I did a second and 3rd look at my direct conversion receiver and felt that I could have done better if had I few more tools on my bench.
Some of the things which I felt needs improvement are
1. VFO Stability
2. Impedance matching between stages of the multiband vfo.
3. RF Isolation between modules.
4. The multiband VFO suffered from non-flat output between bands
To tackle the VFO Stability I needed a frequency counter. I built my frequency counter using an atmega8 and programmed it using a STK500 compatible programmer I built a year ago. I posted the schematic for the programmer here.
The gate time is hard coded to 2sec sample and a prescaler of 16. There are 3 GPIO free on the atmega8 I am using, so adding a user selectable gate time should be fairly easy. As soon as I receive my 74HCT4040 I will lower the prescaler to 8. Without a user selectable prescaler and gate time the last digit is not very useful, but I did a quick compare of the reading of this frequency counter to a lab frequency counter using a 4.00 Mhz crystal controlled source. It was very close to 40 ppm difference.
There are already a lot of homebrewed frequency counter out there so I will not give out any more details except for a few pictures.
During the prototyping I have found the addon of a ribbon cable ISP connected directly to a bread board very useful
The frequency counter main board was built with headers so that it can be programmed and powered by the ISP directly.
I estimated that the actual coding in C using avr-gcc/avrdude under OpenBSD took <0.5 day. Hardware construction took 1 day and a bit of time was spent mapping GPIO, prescaler selection and front end calculations. The work was spread out through a few weeks depending on available free time and waiting for ordered parts to arrive.