment and taking care about overall tolerances as it can easily happen a worsening of performance. The antenna for the BLE is mostly narrowband so just sits in 100 MHz bandwidth. An efficiency is not a problem if the device communicates on few meters but for demanding application it could be an issue. So for a certain application it can be used such small antenna but you have to be aware what are the consequences. When any surrounding material is present then you must count with its dial trick features or absorbing properties. A dial trick material detune the antenna either if it is very close to the antenna or if it touches the antenna. If far from antenna less impact from dial trick. But you can count with such surrounding conditions and to make the antenna design matched it results in a change of trace antenna dimensions. Ceramic antennas can hardly be modified else than by matching network. Subsequent impedance adjustment of matching network will be most likely needed no matter if PCB or ceramic antenna is implemented. This can be the case of some waterproof enclosure where a protective epoxy resin is spotted into the device to cover PCB. In case of absorbing materials for example the plastics consisting of carbon fibers this is a huge attenuator so avoid these layers or solid materials nearby. Therefore a knowledge of housing material in which the device with antenna is placed is needed for proper design. Now we will investigate what happens if we pour epoxy resin on the antenna. It is a common case when we protect electronics from water and possibly shocks. This material therefore directly touches the metal part of the antenna as well as the carrier effort for material. This of course changes the conditions for which the antenna was originally designed. Here on the written loss plot you can see how the edit material increases the resonance length and thus shifts the resonance lower. The Smith chart also shows that we have another resonance there which is also now lower in frequency. While the resonance is lower the written loss becomes worse. I will point out one consequence here. The next resonance is certainly a multiple of the fundamental resonance. We can see that the higher one is better adjusted on a contrary to the basic one. This could emphasize the emission of harmonic frequencies coming from transmitter. What bothers us the most is the frequency shift caused by the dark trick and thus also the adaptation of higher harmonic frequencies. In this case it is therefore necessary to consider tuning and or shortening the antenna for the given type of material and its resulting thickness. The efficiency is a little lower in band so the losses are not too large but the question is whether the efficiency on the double doesn't reach similar or even higher values. So antenna is detuned a little with worse written loss. Efficiency is a bit decreased but nothing what is critical. The antenna needs to be matched by matching network or if possible to change the length of the antenna. When looking to the far field we can only see unnoticeable deviations from the ideal radiation pattern. We must count with a smaller antenna gain due to efficiency. Except for small deviations the radiation pattern has not changed. The antenna has retained its omnidirectional character. A further example is about non-touching dielectric material. Suppose that certain material with dielectric properties will be a little distant from the antenna. For simplicity let's take an epoxy resin material again. It is enough that the dielectric material is in proximity of a few millimeters from the antenna shape. The object is made from mentioned material and the size is just the size of the antenna clearance. Thickness is 1 millimeter. We can see that the written loss is a bit detuned but it is still usable and it should not be noticeable in the radiated power. The written loss remained in the band below minus 10 dB which is rated as a very well matched. This is also evidenced by the VSWR which is not depicted here where the entire band is below the value of 2. And what is important the efficiency did not change much either so almost all the power goes to the antenna. The antenna has kept the same parameters. We do not expect the radiation pattern to be altered in any way. So dielectric materials like this if they do not directly touch the antenna they do not change its properties either in the written loss side or antenna pattern side. Now let's make a few notes on another important area variables. Another challenge is variable devices. An antenna fights against the attenuation of human body there. It isn't possible to reach omnidirectional pattern you can forget to that. The range of BLE signal through the body is very shortened as human body consists of high volume of water. A placement of the antenna is very critical and it must be carefully implemented into such device. It is a case of earbuds or various medical stuff. Impact to performance depends on the distance from human body. There is also a further aspect for the variables. The antenna adjustment should be performed in the presence of the human body. Only this way you don't burn all the radiated energy. In this case it is better to focus on radiation efficiency than to make a radiation pattern omnidirectional. While designing and constructing the device try to place the antenna as far as possible from the human body. Some parts of body impact antenna more some less. It is a matter of composition of human skin and tissue or even bones. Try to get as much power as possible from antenna regardless