Thermal Noise Affects OTA TV Signals?

[silverpig]

According to some quick back of the envelope calculations, I figure that 300K thermal noise on a bandwidth of 8 MHz (ATSC channels are 6-8 MHz in bandwidth right?) is about -105.0 dBm. A quick look at tvfool says I have a number of stations with power at my location of -100 dBm to -130 dBm. Of course the gain from an antenna would amplify the signal and not the thermal noise, but there would still be some channels with a S/N near unity.

Going from 300K to 77K (liquid nitrogen) would, if I did my numbers correctly, give you a 6 dBm reduction in thermal noise.

I know it'd be a bit of a pain to set up, but there are some people here who have some pretty insane setups. You can buy nitrogen liquifiers, and can pump the LN2 through the copper tubing of a GH antenna. Condensation/frost would be an issue, but you could probably insulate the antenna itself with some sort of material that is nearly transparent to RF but insulates well enough.

Thoughts?

[keef]

A TV antenna points at the horizon (rather than into outer space). If I'm not mistaken a lot of the thermal noise for terrestrial reception comes from the ground (horizon) itself. I don't think you'd realize the full 6 dB improvement, though I'm not sure how much you would get - half? (the horizon being half ground/half sky).

[silverpig]

The calculations I did considered thermal noise of a 50 Ohm resistor at 300K so that's my starting point. Aside from all other sources of noise, this thermal noise is added on top of your received signal by the antenna itself being hot.

Removing this noise would consist of cooling the antenna and the preamp such that you receive the signal with the cold antenna, boost it with the preamp, then pass it on to warmer components.

[silverpig]

Oh come on... someone here has to want to try this

[300ohm]

(ATSC channels are 6-8 MHz in bandwidth right?)

6Mhz, same as the NTSC channels.

The calculations I did considered thermal noise of a 50 Ohm resistor at 300K so that's my starting point.

What voltage and amperage did you use ?

Oh come on... someone here has to want to try this

Whats the point ? Youre also getting solar and gamma ray radiation from space that is much more noise.

[silverpig]

It's independent of voltage and amperage. This is thermal noise only. It is noise due to the perturbation of charge carriers in a conductor (can be either holes or electrons) by vibrations in the lattice (electron-phonon coupling).

The thermal noise is given by:

P = (Boltzmann's constant)*(Temperature)*(Bandwidth)

To calculate the power in dBm I used:

P_dBm = -174 + 10*Log(Bandwidth) for 300K

To calculate the power in dBm at 77K just notice that the power scales linearly with Temperature and do the log conversion of the 77/300 ratio.

Thermal noise becomes important when you are either measuring extremely weak signals, or when you are averaging over a wide bandwidth.

There may be other benefits to running cold, such as your copper would be a slightly better conductor, making your received signal stronger.

Solar radiation is largely in the THz range (at least for visible light). Gamma radiation is even higher frequency.

[silverpig]

Ah wikipedia has a decent article.

[holl_ands]

Don't forget to follow links in above Wiki article to Shot Noise, et. al.

Thermal Noise would be across 75-ohm input impedance of Preamp's first RF stage.

NTIA Report 02-390 (Fam in Figure 7) depicts Man-Made Noise compared to Thermal Noise (kTb):
http://www.its.bldrdoc.gov/pub/ntia-...390/02-390.pdf

Man-Made Noise was greater than Thermal Noise except for some Rural areas above
300 MHz and Residential areas above 700 MHz....so lowering Thermal Noise won't help
unless you're in the boonies and shut down nearby fluorescent lights, brush-type
motors (e.g. heaters), arc welders, some dimmers, PC's....(incl. tuners & HDTV???)

Galactic Noise (from the "Big Bang") was always lower than Man-Made Noise.

ITU-R P.372-8 Recommendation re Radio Noise depicts negligible Atmospheric Noise below Ch02.
However, that's a long term average result....I don't have any recent Lo-VHF experience,
but many users report glitches in DTV reception due to nearby electrical storms....

Which is why it is advisable to have excess "Fade Margin" over and above "minimal" signal levels.

[silverpig]

Right.

It doesn't change much though as you just add log_10(3/2) IIRC. It might even be log_10(Sqrt(3/2)) which is even less.

[300ohm]

There may be other benefits to running cold, such as your copper would be a slightly better conductor, making your received signal stronger.

