Here is the current physical stage of my Pixie 2 project.   For those not familiar with the Pixie, its a two transistor and one audio op amp transceiver.  This one is built for 40M.

Its built in what I'm calling Ugly Manhattan.  I put the PCB in the vice and cut grooves with a hacksaw to make the pads.   I like this method, but the pads are bigger than they need to be, and they dictate how the parts can be placed. I've left space for further mods, but if I don't add any then will just cut off the last two rows of pads.

Stage one was to model the Pixie in LT Spice.  If any of you have thought about learning to model in Spice, I encourage you to give it a shot.  Its pretty easy to enter a simple QRP circuit, and its great fun to "Run" it and see the readings at various points.

Once I had the Pixie working in Spice, I tried some mods on it, like adding a phase shift oscillator for generating a sidetone, but couldn't get it to work in Spice.  In fact, I couldn't get it to work on the breadboard either.  I'm going to have to figure out why at some point in the future, but for now I will use my NorCal keyer (which I just built) and it will inject a sidetone into the Pixie's audio chain, hopefully.

Note the two crystals and outboard variable capacator: the Super VXO configuration.  Actually I think it still needs an inductor in series with the capacitor and crystal.  As it is it tunes about 6 Khz.

This is the Pixie, transmitter only, on the Prototype board.  I found this part to be a bit flaky, as I was never really sure if there were good connections at all points, at all times.  To confirm my concerns on this, I note that the rig is much more physically stable when built on the PCB and doesn't have a lot of the noises that I thought might be bad connections on the Prototype board.

Of course the beauty of the Prototype board or Breadboard is that its real easy to plug stuff in and try it and if it doesn't work, then plug in something else.

On the other hand, tracing the circuit when its installed on the board can be tricky.  I made a "map" of the board on paper, drew in the components, connected to their 'row' on the board.  In this way I found one wiring error that I had made.

The whole idea of my "Pixie Project" is to learn.  Learn to use Spice modeling, play with adding modifications, get comfortable with building on the Prototype board and try out different building methods: ugly, Manhattan, etc.

I haven't made a contact with this Pixie yet, but I'm blaming that on messing up my beautiful 550' loop.  A couple of weeks back I added and chopped on this loop to try and get it to load on 160M.  It sort of does that now, but doesn't load well at all on 80M and 40M, so as soon as it warms back up to 0 C, I will drop the loop again and put 50' back in.