As announced earlier today, some information about last nights project will be released as soon as possible. You can have a first glance at the schematics on my webpage http://www.qsl.net/dl1gsj/
under "20m MEPT".
More detailed explanations and keyer source code will follow soon.
Some first impressions of the MEPT in question.
Initially I wanted to squeeze the filter between the 12F629 (8 pin controller in a 16 pin socket) and the 74HCT00, however, I was running out of space... well, ok, in a very first version of this transmitter, the filter actually was there. With the appearance of the 47k resistor in that particular spot, the filter could not be fit in again :-(
Further one can identify the "high quality, high power" dummy load. Well, actually this dummy load really is high power for the setup shown... two quarter Watt 100Ohms resistors in parallel seem good for about half a Watt, right? And well, look at the display of my lab-TV blow...
First of all, it looks like a nice sine signal. The vertical division was set to 0.1V/div, making the signal a whopping 550mV @ 50Ohms or 740µW.
As one further test I will run the two "shaping NANDs" in parallel, just to see what this does.