Building a simple transmitter

[Faranight]

Hi, I'd like to make a transmitter for a project of mine. Yes, yes, I know there are regulations about transmitting and I have some questions regarding this as well.

Anyhow, I would like to build a transmitter for the 433.92 MHz ISM band to communicate some sensor data over a distance. So far I've been using those cheap transmitter boards with an Arduino, but now I'd like to build a proper board with a nice SMA connector (yes, the RF traces will be matched to 50 ohms), sensors, power regulator and a microcontroller on it. A digital output pin on the microcontroller will be used to relay a 3V3 signal to the transmitter and turn it on and off to get the ON/OFF keying signal modulation.

I've googled a bit and found schematics for the TX boards that I'm using (see pic below). They use a 433MHz SAW resonator in a TO-39 package that is connected to a 2SC3357 RF transistor to generate a carrier signal. I thought I'd copy this design, but I'll use RO3101D-1 and BFU590 as replacements. The BFU590Q datasheet and its application note have a nice 433 MHz reference design (evaluation board) with matched components (see pic below), but it uses a coaxial input line rather than a digital one. With the help of some more circuits that I've found online, I thought I'd merge these two circuits together, remove the coax input and replace it with the SAW resonator to allow a digital input to control the state of the transmitter (on/off). I've posted a pic of the circuit (pic 3) that I made in KiCad, but I have some questions.

• The BFU590Q base in the reference design (pic 2) is biased with 0.78V from a filter circuit. I've replaced this part so the transistor base is now biased from a 3V3 digital pin on the uC via a current limiting resistor R3. Is this design correct? In the ref design the base seems always biased even when there is no AC signal present. In my design it will be biased only when the microcontroller is outputting a HIGH signal on the pin. The transistor base bias current is important and must not be saturated. A resistor can be used to limit the current. The datasheet shows BFU590Q some important charts about the transistor electrical characteristics.
  
  
• Do I need to add a pulldown resistor on the uC output pin to avoid accidental transmissions? Suggested.
  
  
• The SAW resonator has a different placement in my circuit compared to the cheap TX board. It will be powered from the microcontroller pin. I've seen some designs do it like this. I assume this is ok as a carrier signal generator? In a pierce oscillator configuration the crystal should be placed as shown in the first picture. A N-MOS could be used to turn the circuit on and off.
  
  
• I removed the caps and inductor that the BFU590Q ref board has on the input. Is this ok or do the inductor/caps need to be there? Some other circuits I've found on google don't use any other components - they connect a data output pin via a resistor to the transistor base and add a saw resonator in parallel from base to ground like in my KiCad schematics. The input caps and inductor on the ref board were probably a matching network to match the 50 ohm input to whatever impedance the transistor base is. This question is redundant when a N-MOS is used to modulate the data into the carrier signal.
  
  
• Since this PCB is going to be used outdoors (wind, rain), do I need to add some protection i.e. a 1MOhm resistor from antenna pin to GND to prevent antenna charge buildup? Probably a RF choke would be a better choice.
  
  
• The BFU590Q is powered from 8V in the ref design. Mine will be powered from a +12V lead-acid battery. What's a good way to reduce the transmitter power in this design? Would it be sufficient to increase the value of R4 in the circuit until the output power drops enough to match the local regulations? Probably the best way, but one can also adjust the base bias current accordingly. That in turn dictates the transmitter power.
  

Thanks in advance

[EggertEnjoyer123]

Your circuit won't oscillate because there is no feedback from the output back to the base. The Vbe of the transistor will always be the same.
If you look at the SAW filter datasheet (https://www.mouser.com/datasheet/2/1100/ro3101d-1916808.pdf) , there is an application circuit for an oscillator on page 2, as well as a RLC model you can use for simulation on page 1.You should probably simulate your circuit first to make sure it works before building it.

[Faranight]

I know, please disregard the questions. I messed up a lot of stuff in this circuit since I didn't understand how it works, and I was mixing up different transmitter designs.
I've done a lot of research since, and learned several answers to my own questions. The circuit type I should be looking for is called "Pierce oscillator" which is exactly what the circuit on the first picture represents.

[antenna]

I know, please disregard the questions. I messed up a lot of stuff in this circuit since I didn't understand how it works, and I was mixing up different transmitter designs.
I've done a lot of research since, and learned several answers to my own questions. The circuit type I should be looking for is called "Pierce oscillator" which is exactly what the circuit on the first picture represents.

Mouser and others carry RF transmitter IC chips specifically designed for those bands.  I even saw one that didn't require a crystal somewhere.  They have entire ISM transceivers on a single little chip for $6...

https://www.mouser.com/datasheet/2/297/NRSAS00034_1-2559940.pdf
https://www.mouser.com/datasheet/2/761/SEMT_S_A0000304737_1-2575441.pdf

[MathWizard]

So at higher and higher frequencies, how much testing, by professionals, is done IRL, if you'd have to solder up test PCBs all the time ? I guess at a big company you'd just have someone print a nice test PCB.

I have copper clad boards. I built a couple of little things on them, but I never compared them to proto-THH boards, or beard boards. Soon I'll make a little AM radio TX/RX. IDK if there's any point to use copper clad for sub-3MHz.