In my amp I used electronic switches in stead of mechanical ones. That was also my first time doing this, so I did also a few measurements to look for interesting issues.

For now, without any changes, it worked ok. But let's look at the real performance. 

By the way the scheme is from QRP-Labs, I got it from the internet with a very good explanation how it works. Look at QRPLabs in the building instruction of the 50W amplifier. By the way, for the money you can best buy the kit from QRP. It saves a lot of time searching for the components and it is very much worth the money.

My measurements on the electronic switch circuit are:

- measured the speed, which happens to be very much fast enough for QSK.

- my capacitors at the outgoing decoupling got warm, the full power goes through both C's and they were out of the junckbox. So get some good ones. Very good are the smd versions of the multilayer high voltage capacitors.

- The diode at the outgoing circuit is also getting warm, it dissipates some energy of course, because of the 0,7 volt over the device. If you want put a bigger one in it.

- I looked with a scope on the signals of the switch circuit, and saw no oscillations or whatever. What I saw was a some kind of start up effect at the beginning of the signal on a dot and a dash. The frequency of it was low, and it was damped out after about a milisecond. I dont have a problem with that. So let it be.

- The switching resistors draw quite some current, and also got warm, even hot after a minute or so. Best is to fix them on the cooling block, and over dimensionate them.

Furthermore the switching circuit works fine. I am wondering whether this kind of circuit works also with 1kW. Probably you need a bigger diode at the end, and VERY LOSSLESS Capacitors.