I want my lazy BLX15 back!

I did replace my old BLX15 power transistor with a modern Power MOS device.
The BLF177   the BLX15 has 6dB gain only at 70MHz.
While the BLF177 should have over 20dB.

After a few days  trying all I could think of to get stable power out of the BLF177 MOS   at 70MHz  the result was  2 melted tuning C's.
It sure can generate power this device  but I did not manage to let it  work at the wanted frequency.
Many combinations of L and C  where tested. The original circuit was  used for 50MHz relatively  close to 70MHz.  I did not expect major changes in the L and C's  in use.
The   2 stage low-pass filter should be almost the same.
Of course the bias circuit is different   but simple only a pot-meter to get about 3V at the gate for 500mA. The BLX15 needed the 0.7 PN diode in parallel to get a 500mA bias.
I have  edited  an input file with the scattering parameters from the specification data for  RFSimm99

I need a working circuit with practical L and C values to continue.
It is very strange that I can not find a tuning for this device.  Only oscillating ?  At one moment it looked ok  I could drive easy 150<  in the dummy. But then suddenly at switching in the tuning C burned out and fuses blown.   At lower Voltage I managed to get 20W with 1W drive but very unstable.

Victims of the shear oscillating power...  Arco type 406  80pF  and a smaller one