unclean waveform while tapping transmitter signal from NPN emitter

Here is another attempt at me making a transmitter at about 433Mhz. Eventually I'll tune L1, C1, and C3, but the major problem I have here is that I took an idea of tapping the signal off the oscillator NPN emitter instead of the collector. When I run the simulation I get a rather choppy waveform (see attachment) showing no real modulation. If I strip everything to the right of the 5p capacitor except the battery (of 5VDC), then I can probe the voltage of the NPN collector and a clean waveform will result, however the voltage reading is maybe 12 at most which won't be much for transmitter distance.

Rather than me upping the voltage, what else can I do to fix this? Is R3 and L2 and C4 and Q2 wrong? I'm lost.

Please attach your spice file. Then some one can look in detail.

I couldn't upload the plot file (because .plt wasn't allowed) but I did upload my circuit file just now. It was made with LTspice V4.

Do you have the files for 2sc3357? I don't have anything. I need something like this:
.model FZT849 NPN(IS=5.8591E-13 NF=0.9919 BF=230 IKF=18 VAF=90 ISE=2.0067E-13 NE=1.4 NR=0.9908 BR=180 IKR=6.8 VAR=20 ISC=5.3E-13 NC=1.46 RB=0.023 RE=0.0223 RC=0.015 CJC=200E-12 MJC=0.3006 VJC=0.3532 CJE=1.21E-9 TF=1.07E-9 TR=9.3E-9 Vceo=30 Icrating=7 mfg=Zetex) ; (C) 1993 ZETEX PLC, Last revision 14/3/97

The .plt files are very large and don't say much.
Thanks for the .asc file.

I downloaded the package from ltwiki.org but they have the individual files on their FTP.

This is their transistor model file: http://ltwiki.org/files/LTspiceIV/lib/cmp/standard.bjt

And download everything in http://ltwiki.org/files/LTspiceIV/lib and replace the lib folder in your LTspice with the folder contents on the server.

i don't think there is a problem with the circuit , altough the wave displays somewhat ??? wavy

i feel myself at home up to 20 ... 30 MHz but not at above it -- i just checked it from "Spice technical" point of view -- can't much credit my oppinion