Digital Space Noise Detector

[jamesadrian]

I would like to make an antenna and circuit that will produce an unfiltered digital output of noise from 100 megahertz and above. I would like to have the output stored in a flash drive. This would be used as a source of binary random numbers. Especially, I would like know people who understand how to do this. I want to market this as a product to be sold at future beacon.com and elsewhere.

Very sincerely.

James Adrian
[snip]

moderators note: removed personal details to avoid spam

 

[Bluejets]

And just what type of noise, amplitude, source are you referring to...likely noise will override your noise....much more detail required methinks....

 

[Delta Prime]

I'll let you have my 5th grade science project it has a 1 milliwatt output.

 

[jamesadrian]

And just what type of noise, amplitude, source are you referring to...likely noise will override your noise....much more detail required methinks....

Bluejets,

Here is one thought for the antenna: A one-yard diameter parabolic disk does not efficiently capture waves below some frequency, but as far as I know, it does not bias frequencies above that frequency.

Concerning the input circuitry, it should not narrow the frequency range it passes to select a particular FM station. Its output is compared to a voltage level (ground). The output of the comparison circuit goes to the high end of the power supply when its input is above ground and goes to the negative extreme when its input is below ground.

The ability of the comparison circuit to react is not unlimited. If all of the frequencies in its input are too high for it, the output it produces will always be too close to ground and no digital ones and zeros will be produced. I imagine that you have a solution for that.

We can discuss the purely digital circuitry that follows the circuitry mentioned above, but I think this much should be addressed first.

Thank you for your interest. This will be a VERY successful product.


James Adrian

 

[danadak]

Noise source design question aka 'How Agilent does the magic trick?' - Page 1

Noise source design question aka 'How Agilent does the magic trick?' - Page 1

www.eevblog.com

Just out of curiosity if you are storing the noise sampled data values in a flash drive why not
just gen them with an algorithm and reduce the problem to D/A side of circuit ? And how are you
going to get the flash speed to regurgitate the RF noise at such high frequencies ?


Regards,. Dana.

 

[jamesadrian]

Noise source design question aka 'How Agilent does the magic trick?' - Page 1

Noise source design question aka 'How Agilent does the magic trick?' - Page 1

www.eevblog.com

Just out of curiosity if you are storing the noise sampled data values in a flash drive why not
just gen them with an algorithm and reduce the problem to D/A side of circuit ? And how are you
going to get the flash speed to regurgitate the RF noise at such high frequencies ?


Regards,. Dana.

Danadak,

No known algorithm is truly random. The noise from the Sun, for instance, is truly random. It can be used in scientific experiments without objection.

James Adrian

 

[danadak]

No known algorithm is truly random. The noise from the Sun, for instance, is truly random. It can be used in scientific experiments without objection.

Not a physicist, but would not a Galaxy have more randomness in its noise
than a sun ? More total possible states due to more quarks, leptons, atoms.....

These questions ?

Just out of curiosity if you are storing the noise sampled data values in a flash drive why not
just gen them with an algorithm and reduce the problem to D/A side of circuit ? And how are you
going to get the flash speed to regurgitate the RF noise at such high frequencies ?

Regards, Dana.

 

[jamesadrian]

Not a physicist, but would not a Galaxy have more randomness in its noise
than a sun ? More total possible states due to more quarks, leptons, atoms.....

These questions ?




Regards, Dana.

Dana,

Any star will do for randomness. If a file is truly random, I cannot get more random. A 5 kilobyte file that is truly random is just as random as a 5 megabyte file that is truly random. Adding truly random sources just gives you more random digits.


James Adrian

 

[jamesadrian]

Not a physicist, but would not a Galaxy have more randomness in its noise
than a sun ? More total possible states due to more quarks, leptons, atoms.....

These questions ?




Regards, Dana.

Dana,

Te digital circuit needs to accept bits art the high rate of the noise production, but it also needs to store serial data and provide it in parallel words for the slower digitals storage.

I am not here for the digital expertise just for the analog front end.

James Adrian

 

[AnalogKid]

Besides all the nonsense.lf you are in the United States then I hope you know you must be in compliance with FCC part 101.115 covering directional antennas especially the front to back ratio, beam-width and cross polarization discrimination for WCS and BRS/EBS band. If not in compliance then you're just blowing smoke!

Per an act of Congress in 1984, exactly 0% of that applies to a ***receiving*** antenna. Local zoning ordinances or HOA rules might apply, maybe, but only for non-electronic reasons. And even then it's probably not a problem unless you want to mount it permanently on a roof. DirecTV's lawyers worked through all of this in 1994.

Of more concern is the mention of a "1 yard" dish. That will be very inefficient at 100 MHz. Ideally you would want something in the 3.5 to 4 meter range. Now THAT will get the neighbors talking.

ak

 

[AnalogKid]

I would like to make an antenna and circuit that will produce an unfiltered digital output of noise from 100 megahertz and above.

