[time-nuts] Square to sine wave symmetrical conversion (part 2)

Alexander Pummer alexpcs at ieee.org
Sat Jul 25 19:05:33 UTC 2015


it is relative easy to make a perfect 50% square wave from almost any 
input wave form

U1 could be any-- fast enough  for the desired frequency--comparator or 
a transistor pair similar to Charles Wenzel's  circuit, R1 C1 is a long 
time integrator,[ RxC >> 1/f of the incoming frequency] the voltage at 
"A" is proportional with the duty cycle, U 2 is some high gain low 
noise, low input offset voltage high input impedance amplifier, the 
duty-cycle is set by R4/R5,  fine tuning with R6, C2v removes the noise 
Vr is well stabilized reference voltage,
To set up the circuit the output should be connected -- with DC 
decoupling -- to a spectrum analyzer's input for watching the second 
harmonic [a perfect 50% duty cycle square wave lacks of even harmonics 
..] of the input frequency, which has to be adjusted to minimum with R6, 
using the same stile resistors for ±0,1%, R4 and R5, with value 100 
times of R6 a very good temperature stability could be achieved. for 
better short time stability R2 's top could bealso connected to Vr,
If the drive capability of U1 is not enough for the load non-inverting 
buffers could be inserted to U1's output, of course that circuit wil 
contribute some phase noise/ jitter too.
73
KJ6UHN
Alex


On 7/25/2015 1:34 AM, timeok at timeok.it wrote:
>
>> bypassing the inverter you will improve phase noise. Yuo will probably
>> need a sine buffer at 10MHz to drive 50 ohms.
> This separator con be the solution:  http://www.timeok.it/files/hp5065AoptH10v200.pdf
> high input impedance, output 50Ohm, high power handling and low additive phase noise.
>
> But before measure using an oscilloscope the output of the GPS OCXO directly on the output pin to verify if there is a sine-wave or square
>
> Luciano
> timeok
> Message sent via Atmail Open - http://atmail.org/
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.




More information about the Time-nuts_lists.febo.com mailing list