Hello everybody,
This problem solved, more poping out ! ( No problems, no fun

)
I got the mentioned problem solved, thanks to the inputs of this groups and an excellent customer care by Matti. Thanks Matti!
It was a faulty ACC cable as a major cause. After getting tired of checking and re-checking the quality of my own soldering skills on the Db15 connector I have decided not only replace the DB15 connector on the ACC cable supplied from the factory but make an entire new cable of my own. It means a new db9, db15 connectors and a 1m of 8 wire computer cable. Bingo! The erratic behaviour of the K3 stopped immediately and never came back! It was one rebaund when I changed back to the Samlex 1223 switching power supply. It is most probably one of this " floating type" accordingly to Matti and doesn't go together with the PA1000 design. So I stay now with an old good Kenwood PS-20 supply and it works just fine. After a week now of intensive use including the Baltic Contest the forementionned problems do not come back.
BUT! During the Baltic contest on this past weekend a new interesting phenomena poped out! For the contest, I took my JUMA PA1000/K3 set up to another location in the countryside. All the interconnections were the same as at home - good ground wire between the rig and the PA, the new good ACC cable, the old good PS-20 power supply. After switching on the gear and not yet have transmitted anything to the air, I encountered with the strange behaviour of the JUMA PA1000. With the OPER button on, the TX LED started glowing slightly, then started blinking with the regular intervals like 2 times a second, together clicking the PTT relay and after minute or more totaly went to the TX mode with the TX LED glowing full color. Pushing OPER again made the PA come back to the RX. pushing again - the PA goes into TX. The OPER button have become like a PTT pedal really,

. The whole PA was woking perfectly as usual, I was taking part in the contest very nicely just had to manualy switch the OPER button for eqch qso,

. The phenomena went away for an hour or so during the contest, but then came back again. Most interesting part - when I came back home from the field position and reconnected everything exactly the same way in my fixed position - no problems, the OPER button and the whole PA is working just great!
So, it is very interesting to know the possible origins of such an event because I intend to carry the PA to possibly more other locations.
In the countryside I was using the Vertical antenna with the radials on 80 m with a perffectly tuned SWR to resonance and the RF choke. I exclude thus the " RF in the schack " problem and because this behaviour was in place even without any antenna connected to the PA and in the RX mode. I exclude also the " check the cables " advise because the cables are good. The erratic behaviour though stopped when I disconnected the ACC from either the rig or the PA or switched the K3 off. The restarting the PA or the K3 made no change.
The house itself is a newly built with an up to date electrical installations with a standard AC outlet sockets with the ground wire.
I wonder could it be that the OPER button circuitry reacted to some interference from the computer modems in the house or security systems or the LED lightning? But the whole PA is in a perfectly shielded box! In general it is a very clean conditions there for the HF and there is no local QRM for the reception.
Thanks again everybody and looking for more food for thoughts!
73 Linas LY2H