Gigahertz square wave

Discussion in 'The Projects Forum' started by Peter Frazier, Jun 3, 2015.

1. Peter Frazier Thread Starter New Member

Jun 3, 2015
6
0
Can anyone please suggest a circuit that can output a squarewave 1 gigahertz or higher?

Thanks

File size:
18.2 KB
Views:
55
2. JDT Well-Known Member

Feb 12, 2009
658
85
Consider that a square wave consists of the fundamental frequency plus a large number of odd harmonics. The more the better (squarer). So to have a 1Ghz square wave you need to be able to output and transmit at least up to 15Ghz and preserve the phase relationship.

This is done on the surface of computer processor chips but the signal is only travelling micro-meters and is not square anyway!

3. Papabravo Expert

Feb 24, 2006
10,340
1,850
At those frequencies a demonstrably square wave is an impossibility with lumped parameter circuit elements. You need microstriplines, a CAD program, and test equipment that would strain any hobbyist's budget.

4. Peter Frazier Thread Starter New Member

Jun 3, 2015
6
0
What is the highest stable square wave possible?

I have used a 2.7 mhz 555 astable circuit.

Is there an IC that can produce a higher astable output?

What is the maximum stable transistor astable multivibrator circuit?

Thanks.

5. AnalogKid Distinguished Member

Aug 1, 2013
4,685
1,297
Your curcuit as drawn probably is not possible. Standard LEDs have turn on and turn off times of 10 to 50 ns, over 10 times too slow to be driven by a gigahertz signal. Also, a discrete transistor driven as a saturated switch can not turn off fast enough to do what you indicate.

What is this for?

ak

6. RichardO Well-Known Member

May 4, 2013
1,336
429
You may be able to drive the LED with a very fast pulse if that pulse does not have to be 50% duty cycle. An avalanche pulser can give sub-nanosecond risetimes. The limitations are that the pulse width is only a few nanoseconds and a pulse repetition rate of more than a few hundred KHz can be difficult.

7. Papabravo Expert

Feb 24, 2006
10,340
1,850
Just for fun ask yourself what effect a little bit of series inductance will do to a 1 GHz signal.

2*pi*1 GHz* 10 μH ≈ 62.8 kΩ

That is a fairly healthy impedance for a source to over come.

How about the impedance of a 1 μF capacitor.

1 / (2*pi*1 GHz*1μF) ≈ 159 μΩ

That is a virtual short, hopefully not to GND

You begin to appreciate the difficulties in circuit construction with lumped components at these frequencies.

8. Peter Frazier Thread Starter New Member

Jun 3, 2015
6
0

What then is the maximum stable frequency for a square wave that can be generated?

Like I said I have found a 555 ic that puts out 2.7 mhz. Is there another ic that can generate a higher stable square wave frequency?

Thanks

9. Papabravo Expert

Feb 24, 2006
10,340
1,850
I cannot assign a precise meaning to "stable square wave frequency". Stable is different than square. For example you could say that to be square, the rise time must be less than 1% of the period. So for example, a 1 nanosecond rise time would require a 100 nanosecond period which is the same as 10 MHz. I think that making a square wave in excess of 10 MHz. is actually a fair challenge. Most of the oscillators and clocks above that frequency have rounded shapes and start to look more like sine waves than square waves.

10. Peter Frazier Thread Starter New Member

Jun 3, 2015
6
0
Thanks for the response.

11. DickCappels Moderator

Aug 21, 2008
2,758
667
Maybe if you give more detail about what you want to accomplish, some potential solutions will be offered.

Aug 23, 2012
4,925
777
13. Peter Frazier Thread Starter New Member

Jun 3, 2015
6
0
Can you suggest and IC or circuit that is capable to go up to 10 mhz?

14. DickCappels Moderator

Aug 21, 2008
2,758
667
I was just looking at the MM74HC4046. Good for 12 MHz at least as either a VCO or PLL.

15. Peter Frazier Thread Starter New Member

Jun 3, 2015
6
0
Thanks for the response.