radiation pattern. It is especially about efficiency, antenna size and correct impedance matching. Now let's look at how the human body affects the performance of the antenna. It is a slightly more complicated problem because human tissue is composed of layers of tissue that have different dial trick and conductive properties. At the same time the thicknesses are different for each person and are also different in different parts of the body. For this task I selected this composition which worked well for me in the designs. I just note that the antenna is not directly on the human body but there is a certain distance between them. First the antenna is heavily detuned. Basically it works quite well but in a different band. However it has noticeably worse adjustment to the desired frequency. What is unusual is that a weak resonance appeared near the desired band. Of course it depends on the distance from human tissue. This needs to be evaluated because such secondary resonance can be only present near human tissue and can completely disrupt the matched state of the antenna. When looking into VSWR that is not depicted here we can see the tuning of the fundamental resonance but the secondary resonance located close to band and resonances at higher harmonic frequencies as well. Another aspect is efficiency which drops significantly. Some part of the efficiency comes from mismatch but even if you manage to get the antenna to the right resonance frequency you still won't get good efficiency. You simply burn a lot of energy thanks to the human body. We could already see the consequences on return loss. This is shifted in frequency quite a lot and the efficiency is affected a lot. We must modify the antenna in shape to be able to resonate in desired band even thought the efficiency doesn't reach a good level. For sure the optimization of efficiency is a priority as well. The change is also noticeable in the radiation pattern. There was a significant rotation of the characteristic and the directionality became more visible. The problem is therefore the significant influence of human tissue on all antenna parameters. Therefore it is not easy to adjust the antenna in a straightforward way so that it meets our needs. Additionally, the complication is that while the antenna works well in the presence of the human body it will hardly have a good match when the human body is not present. It must be on mind and check the performance of antenna while the human body is distant from the device and an operation is requested in the time. In the pictures are visible changes in reactivity that wouldn't be so much an issue but you must count to a significantly reduced efficiency. So antenna pattern has changed a lot and became directional. Affected efficiency is the cause of less range. It is advisable to correctly position the device and dust the antenna in relation to the partner device with which it communicates. As you can see a decrease in gain in the direction of human tissue. We can see that the implementation of the PCB antenna is not easy if we want to maintain its performance. Maybe it would be easier to use a ceramic antenna. Let's look at this in more detail. Manufacturers of chip antennas are giving a recommendation of layout for their antenna. The shape of clearance and size of evaluation board has certain meaning and this information should be read out as well not only the table of parameters. This evaluation PCB mostly says what are the minimal or optimal requirements for your PCB. If you want to reach the parameters stated in a datasheet you should follow the evaluation board size and location of the antenna. When decreasing the PCB size you will lose efficiency and bandwidth. In case of changing location you will observe a change of radiation pattern shape, directivity and the efficiency and bandwidth as well. When increasing the PCB size you will probably get more efficiency but far field will be more fragmented and to make matters worse it will be different in X and Y directions. It is always better to imitate the recommendation to your board size and preserve the PCB shape. But don't be sad, similarly it is for PCB antennas. Let me summarize the lesson learned from both parts of the video. Point by point I will remind you of the most important aspects of PCB antenna implementation. Although there is an increasing need to install the antenna in a small space on the PCB believe that this worsens the antenna parameters in terms of achievable efficiency, bandwidth and return loss. So better to use as large as possible antenna. With the size of antenna the size of ground plane relates to. In many cases such type of antenna is used that sufficient surface of ground is needed as well. In other words you cannot use monopole antenna on PCB surface which is smaller than antenna itself. Antenna needs a space where to radiate. It is mostly at the edge of PCB only in a couple of cases for some antenna types you should consider about middle of PCB. There is mostly no possibility to put the antenna in direction that the antenna would be above ground plane. It is almost always in the same plane as ground plane is. It should be our concern to make ground as large and as visible for the antenna. Thus layers in ground around clearance has to be placed as well. Antenna should be placed at the edge of PCB according to recommendation of manufacturer. However this may not always be the case. On the contrary a microcontroller is better to place somewhere farther from the edge as it needs the components and appropriate roads placed around the MCU. In