Hey ho. Been quiet for a couple of days making test gear. ULTIMATE TEA. Something simple for the weekend. Missing a calibrated signal generator at the moment which is a big problem and to be honest getting annoying now. Mainly because you can't couple a frequency counter to a VFO circuit because it shifts the frequency without massive buffering. This is an issue because my VFO is a black box oscillator (half an SA612AN) so I can't get at it without resorting to friggery or moving the VFO out of the mixer. Function generator isn't stable enough for such tasks. Gah.
1970s solution incoming.
So in the junk boxes I went digging and found some 4MHz crystals, a tube each of 74LS74 and 74LS90 ICs. Scratchy head says if I knock out a square wave from these at say 1MHz, 100KHz and 10KHz I can then get calibration marks at all harmonics (thanks Fourier!).
So little circuit. I didn't bother doing a schematic but basically it starts with a 4MHz colpitts crystal oscillator. The loop capacitance was totally wrong when I picked capacitor values out of my arse to start with so it oscillated on an overtone of about 11.93MHz. Quick frig around with the loop capacitance to get it to roughly the load capacitance of the crystal (30pF) and it's bang on 4MHz. Buffered this with a simple transistor switch with some feedback to keep it stable, then whack it into two stages of D-type flip flop to get it down to 1MHz. This is then let out via a 100pF cap and the rotary switch. Then into a 74LS90 to divide by 5 and then 2 (not 2 then 5 - quiz: why not?). Out via another 100pF cap to 100KHz. Then into the final 74LS00 to err... hmm. Never assume that the IC you want is actually in the tube. This was promptly desoldered and widlarized. Another actual 74LS90 was then soldered in and out via another 100pF cap to get 10KHz. RG174 was used (grounded at one end) to stop crosstalk between the output lines. The switch controls power as well so you select (1) OFF (2) 10KHz (3) 100KHz (4) 1MHz.
Popped it on the counter and poked the trimcap until it was spot on. And it sits there on the 1MHz out at 1.0000000MHz according to my TF930. 1Hz drift either way occasionally. Pretty good!
Ugly but it works. PCB is some FR4 offcut carved, tinned, cleaned. Enclosure was from another unfinished project. I need to find the lid for it. I'll stick a nice label over the big hole in the front one day.
Here's what it sounds like. Note that the harmonics are good to 6m band (50MHz!) from just the 100KHz output. Set to CW with narrowest DSP filter width there. If you look at the S-meter at the top left you can peak it on that rather than the audio.
Right. Bed.