Cheap RF 433.92MHz design, PCB track antenna, values of capacitors

Hi.

I'm doing a RF remote control project using a PCB with microcontroller and with simple onboard antenna layout as transmitter and a 433.92MHz receiver RWS-371F-6 model. Here is the datasheet of the receiver: https://www.es.co.th/Schemetic/PDF/RWS-371F-6.PDF

I don't have knowledge about RF design, just electronics and MCU programming. I'm using manchester encoding/decoding @ 1000bps to improve the communication distance from receptor board to the transmitter. 'RF-TX' is a 0V~5V signal.

The circuit of RF was copied from a board of a garage door opener, but I included my own microcontroller to the PCB schematic and I have some prototypes of the board. The values of resistors (150 Ohms, 47K, 47 Ohms) was also copied from the garage opener transmitter board. Here I am showing the layout of my board and the values/part number of the components. I've mounted a few pF capacitor combination values (0805) for C2/C3. I have a capacitor sample book with many values on pF range, many values from 1 to 22pF. The communication TX/RX is working correctly, but I want to check if I can increase the distance range by using adequate values for C2/C3. Currently I'm using both as 1pF.

At the receptor, I'm using coil loaded antenna with 1/4 of wavelength, the same of the picture, tomorrow I will try a straight rigid wire of 1/2 of wavelength.

Well, as said, my doubt is around the correct values of C2/C3, can somebody who understand of RF desing calculate them easily? Or does this generate a kind of service that I have to contract a RF designer to calculate them to me?

Regards.

In my opinion, it will be easier to try capacitor values (while keeping C2 = C3) experimentally,than to try to calculate them.

On the referenced PDF datasheet, the receiver has a linear output on pin 3, which you could connect to a voltmeter to measure relative signal strength. You can test this by moving the transmitter towards and away from the receiver while watching changes in voltmeter reading. If this works, you could keep the distance fixed, and vary the capacitors for maximum signal.

I recommend you purchase a copy of the ARRL Handbook from www.arrl.org to begin your learning of radio circuit and antenna fundamentals from a practical perspective. Much of the content is specific to amateur radio, but the fundamentals chapters are applicable beyond amateur radio.

The two capacitors you mention are critical in the operation of the colpitts oscillator, you change them at your peril. There is quite a bit of work involved in getting a SAW based oscillator of this type sorted, they are deceptively simple, there is a paper here which goes into some detail.  Looking at your PCB, there is a lot of switched RF energy flying around the place, you may well need to be more careful about how your grounds are arranged. You may be better to buy and mount the transmitter module to your board. On your schematic, you show the antenna as a collector load for the transistor, I believe that it should be an RF choke, and that the  antenna is probably fed from the collector via  matching  components.Keep in mind that this frequency is largely reserved for HAM use in many countries, so you need to avoid creating interference, and it should not be used in a mission critical application where the receiver could be blocked by local more powerful transmitters.
Cheers,Richard

Hi. Thanks for the kindly help of yours.I also posted the same topic in another forum (edaboard), and a user said: to get more power you can decrease R3 (I decreased from 150 to 120R) and increase C3 (I increased from 1pF to 1.5pF), and R1 was originally 47K and now it is 22K. The schematic below shows the values that gave me the best results in terms of range until now. So then, changing C1 to 2.2pF with R3 = 120R and C2 = 1pF the results seemed to be worst but this can be just impression.
I'm doing a new layout for that PCB of the transmitter, and will produce new prototypes soon. Below is my current drawing. I removed red/top layer tracks apart from the RF circuit, and added a GND plane on top layer (red) to lower the impedance from the bottom pad of C4 (ground) to the ground of the amplifier circuit. A have a doubt: should I place ground plane where the components of RF circuit are placed? Currently (on new layout) I have a plane at the other side of the RF components (the red fill). Until now, I know that I only can't place ground fills where the antenna track is.From the first prototype of the board (antenna lenght = 1.91 inch or 48.5mm ) to the second layout that I'm currently doing (antenna lenght = 1.45 inch or 36.9mm), I've changed the antenna drawing as can be seen.
The goal: I want the antenna to broadcast at maximum strength, to achieve resonating action, producing standing waves, but unfortunately I have poor knowledge in RF.The main questions now are: What should be the 'best' or 'suitable' shape that I can do for the antenna in my new layout? For example, with a specific width and drawing? Is it possible to calculate/predict the antenna track and calculate its inductance and then calculate the values for C2 and C3 to reach the goal? Of course I can test the values of capacitors in practice, but I'm wanting a starting point about the antenna shape for the case of my PCB and my schematic to finish the new layout.
Tomorrow I'll check the linear output of the receiver...Regards.
   

