Do you say that frequency drifts of some MHz and noisy signal are acceptable? What is the application of your oscilllator?
My application is to measure distance to objects that do not move (no doppler effect). Maximum distance is 10 meters, actually I need only 5, but i used "double rule": If it works better two times, then it easily get that 5 meters. In this case MHz drifts are not acceptable. But what I thought of is it can be self-compensated. Here is my "theory":
Code:
light_speed=299792458 ' light speed in m/s
light_speed = 299792458
distance_to_object=10 ' maximum distance to object in meters
distance_to_object = 10
' s - distance to the object
' v - speed of RF wave
' multiply by 2 because wave travels s two times:
bounce_time(s,v)=2*s/v
time_interval1=bounce_time(distance_to_object,light_speed)
time_interval1 = 0.000000067
' result in seconds
I am going to use Wilkinson power divider, put one half of signal to TX antenna, and the other one into the Schottky 2-diode mixer. RF wave travelling time is small (0.000000067 seconds), so I hope local oscillator will not provide too rapid glitches before transmitter wave come back from RX antenna. Of course some noise will occur. But then I thought a new idea: as i do not need so measure doppler shift, only small distances, then i can do more rapid modulation. More rapid means faster voltage changes in modulation "saw". As modulation is faster, beat frequencies for 1..5 meter distances will be more different. Even if i get some glitch it will be in near-distance readings. But phase noise still occurs on the modulation period, wich is much bigger than 0.000000067s. Modulation period is something between 1..10ms and if big phase noise comes then i get wrong measurement. But then I thought another good idea widely used in spectroscopy: if nothing moves - then accumulate. For example, I get 100 readings, it is only 100*10ms = 1 second. I dream that can get very acceptable results with that heavy modulation and accumulation. After all I want put this "idea" in the matchbox with some low-power ARM microcontroller, Li-accumulator and bright LCD on the back side with some "1,2,3,4,5m" readings. Maybe eraphones jack on the left side to hear some lo-frequency glitches or to connect to PC sound analysis software..
At the moment I did not learned much in microwaves, i stucked with searching proven microstrip design photos. Actually I can't find any images in google that I did not saw before. I have a wide collection of this photos now, some TV/sat recievers, motion detectors, mixers, filters. Most of them using DRO. Of course DRO is "ideal": it is small, it is easy. I read about temperature frequency drifts, and how DRO is good at handling this to few MHz for wide range of temperatures and even smaller, easy feedback design, etc. So tell me can I do it without DROs?
update: i am trying to check simple low-freq oscillator higher harmonics theory (as in my first message), as it seems one of the easiest way, but with many new problems. I was able to hear something on 10GHz, but the signal is very low. Lo-freq oscillator directly connected to the mixer input through capacitor. The other input is voltage regulated DRO at 10GHz. Because signal i can hear is too low, I am afraid it is some interference with other harmonics and/or from power source or computers sound card line-in input, etc.. Can 1GHz oscillation circuit provide lower harmonics somehow? The most difficult part is to tune VCO to find this frequency, i use about 30 minutes warming-up of both circuits and the same time for tuning. Signal analysis performed is 8192 points FFT gives very poor result. If I use a screwdriver to "tune" lower-freq generator by toching vital parts, i hear only noise, so it gives me some hope about true 10GHz harmonic found. Yes, i know, very silly experiment and waste of time..