ouagadougou said:
A Noise Figure of 0.8dB is still a realistic minimum at ku frequencies. The gain of successive transistor devices has improved over the years so that when you cascade them in an LNB they give a better overall noise performance and gain than the older lower gain devices.
Do you know exactly how they measure the noise? I'm curious because SMW do also make those X-Line (old Wideband type) LNBs, and for many years* they were quoted as having a noise figure of 0.6 dB (actually those are the ones I said appeared in 1995) whilst the newer WDL ones, and also the Quattro (= OMT + 2 WDLs), were only rated at 0.8 dB and when I asked SMW about this, they told me that they were in fact all of them (X-Line, WDL/Quattro) using the same transistors, it was just that the design of the WDL/Quattro was different or sth (he didn't really give me a good answer that day, herr M**** Kj*******
).
* To my surprise their noise figure is now only quoted as typical 0.8 dB, whilst the WDL NF is still quoted as 0.8 dB as before, but the Quattro has gone up from 0.8 dB to 0.9 dB! So have their LNBs suddenly become poorer now (poorer manufacturing), or have they changed the way they measure the NF (yet again, to their disadvantage?) or what?
ouagadougou said:
If you have the equipment you can quite easily measure the noise output increase of your LNBF when you put your hand over the feedhorn becuase the LNBF noise temperature is lower than the physical temp of your hand.
I wish I had the equipment, I would be bery keen to conduct this sort of experiment.
Channel Hopper said:
The idea of dB in the signal path is to simplfy the equation of total losses by addition.
Once you remove the average 0.6dB from the rest of the losses (atmospheric noise, temperature, dish, cable, connectors and tuner), you find there isn't that much improvement anyway.
Well, according to the formula I use (and ouagadougou too, as I can see from his last post), if you DO remove the noise (T = 0), then C/N suddenly becomes quite large ... :-hammer
(I do stand open for correction though, maybe I'm using an oversimplified formula.)
Spiney, do you know how much noise (in Kelvin) we could be talking about with one of those masers? Do you think it could ever be made available to us consumers? (Perhaps this is what the MOD or DOD use?)
An LNB with a genuine noise figure of just 0.1 (not 0.01) would really make a world of difference. For me and the Invacom I'm currently using it would be like quintupling (yes, that is 5 times) the area of my dish (more than doubling the diameter).