In one of his novels, Tom Clancy gets into the encryption of secret communications and using the sun as a source of ultra-random number sequences that are distributed on CD's. I have not done any surfing on this, but I'd guess you can buy such items today - unless they've been supplanted by better and better algorithms.

And finally - why 100 MHz and above. That's a pretty high low-frequency limit for something driving standard logic. Do you really need random-value 8-bit bytes at a 10 MHz rate? That is a massive simplification of the true nature of a noise bandwidth that *starts* at 100 MHz, but it does frame a useful question.

ak

 

[jamesadrian]

In one of his novels, Tom Clancy gets into the encryption of secret communications and using the sun as a source of ultra-random number sequences that are distributed on CD's. I have not done any surfing on this, but I'd guess you can buy such items today - unless they've been supplanted by better and better algorithms.

And finally - why 100 MHz and above. That's a pretty high low-frequency limit for something driving standard logic. Do you really need random-value 8-bit bytes at a 10 MHz rate? That is a massive simplification of the true nature of a noise bandwidth that *starts* at 100 MHz, but it does frame a useful question.

ak

AnalogKid,

If you were to select a range of frequencies to monitor from f to 3f, what range would you select?


James Adrian

 

[jamesadrian]

When two sine waves, A and B, are superimposed, as when they are amplified by the same transistor, both the sum and the difference are created (A + B and A - B). This means that an input of noise can be obtained by a low-cost antenna that receives frequencies higher than one wants, and yet there is a way to select the noise in a comfortable range. Is anybody here familiar with this method?

James Adrian

 

[davenn]

I'll let you have my 5th grade science project it has a 1 milliwatt output.

View attachment 58622

Besides all the nonsense.lf you are in the United States then I hope you know you must be in compliance with FCC part 101.115 covering directional antennas especially the front to back ratio, beam-width and cross polarization discrimination for WCS and BRS/EBS band. If not in compliance then you're just blowing smoke!

you totally misunderstood the OP's requirements

He wants a receiver with a digital output !! not a transmitter

 

[davenn]

Here is one thought for the antenna: A one-yard diameter parabolic disk does not efficiently capture waves below some frequency, but as far as I know, it does not bias frequencies above that frequency.

at 100MHz to 1000MHz a dish is the last type of antenna you want. Just use yagi antennas

You need to be looking at the wealth of SDR receivers that are already out there

 

[AnalogKid]

If you were to select a range of frequencies to monitor from f to 3f, what range would you select?

a) It doesn't matter what I would select. It's not my project and I have no operating parameters.

b) Since I have no idea what the intended application is, I have no idea.

After seven posts, we still have no idea what you are trying to achieve. We do not know if the white noise bit stream is going to the shift register's data input or clock input. If it is the data input, we do not know the clock frequency. The clock acts as a lowpass filter, significantly altering both the "whiteness" and the randomness of the result. This bears directly on the analog front end design.

Or, the white noise bit stream could be driving the clock input, to randomly (in the time domain) sample some other waveform for some other reason. All of this affects the required bandwidth of the noise source.

Three days, seven posts - not one number.

ak

 

[Delta Prime]

I would like to make an antenna and circuit that will produce an unfiltered digital output of noise from 100 megahertz and above.

Apologies to the OP after rereading, it's still reads like he wants a transmitter!
I just wanted to show off my little antenna

 

[Delta Prime]

Per an act of Congress in 1984, exactly 0% of that applies to a ***receiving*** antenna.

Well Kid... I was shooting for 0.001%
Again apologies to the OP.

 

[bertus]

Hello,

at 100MHz to 1000MHz a dish is the last type of antenna you want. Just use yagi antennas

For a wideband antenna I would use a discone for omni direction or a logaritmic periodic antenna for directional use.

discone antenna - Google Search

log periodic antenna - Google Search

Bertus

 

[hevans1944]

Convolving information with randomness produces a truly random signal that nevertheless contains all the original information. The original information can be recovered by convolving the randomized information with the the exact same "randomness" used in the original convolution. This is known as a "one time pad" cryptography method. It's main disadvantage is the necessity to transport the random encryption set of numbers to the receiver's location for decryption of encrypted information. The convolving operations are usually performed digitally with exclusive-or logic: If A (un-encrypted) information) ex-or B (random numbers) equals C (encrypted message), then C ex-or-B equals A.

Algorithms exist to generate extremely long "pseudo- random sequences" but such sequences are finite and never truly random. Using "noise from outer space" to create random binary numbers is (probably) a truly random process, but hardly necessary for secure communications.

Could the OP please tell us why they need really good random numbers, instead of just "gud enuf" pseudo-random numbers?