# Highest frequency mankind has ever generated

1. Aug 1, 2012

### y33t

Does anyone know the exact value of the highest frequency generated by man ? The waveform and amplitude doesn't matter, it can be sinusoidal, PWM or what ever.

2. Aug 1, 2012

### Staff: Mentor

Frequency of what, measured where?
Photons in the lab system? At least ~100 GeV, which corresponds to 2*1025 Hz. I would expect that the LHC collisions produced some photons with up to ~1 TeV, too, which adds one order of magnitude to this value.
Photons in any arbitrary system? There is no limit to the frequency, as it just depends on the system where you calculate the frequency.

Relativistic, quantum-mechanical frequency of particles in the lab system? 4 TeV, or 1027 Hz for protons in the LHC. Again, in other systems this number is different.

Rotation of a macroscopic object? 1MHz, with graphene.

Something else?

3. Aug 1, 2012

### y33t

I am talking about highest electrical signal frequency achieved, not optical or etc.

4. Aug 1, 2012

### Staff: Mentor

Send visible light on a ~500nm-antenna, and you have electric alternating current of 6*1014 Hz. Use ultraviolet light and a 10nm-antenna, and you get 3*1016 Hz. The limit is just the material - a wavelength as short as the distance between atoms does not make sense as "electrical signal".

5. Aug 1, 2012

### y33t

Thanks for the information mfb.

I am asking about electrically generateable upper frequency limit. I know this limit is determined by the latest technology available but I don't know the exact numbers.

Answer doesn't involve anything related to optics I guess because the highest frequency generated electrically is around 10-20THz as far as I can remember.

Optical antennas is a different topic and I believe there can be encountered so many difficulties generating that scale of frequency (6*1014 Hz). If it was 'that' easy, 600THz processors would be on the market by now.

6. Aug 1, 2012

### Staff: Mentor

The limit for computing is not the generated frequency, the issue is to work with it. You can produce signals with 600THz, but the logic gates are not quick enough to handle this, and signal delays are another problem.

7. Aug 1, 2012

### y33t

We can generate signals with microcontrollers, FPGA's, external DAC's and etc at any frequency depending on the boundaries of that IC. For example a PIC mcu can generate a 20KHz signal but it can't generate MHz or above. A midclass FPGA can generate a few MHz or GHz but not above. What I was trying to ask is that what is the limit that can be generated electrically . At this point I think we come up with your expression above, changing signal amplitude over time is the generating part and this is determined by the response time of the switching gates.

How about non-contact relays (magnetic) and so on. Can you point me to a paper or writing for this particular problem you mentioned above (lack of speed originating from logic gates) ?

On the other hand, I haven't come across any article or study that they could generated 600THz electrical signal by exposing a monochromatic light source to the wavelength-equivalent antenna. Is it really that easy, hold some laser to a very tiny antenna and you have your 600THz clock ?

8. Aug 1, 2012

### Staff: Mentor

Yes, and the reason is simply that a higher frequency would not be useful.

I do now know why you dislike the optical approach. However, 3 NOT gates in a row can give an oscillating signal, and transistors are close to 1 THz.

See the fastest transistor link.
1THz corresponds to ~200µm maximal signal propagation, and real lines have resistance and capacity in addition to that, which reduces the practical speed.

You have an electric potential oscillating at this frequency. How to use this is a completely different question.

9. Aug 1, 2012

### bcrowell

Staff Emeritus
For microscopic rotation, I believe the champion is rotating atomic nuclei produced in heavy-ion collisions, where the frequency can be on the order of (1 MeV)/h, or about 10^14 Hz.