FM modulator - Some transistors does not oscillates

[ppTRN]

Hy everyone,
I built for didattic purposes an FM modulator that works perfectly in a range of a few tens of meters, even if built on a breadboard.
I wanted to try different transistors for the oscillator but I found out that even if almost identical to the BC547, many does not oscillates.

Here's what I tried and worked:
BC547
2N2222
BC108

What did not work:
BC149
2N3904
S8050

Is this behaviour only related to the transitioning frequency of the BJT or are there other parameters (stray capacintances or input capacitances of the transistor?) that does not allow the circuit to work?

[jwet]

BC149- higher power device- large capacitances,  real slow
2N3904- should work- similar to 2N2222, might need to kick up bias current a little
S8050- higher power device- large capacitances, gets slow

Two of those transistors are power devices and have large capacitances and marginal Ft's.  The Ft's are also specified at pretty high currents- 10 to 50 mA.  You don't show Vcc but assuming something like 9v, you're a little light on Ic.  Luckily oscillators only need a gain a bit over 1, so a transistor with Ft of 200 or so should do ok if biased where it has this gain.  In your circuit, with 8.5v input (9v battery), the transistor is biased 4.35 Vb, 3.6v Ve and 3.6/.56 or about 6 mA, Ic.  This is likely right on the line for the 3904.  A nice UHF transistor used up to 500 Mhz is the MPSH10, a notch faster than those you're using, lower capacitance, etc.   Directly driving an antenna is not a good load for an oscillator- in order to get good Q, the impedance of the collector circuit should be high for an oscillator but driving an antenna is a low impedance need .  Antennas have a capacitive low impedance- especially if they're short (< wavelength/4 about .75 m)- a matched antenna would have an impedance around 100 ohms- can kill oscillator.  The trick in these little circuits is to lightly couple power out of the oscillator so it keep oscillating but gives you something to radiate.  Play with the antenna a bit- longer/shorter, cap in series (47 pF), etc.  You can also couple the antenna through a secondary coil wrapped over the oscillator inductor to couple out some signal- maybe 6 turns of small wire over the coil to a balanced dipole that is near resonant.  If you're really trying to get more power/range, consider using an oscillator (like you have) feeding an emitter follower stage that feeds a power stage- maybe a common base amp.    The Ham's  ARRL handbook- any year is a good reference.  Big topic, good luck- have fun.

[G0HZU]

The internal capacitances of the BJT will affect how it performs in your circuit and having a faster part with less capacitance might not be a good thing in this case.

Given the right circuit, a genuine 2N3904 can typically oscillate up to about 1GHz, and an MPSH10 can probably oscillate up to about 2.2GHz. I think that a BC108 might manage to oscillate as high as 500-600MHz and a 2N2222A might manage 500MHz.

It all depends on the circuit around the transistor.

[szoftveres]

While studying many older VHF-UHF receiver circuits I've observed that usually lower ft devices are preferred in oscillators, since they rely on higher internal capacitances.
In the above case, the Cce (5pf) and the Cbe internal capacitances complete the Colpitts oscillator. I'd try increasing the 5pf to something like 7pf or 10pf and see if those other transistors will start working as well (the result might be however that the previously good transistors will stop working).