Circuit to generate 2 non-overlapping clocks in GHz range

[shockingshockley]

Hello. I am working for a circuit to generate 2 non-overlapping clocks in GHz range. However, the circuit only works in the MHz range. What are the probable reasons why it does not work? Or do you have any suggested circuit for this purpose? TIA

 

[KlausST]

Hi,

You don't show a schematic, nor a code, nor any simulation results, nor do youcsay which hardware you use.
What is your idea? How can we help?

Klaus

 

[shockingshockley]

Hi,

You don't show a schematic, nor a code, nor any simulation results, nor do youcsay which hardware you use.
What is your idea? How can we help?

Klaus

Sorry, here is the schematic of the inverter and simulation result. I have a triangular output instead of the pulse output.

 

[d123]

Hi,

I shouldn't be here - my speciality is spectaculary bad analog design full of hobbyist mistakes, but still, in case the main comment is useful...

I'm used to seeing non-overlapping clock generators done with logic gates of varying complexity, this being the typical simplified one you'll be familiar with:

The other day, not seriously looking into discrete non-overlapping clock generation using standard logic and BJTs, I noticed that the trailing edge of both signals needs to be shut down by its rising complement, otherwise they overlapped (throwing capacitors at the circuit does not solve the problem, just displaces it in time) - I see that yours also overlaps in that way in your space invaders picture. In my wayback machine cartoon you can see what I mean. I doubt the wayback circuit even works in reality, btw, and the (1uF) risetime delay caps have to be selected based on the actual clock speed, it's a NRND circuit, it's just to show what I'm referring to:

How you'd get your circuit to GHz speed, no idea, surely that's related to MOSFET capacitances?

 

[dpaul]

This is just posted at the wrong sub-forum!

 

[Grimpeur]

Which forum is best for logic circuit questions like this?

 

[d123]

Hi,

It depends what the person who started the thread is doing. Maybe it would be better classed under 'Analog Circuit Design' or perhaps 'Analog Integrated Circuit (IC) Design, Layout and Fabrication', or possibly 'RF Microwave, Antennas and Optics'. I have no idea as I do not know what their end goal is.

 

[Grimpeur]

Hi,

It depends what the person who started the thread is doing. Maybe it would be better classed under 'Analog Circuit Design' or perhaps 'Analog Integrated Circuit (IC) Design, Layout and Fabrication', or possibly 'RF Microwave, Antennas and Optics'. I have no idea as I do not know what their end goal is.

Definitely not "RF Microwave, Antennas, and Optics." "Analog ICs" could work in principle, but we really need a separate space for digital circuits, layout, logic gates, and logic solutions, and nothing looks obvious in the top-level categories. Analog guys are all about PLLs, power supplies, bandgaps, antennas, transmission lines, etc. In digital, we generally try to pound all analog signals into square waves or flat lines, and pack everything as densely as possible. This is relevant to ASICs far more than to Analog ICs.

I've seen a few digital circuits/gate solutions appear in "Digital Signal Processing," but that forum is primarily about the math. Also in "Elementary Electronic Questions," but these concerns are beyond elementary.

It is a distinct subset of ASIC design methodology, but I'd much prefer to have a category for "Digital Circuits and Logic."

On a related note, where would "Mixed Signal Design" belong?

 

[dpaul]

@Grimpeur ,
Either at Analog Circuit Design or Analog Integrated Circuit (IC) Design, Layout and

​