Need help identifying these RF oscillator circuits

[VooDust]

I'm using a simplistic 433MHz transmitter module from seeedstudio:


https://www.seeedstudio.com/433Mhz-RF-link-kit-p-127.html

It is working well and I would like to recreate the circuit on a custom PCB to create my own product.

However, I have trouble understanding the circuit. I know how Colpitts Oscillators work in common-base, common-emitter and common-collector configurations, but I can't figure out this one circuit.

This is the circuit:



Edit: I got an LCR meter and measured these exact values:

• C1: 6.1pF
• C2: 8.2pF
• C3: 3.2pF
• L1: 52nH
• L2: 92nH
• R1: 47kOhms
• R2: 47kOhms
• T1: RF Transistor: http://www.unisonic.com.tw/datasheet/2SC3357.pdf (not verified)
• SAW: 433MHz resonator: http://www.msilverman.me/wp-content/uploads/2014/01/K1037299581.pdf

I assume these values are +/- 50% because of the manual testing method.

• Given that C1C2||L1 would resonate at 342MHz, I think the values are in the right ballpark, because for resonance at 433MHz, values of 6pF for C1C2 and 24nH for L1 were required. IF this were an LC tank. But it's NOT. Or at least I can't see it in the circuit diagram.
• Seeing how C1 and C2 measured so close and R1 and R2 used the same value, I assume that C1, C2, and C3 are probably of the same capacitance value because I read that using as few different components as possible brings down manufacturing cost, and this is a pretty cheap module, so...
• L2's purpose could be to either decouple the RF frequencies from going into the power supply (but it's battery powered anyways) or to allow some voltage to be developed at the point between L1 and L2, which is needed for gain. At 433Mhz, L2 would equal to 158 Ohms which is a magnitude I have seen in other circuits as well.
• The DATA pin is just to turn the transmitter on or off, I'm using simple on-off-keying for transmission

With that analysis, my conclusions are:

• R1 is a cheap-ass way to bias the RF transistor
• The node L1L2 is the output of the amplifier circuit and the input of the feedback network
• C1, C2 and the SAW resonator would comprise the resonating circuit tank circuit likewise found in Colpitts crystal oscillators. But here's the catch:
• Intuitively, I thought a positive going input at the top of the SAW (intersection L1L2) would lead to a positive going input at the transistor's base, but that would be out of phase! i.e. negative feedback, but positive is required.
• Thanks to the wonderful analysis of https://youtu.be/I4bAfDu6F1k?t=207 I have to revise my statement. Because the feedback network's output is not taken from between the capacitors (the often-cited "capacitive voltage divider"), but between the capacitor and SAW (i.e. an inductor at resonant frequency), the output of the feedback network is negative going, but with equal amplitude. i.e. the feedback network is phase inverting, but with a gain of 1.
• So to fulfill the Barkhausen Criteria for oscillation, the RF transistor needs to have gain > 1, which should be no problem since it's a common emitter amplifier.
• So what's the gain of the amplifier? I have no idea. Datasheet says 50-300 DC gain, so if we take a gain of 200, quiescent current would run from 10mA to 48mA for VCC from 3V to 12V (verified with my bench supply). So if the gain of the common emitter amplifier is its transconductance times collector load (158Ohms see above), so this means 10mA/25mV*158 right? i.e. gains from 63.2 (@3V) to 303.36 (@12V) ?!
• What's the role of L1 then? I think it acts as an antenna, but it's less turns than L2...? By the way I have NOTHING connected to the ANT pin, the device works well without one. But I might add an antenna later, maybe on PCB?

Okay, so as you can see I figured some of this out while writing this post. But lots of questions remain, and now the fun part starts: For the R433A SAW resonator, the datasheet suggests a completely DIFFERENT circuit:

• What advantage could this circuit have over the other? Stability, drift, parts count...?
• IF the RF bypass would short the inductor to ground, C1, C2 and L1 would form an LC Tank circuit like here:

But then, how would the SAW resonator provide frequency selection? And where would the LC tank circuit source current, if it's connected to ground for AC signals?

• At resonance, the resonator would act as a high impedance, allowing all of the bias current through R1 to enter the transistor
• Likewise, for frequencies other than 433Mhz it would shunt some of the bias current
• But I only see a gain loop between collector and emitter, using the "conceptual" LC tank circuit, feeding some part of the input back into the emitter (in-phase). But it's weird that the resonator seems to be left "on the side", not being part of the feedback loop...?

I hope some of you can add some critical analysis to the above statements

[atferrari]

The first one you altered, but not exactly as is, I have it working on my bench right now. The tank is connected between base and common. The feedback coming from the capacitive divider goes to the emitter.

[PA0PBZ]

If you think of T2 as a resistor it's quite close to a Hartley, but how they tune the coil 

[ledtester]

It is working well and I would like to recreate the circuit on a custom PCB to create my own product.

Those 433 Mhz (also 315 Mhz) modules are widely available on alibaba for < $1 a pair:

https://www.alibaba.com/product-detail/Wireless-Transmitter-Receiver-Kit-433mhz-rf_62127437452.html?spm=a2700.galleryofferlist.0.0.736229f1rkyTmg

I'd also look on aliexpress.