Will this FM transmitter work?

Question

I found this schematic on circuit digest.

I have tried to simulate it in LTspice.

That's the wave I got from the simulation. Most of time it's just DC and does not oscillate at all.

Is this going to work as an FM transmitter?

In real life, I want to use this BFQ19SH6327XTSA1 from Infineon instead of 2N22222. (link datasheet) ([download link for lib])5

***************************************************************
*   Infineon    Technologies    AG
*   GUMMEL-POON MODEL   IN  SPICE   2G6 SYNTAX
*   VALID   UP  TO  6   GHZ
*   >>> BFQ19S  <<<
*   (C) 2014    Infineon    Technologies    AG
*   Version 2.1 October 2014
***************************************************************
*.OPTION TNOM=25, GMIN= 1.00e-12
*BFQ19S C B E
.SUBCKT BFQ19S 11 22 33
*
CBEPAR 2 3 1.941E-014
CBCPAR 2 1 5.476E-013
CCEPAR 1 3 7.843E-013
LB    2 20 2.448E-009
LE   3 30 1.706E-009
Rci 1 10 0.002258
CBEPCK 20 30  8.568E-016
CBCPCK 10 20  1.201E-014
CCEPCK 10 30  8.448E-014
LBX    20 22 7.525E-013
LEX   30 33 1.178E-010
LCX   10 11  2.449E-010
*
*
Q1 1 2 3 M_BFQ19S
*
*
.MODEL  M_BFQ19S    NPN(
+   TNOM    =   25
+   IS  =   1.734E-015
+   BF  =   126.2
+   NF  =   0.9968
+   VAF =   52.15
+   IKF =   1.035
+   ISE =   7.715E-015
+   NE  =   1.81
+   BR  =   13.1
+   NR  =   0.9968
+   VAR =   3.079
+   IKR =   0.1214
+   ISC =   6.257E-015
+   NC  =   1.6
+   RB  =   5.552
+   IRB =   2.4E-005
+   RBM =   1.524
+   RE  =   0.2261
+   RC  =   1.814
+   XTB =   0.854
+   EG  =   1.11
+   XTI =   9.516
+   CJE =   4.551E-012
+   VJE =   0.7397
+   MJE =   0.3271
+   TF  =   2.467E-011
+   XTF =   4.69
+   VTF =   6.79
+   ITF =   1.027
+   PTF =   9.546E-017
+   CJC =   1.164E-012
+   VJC =   0.6791
+   MJC =   0.3983
+   XCJC    =   0.007944
+   TR  =   3.412E-008
+   CJS =   0
+   VJS =   0.3601
+   MJS =   0.02997
+   FC  =   0.567
+   KF  =   2.2E-14
+   AF  =   1.6)
***************************************************************
*
*
.ENDS BFQ19S

Answer 1

Made with microcap v12.

Something like this could be "used" ... As pointed out by @Marcus Müller.
It requires some add-on on the resistor values specifically adding "parasitic" capacitors.
See my C4, C3, capacitors and C5, R3 resistor (~ antenna).
The resulting wave "could" be an FM oscillation depending on V2.

Added picture with BFQ19 (= BFR96 ? with some adjustments), modulated 100 kHz ...

.MODEL BFR96 NPN (LEVEL=1 AF=1 BF=9.80000E+001 BR=8.53970E+001 CJC=2.38386E-012

• CJE=5.15057E-012 CJS=0 EG=1.11000E+000 FC=500m GAMMA=10p IKF=2.00000E+000
• IKR=3.08736E-002 IRB=1.00000E-006 IS=2.44463E-015 ISC=3.07780E-015
• ISE=3.95722E-014 ISS=0 ITF=2.48030E-001 KF=0 MJC=2.14438E-001
• MJE=3.77782E-001 MJS=0 NC=1.12019E+000 NE=1.41333E+000 NF=9.87224E-001
• NK=500m NR=9.78180E-001 NS=1 PTF=0 QCO=0 QUASIMOD=0 RB=3.00000E+000
• RBM=3.00000E+000 RC=6.62000E-001 RCO=0 RE=7.17149E-002 TF=1.61902E-011 TR=0
• TRB1=0 TRB2=0 TRC1=0 TRC2=0 TRE1=0 TRE2=0 TRM1=0 TRM2=0 VAF=2.66948E+001
• VAR=4.69684E+000 VG=1.206 VJC=2.98165E-001 VJE=6.00000E-001 VJS=750m VO=10
• VTF=2.83687E-002 XCJC=1 XTB=0 XTF=7.50000E+001 XTI=3.00000E+000) .ENDS

Answer 2

I would at least suggest a high-frequency ferrite bead or resistor in series with the base if using an RF transistor.

Using anything beyond MMBTH10 is overkill and inviting trouble, really. It's not a fancy circuit and it doesn't need or want much.

The simulation is missing parasitic layout elements. Consider the loop of three capacitors here:

^{simulate this circuit – Schematic created using CircuitLab}

These values are very much straining CircuitLab's capability, but for sake of representation, it does work sometimes:

That's about a 1GHz tone riding on (or perhaps dominating over?) the 100MHz mode. The key change is LS1/2 due to just a few mm of trace length between components, including their own lead and body length. Stray inductance accumulates in the ballpark of 0.5 nH/mm so it doesn't take much when you're talking GHz fT transistors.

The key pattern is the capacitors to all terminals of the BJT, making a low-impedance loop that inevitably has some inductance, and HF oscillation (as in, higher than intended) is perhaps inevitable.

Breaking that loop, by adding resistance to one or more pins, particularly resistance only dominant at such frequencies, is the usual strategy. A different topology can also be chosen that is less susceptible to such modes of oscillation.

Note that a ferrite bead acts as a parallel RLC, having low impedance at low frequencies, rising to some (resistive) peak at some mid frequency, then falling above there. When the peak is positioned over the HF oscillation frequency, the impact on the intended mode is small (the FB will still be lossy at 100MHz, but not as much, and more loop gain is available at lower frequencies, making the oscillator more tolerant of that loss) while the impact on HF mode is strong. To use an RF transistor here, consider a small chip FB rated 20-50Ω @ 100MHz, or preferably 50-200Ω somewhere in the 500-3000MHz range.

2N2222 and other general-purpose types are mostly too slow (internal resistance, capacitance, other time constants) to encounter such HF oscillation, hence the circuit is fairly safe to use with minor variations.