front of the antenna wouldn't be any obstacle that would defend the radiation. It can be metallic object. Select a place according to recommendation of manufacturer. You should also consider mounting or wearing side to choose the best direction. Clearance is the space on PCB underneath the antenna footprint where no metallic plane, trace or object is present. It must be defined for the antenna according manufacturer recommendation. Decreasing this clearance causing smaller performance of antenna. We sure can imagine as the antenna is radiating element that can simply interfere with other roads and components. The more sensitive circuit the more consequence it could have. This is about very low power signals and high frequency signals and about power supply as they are directly connected to RF part of transceiver. But it may be an issue even by an opposite direction. I mean that distribution could come from different signal to RF, let's say to receiver. Any metal object placed in proximity of antenna acts as an obstacle or can make the antenna impedance mismatched. A particular form of effect will depend on a size of the object. What is important is not to use a metallic enclosure because it is a really huge obstacle. The radiated power will be completely reflected if no free space is present at all. You must provide an area where antenna can radiate through. There exist some options how to get the radiated power from enclosed metallic housing. But it needs to have EM simulator and do that really accurately as this is a risky task. We are getting to matching network. It should be located just at the antenna port due to feeder which is designed to exact impedance and this impedance is kept till this point where antenna is feeded. Even though an antenna is designed to have 50 ohm it may not have this impedance and from the reason it is impacted by surrounded conditions. So you need to adjust the antenna impedance to the feeder by the matching network. The stp change is just at the antenna port so this is the right point to locate matching network. If you design a matching network somewhere else you have to include a feeder in calculation of matching components values. It basically acts as a phase shifter so it can completely change the complex number of impedance seen by the antenna. You can take a look in some smith chart to try it. Another matching network is usually located in front of the MCU. It is between the RF output and a filter. This is present there due to different optimal impedance provided by end stage of transmitter. To get as much power as possible you need the matching. But I have met severely with wrongly placed matching network or forgotten matching network. This is about misunderstanding of an RF stuff. Try to take care about positions of the components in matching network. If two parallel components are used they should be tied to the same side of ground strip to create common ground point. This will limit parasitics. They should be located close to each other from the same reason. Often there is only limited space on PCB to place RF part. It can lead to unwanted coupling between signal traces. Try to completely clear away a surrounding of RF path or at least route them out sufficiently far from it. There must be performed trade-off between PCB size, length of RF line and limiting a coupling among signals. An RF line should be kept short due to transmission losses but if you have space rather prioritize a bit high loss then unwanted signal couplings and interference. Although the antennas especially the PCB ones have often different designs impedance than 50 ohms as their natural impedance cannot be exactly this one. The rest of RF path and components inserted are designed in 50 ohms and RF line must comply this characteristic impedance. The antenna will be adjusted by matching network to impedance of the feeder. Instruments, cables and connectors are commonly designed in 50 ohms, sometimes they can handle different characteristic impedance. There are certain aspects that influence the resulting performance of an antenna. One of the important thing is to design antenna to not have thin trace as manufacturing tolerances could cause large change in impedance then. Also it is important to count with different die-trick stick up and thicknesses of individual layers. Count with different thicknesses in this region where clearance is placed due to the absence of metallic layer there. Ask for specification in this area separately from the common places. Copper thickness can play a role as same as thickness of solder mask coating. This has die-trick properties and it can detune the antenna. If not necessary do not use it on the copper part of antenna as manufacturer cannot guarantee a thickness during series but small changes caused by the above aspects can be adjusted by matching network in case where it is not affected largely. If not properly designed or antenna resonates far from demanded frequency then matching network may not be enough and mechanical adjustment of antenna or even its redesign will be needed. The inverted antenna is widely used in RF designs for BLE. We can meet with various modifications especially with folded versions that save space on PCB. Even this folded one reaches enough bandwidth to cover Bluetooth band. But if you have space on board you should use the space as possible for unfolded version. First you get efficiency and second a bandwidth. Do not forget to use sufficient size of ground. Thank you for watching the second part of the video. We will see you next time. Take care, bye.