The latest progress is encouraging but the practical limits are reached. I need the following things :
- get a proper frequency generator.
Edit (20160328): done !
I have a waveform generator that is useful for certain measurements but it's not practical to generate a given frequency or sweep frequency ranges. My solution : use one of these 16.384MHz oscillators I found, add a 74HC4017 as a variable predivider, and a couple of 74HC4040 to divide down to about 1Hz...
That, or find a cheap generator, but I have most of these parts and it could make a nice HaD project page.
Edit (20160328): done too :-) #Quick & Dirty Frequency Generator - check and compare the differences between two topologies of the 10T FF, with the clock at the ground, or between the bistable and charge pump. I should build both on the same board and spy the differences... Gate capacitances certainly play a role. Feedback resistors too must have some influence... What's the best value ?
- design a faster 10T FF so it can reach 100KHz easily. Maybe a couple of Schottky diodes will suffice. Maybe not.
Edit (20160329): done as well :-) A faster Generalised Flip-Flop - Design/layout a single stage of the 10T FF so they can be stacked :-)
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.
A Rb clock have a relatively low jitter compared to the the 1pps output on a GPS timing unit. But the Rb got a long term drift and also varies the frequency with the room temperature even when properly shielded and ovenized. But the GPS have excellent long-term stability (hours/days/years) since it's are based on cesium clocks in the satellites. So by having the Rb clock slightly pulled into submission by the GPS on a long-term averaging basis excellent results can be achieved.
A bog-standard Rb source is more than enough for any sane persons needs. I'm not a time-nut,but I already had all the parts already so I decided to put it all together just for fun...
Are you sure? yes | no
I'd love to have a nice and small signal generator for digital electronics. Today I've got one of those boat anchors - the HP 3325A. It's kinda nice with microhertz resolution, sweeps, phase adjustment , output voltages up to +-40 volts, and other stuffs. But it's huge and limited to 20MHz sines and 10MHz square.
If be happy with a unit with just 1Hz resolution at a range of 1Hz up to 100MHz. And then have just square output with fixed voltages of 1.8, 2.5, 3.3 and 5 volts. Maybe it's time to design a unit like that based on one of the SiLabs PLL thingies. Not LCD/OLED display though - I like the big and highly visible 7-segments :-)
Are you sure? yes | no
Nice rack ;-)
I have something else in mind, using 74HC stuff first. Powered with a couple of AA batteries, square output. Most of the complexity is with selecting an output and applying a crude predivider... simple, cheap, no programming, useful for testing my clock and the #Discrete YASEP
Then, I would make a sort of DDS with a FPGA. I'm browsing the Rubidium timebases on eBay at this very moment...
Are you sure? yes | no
I've got (or actually had - since it's all boxed up now) the external input of the 3325 hooked up to a 10Mhz Rb-unit that I have connected to a GPS timing module for disciplining. So I have a GPS antenna on my roof... One of my neighbours recognized the shape of the antenna and asked why I wanted to know the location of the house :-)
Are you sure? yes | no
I have some timing GPS receivers (somewhere in my boxes). But no access to the roof, inside my workshop in a large industrial-era building... a Rb osc would be more handy and would cost as much as buying 10s of meters of coax cable. I have an amplified antenna but no clue if it's 50 or 75 Ohms......
Now, I read and suppose that Rb outputs have some jitter, due to the output DDS.
But for now, a tiny TCXO is more than enough, right ?
Are you sure? yes | no