I have used a few 433MHz ASK transmitters. They're cheap, readily available, and work reasonably well. Based on this application note from Murata I tried to build my own. I replaced the NE68030 transistor with a BFR93A, and the RO3101E SAW resonator with a generic one. The antenna I'm using is this one.
The appplication note has instructions to calibrate it. The SAW needs to be removed and have a chip capacitor soldered instead. Then measure the free run frecuency. In my case I couldn't really get a stable frequency. In the datasheet of the SAW resonator I noticed they don't use L2, so I removed it. Once I did that, the free run frequency jumped to 470MHz or so.
Why does this happen? Is it because I'm using an antenna that is already matched to the frequency and does not require L2? (I'm guessing this transmitter is designed to be used with a 30-35cm lenght of wire as an antenna).
So anyways, I tried to solder the SAW. Once I did that nothing happened. It wouldn't transmit. Until I touched the antenna, and it would be near 433.92MHz. So I soldered L2 back again, and no output. I tried with different values for L2 and finally a 56nH value seemed to work. The frequency was measured near 433.90MHz, but it seems to be very weak. The only tools I have are a frequency counter with an antenna connected to its input, and a UHF portable radio. I have no VNA or even SDR.
How can the output be improved? The datasheet says L1, C1 and C2 need to be tuned for the particular application. In my case it was actually L2 that made it work. The only difference is that I have no 6pF SMD capacitor handy so I just used a really old "lentil" type TH capacitor I found.