I am no expert in this area. However, here are my comments for what they are worth. Your antenna  should  be a multiturn  loop to to be a more effective radiator. The inductance of the loop has a bearing on the performance of the circuit, so how feasible a multiturn loop is in terms of the required collector inductance would have to be investigated. At present you have a single turn loop on a PCB,  a multiturn loop  would need to be formed from some heavy copper wire and soldered onto the board.  At the frequency you are operating, however,  the  loop  is  small relative to the wavelength, and together with skin effect  it will not be an efficient radiator. To use a whip antenna, which should be more efficient,  you need to replace your pcb track  antenna with an rf  inductor , and via a capacitor couple you antenna to either emitter or collector of the transistor, as in the app note (see page 9). There are further comments about  the value of the collector inductance and the trade offs in using collector or emitter coupling in the app note. Have you tried reading the application note, it is quite detailed, however, if you don't have the background it will  appear quite daunting. It does have some quite readable sections,  pages 8 and 9 are the important ones. I suggest you go through it and extract the information where they deal with some practical circuits, and see if you  can get a better understanding. If you are not familiar with j notation, there are plenty of good basic text books on the subject.
As a final comment, the transmitter board  (TWS-DS) supplied with the receiver you are using works with a whip antenna, you may be able to copy their circuit.Cheers,Richard

To keep the same resonant frequency of the PCB antenna loop, I recommend you keep the same shape in your loop. The wiggles in the original loop are there to fit the needed length of trace into the space available. Making your loop larger to give the same length as the original (following the wiggles in measuring original length) should work also, and the larger area would give more signal. Calculate the wavelength of 433.92 MHz in air to see how short the wavelength is. The wavelength on your PCB will be shorter by a factor sqrt(PCB material dielectric constant).

Hi.The PCB material is (and always will be for this PCB) FR-4 with 1.6mm thickness. I will check with my PCB supplier what is the material dielectric constant and post it here later.
I also posted the same question of the topic in another forum (edaboard) and a user recommended to add the C5 capacitor (check the new schematic attached, its a capacitor connected from collector to ground). He said: "Add the extra capacitor at the bottom left end of the antenna. Ideally, it should be directly across the two ends but with the ground plane (polygon pour) in place it should be safe to add it between the ground plane and transistor collector pin. The idea of C4 being 470pF isn't because that value is critical but that it has a high enough value to be seen as a short circuit to the signal at 433MHz. According to the XC formula, it has a reactance of only 0.78 Ohms at 433MHz! That means that the antenna is effectively grounded at one end and driven by the transistor at the other. Connecting the new capacitor from collector to ground plane is therefore electrically very similar to connecting it across the antenna ends "
Question: What value I can use for C5 as a starting point?
Actually, I'm going to produce new prototypes on next weeks, and I'll produce 3 versions of the transmitter board:- Rev. B = original prototype with little modifications, but same antenna shape.- Rec. C = new layout with curved/wiggle shape of antenna - Rec. D = new layout with straight shape of antenna

Peter, you said "The wavelength on your PCB will be shorter by a factor sqrt(PCB material dielectric constant)" How do I calculate the total lenght that the antenna track should ideally have? If I use the unit in milimeters, 433.92MHz have 0.69m or 691mm wavelenght, and sqrt of 691mm = 26.3mm. Is that the ideal lenght that the track should have? If not, how do I calculate it?
Question 2: About revision C and D layouts, for rev. C I'm using the antenna track width as 1mm. and 1mm on rev. D I'm using 1mm. Should I use other value? What is your suggestion about this?
What other suggestion do you have to improve my layouts (Rev. C and D)?The ground plane is equal on C/D layouts.
Regards, Jeferson.

And I just received the response from the PCB supplier, as follow:The dielectric constant of common FR4 board is ≤ 5.4, usually between 3.8 ≈ 4.6

Do I have to place ground plane on the other side of the board (red layer of layout) where the RF amplifier compinents are pkaced, this is important to know.

I recommend that you do include red ground plane where the oscillator components are located, as you show for rev. D. I would cut the ground plane back a bit, so there is more open area around the antenna, while still including ground plane where the oscillator components are.
Your board's dielectric constant is about 4, so the refractive index is about sqrt(4) = 2, and the wavelength is about 690 mm / 2 = 346 millimeters on the PCB. The length of your antenna track is small compared to the wavelength on the PCB, so can be treated as an inductor.
Your challenge, is to adjust the shape, width, and length of your antenna track, and the value of C5, to resonate the antenna at 433.92 MHz. This can be modeled, but I have no experience with the appropriate software. Do you have a means of measuring the RF output of your transmitter? If so, you can adjust the antenna and C5 experimentally.
You might want to replace the antenna track with a length or coil made of #12 bare wire, which you can shape or bend as needed.
Be sure your transmitter does not exceed FCC legal limits.

Hello dear can you please share your final result which one worked well I am also in a similar situation it might help me and others