N8YX
06-26-2017, 11:32 AM
Anyone who's owned or operated an early generation Kenwood or Yaesu hybrid is undoubtedly familiar with the provision for operating the rigs on a fixed frequency, via a crystal which is fabricated to a specified value then inserted into a socket within the radio. The crystal (or crystals) are then selected by means of a front panel switch.
Sounds easy and cheap enough, right? 'Twas back in the day, when we had a variety of custom crystal manufacturers from which to choose. Nowadays, we have precisely one - Quartslab. At $20/rock, filling up a bank of four in your Kenwood might be acceptable
15392
but loading up all 11 channels in an FT-301 costs as much as a used example of the radio itself!
15393
What to do?
I gave this a little thought over the past week or so. The use of a DDS as a replacement for a 5.0-5.5MHz analog VFO is getting to be fairly common practice, so why not extend that concept and use one to generate the same frequency ranges for those crystal oscillator circuits?
One could use a PIC to scan the crystal-select lines. Given the devices only have a few inputs, a 16-4 encoder can be employed for the Yaesu setup. A set of data words (dependent on the DDS) is programmed into the PIC. Pseudocode is along the lines of: While the Select circuit is NOT in VFO mode, scan the Crystal Selection lines and see which position is selected. Then, send the appropriate string to the DDS. While that particular Select line is enabled and the current working frequency is the same, do nothing. If it changes, send the new data to the DDS.
Upon selecting VFO again, an initialization sequence is sent to the DDS so it'll go into dormant mode and not generate unwanted output.
There is a bunch of information online about getting a PIC to send serial data.
Anyone else think this is a neat concept?
Sounds easy and cheap enough, right? 'Twas back in the day, when we had a variety of custom crystal manufacturers from which to choose. Nowadays, we have precisely one - Quartslab. At $20/rock, filling up a bank of four in your Kenwood might be acceptable
15392
but loading up all 11 channels in an FT-301 costs as much as a used example of the radio itself!
15393
What to do?
I gave this a little thought over the past week or so. The use of a DDS as a replacement for a 5.0-5.5MHz analog VFO is getting to be fairly common practice, so why not extend that concept and use one to generate the same frequency ranges for those crystal oscillator circuits?
One could use a PIC to scan the crystal-select lines. Given the devices only have a few inputs, a 16-4 encoder can be employed for the Yaesu setup. A set of data words (dependent on the DDS) is programmed into the PIC. Pseudocode is along the lines of: While the Select circuit is NOT in VFO mode, scan the Crystal Selection lines and see which position is selected. Then, send the appropriate string to the DDS. While that particular Select line is enabled and the current working frequency is the same, do nothing. If it changes, send the new data to the DDS.
Upon selecting VFO again, an initialization sequence is sent to the DDS so it'll go into dormant mode and not generate unwanted output.
There is a bunch of information online about getting a PIC to send serial data.
Anyone else think this is a neat concept?