Another good reason to use copper or aluminum for the driven elements of an antenna build, they conduct/dissipate heat better than most other electrically conductive metals, except silver.

Also lets not forget the noise generated by the TV set itself, thru the Power and IF and Flyback circuits (CRTS) and not adequately shielded by the tuner.

I know it'd be a bit of a pain to set up, but there are some people here who have some pretty insane setups. You can buy nitrogen liquifiers, and can pump the LN2 through the copper tubing of a GH antenna. Condensation/frost would be an issue, but you could probably insulate the antenna itself with some sort of material that is nearly transparent to RF but insulates well enough.

Hmm, the only practical way I can think of to do this experiment if you want to do it, not me, heh (but its insanely impractical as well, heh) is to imerse the antenna in the same solution (some type of special electically inert freon (or something that starts with an F) that I dont remember the name of. But search thru overclockers.com.au) that insane overclockers imerse their motherboards in to lower the temps of the whole complete system. I dont know the RF properties of that stuff, but you can get pretty darn cold with it. Wear heavy, heavy gloves, heh.

[silverpig]

Cool. This is what I was looking for. There are probably ways to remove environmental noise (actually there definitely are). fluorescents are noisy yes but they operate at 30kHz or something like that and run on 60 Hz AC. When you're looking for a signal at 500 MHz with a 6 MHz bandwidth, it shouldn't be too difficult to throw a few capacitors in line to filter out all the low frequency signals.

[silverpig]

Originally Posted by 300ohm

Another good reason to use copper or aluminum for the driven elements of an antenna build, they conduct/dissipate heat better than most other electrically conductive metals, except silver.

Also lets not forget the noise generated by the TV set itself, thru the Power and IF and Flyback circuits (CRTS) and not adequately shielded by the tuner.


Hmm, the only practical way I can think of to do this experiment if you want to do it, not me, heh (but its insanely impractical as well, heh) is to imerse the antenna in the same solution (some type of special electically inert freon (or something that starts with an F) that I dont remember the name of. But search thru overclockers.com.au) that insane overclockers imerse their motherboards in to lower the temps of the whole complete system. I dont know the RF properties of that stuff, but you can get pretty darn cold with it. Wear heavy, heavy gloves, heh.

Well the idea would be to get a clean low noise signal, then boost it with a cold preamp. You start with a very low S/N, boost it, then you can add tv noise and other stuff down the line, but because your signal is already so high it doesn't matter.

Well I know HOW to do the experiment. I'd probably suggest wrapping the driven elements in some kind of thermal insulation (foam or something) that hopefully wouldn't drop the signal too much. Get a LN2 dewar and hook up a rubber tube from the LN2 output port on the dewar to one of the copper tubes of the active elements. Then run a second rubber tube from the bottom of that element to the bottom of the next element, and one more from the top of the second element back to the dewar. You basically want a series of tubes such that the LN2 can leave the dewar go through one element, through the next, and then back to the dewar.

The evaporating LN2 from cooling the antenna will pressurize the dewar, forcing more LN2 out and flowing through the system. Of course the dewars are equipped with blow-off valves to maintain a safe pressure.

I suppose before returning the nitrogen to the dewar you should build a chamber containing the balun transformer and the preamp to keep those cool as well.

LN2 is cheap. Rubber tube is cheap. A dewar is expensive, but you can probably rent one. I know how to do this all but don't really have the time nor the means at the moment to pull it all off. If only I could do it in the lab :P

[300ohm]

You basically want a series of tubes such that the LN2 can leave the dewar go through one element, through the next, and then back to the dewar.

Yeah, that will do it to, heh.

[holl_ands]

But first you'll need to cool the biggest thermal noise contributor: the Preamp transistor.
And don't forget the feedline between antenna and Preamp....

But by then, you'll figure it's cheaper to subscribe to Cable or Sat....

[silverpig]

Originally Posted by holl_ands

But first you'll need to cool the biggest thermal noise contributor: the Preamp transistor.
And don't forget the feedline between antenna and Preamp....

But by then, you'll figure it's cheaper to subscribe to Cable or Sat....

That's easy enough to do... just have a little box with that stuff in it, flow the LN2 in, flow it out...

All you need in terms of hardware is some rubber tubing, some foam insulation, an insulated box with 2 input and output bibs... When you're building an antenna anyways this stuff will only add to the cost by oh 30%.

The dewar will be expensive. LN2 is cheap.