How important is the noise rating on an LNB, really?

[ancient]

I'm just curious, if you have two LNB's and one claims to have an 0.4 dB noise figure while another claims 0.5 dB, is that difference truly significant or is that just mostly marketing hype?

 

[USDownlink]

For a difference of 0.1dB, it's marketing hype. The manufacturer is going to use super-ideal conditions and cherry pick the noise floor data in order to arrive at their figure. Further, if you're relying on a 0.1dB margin to receive a signal, you're going to be steadily disappointed with the normal turn of events in the process of receiving the signal, like wind and rain and mosquitoes flying in front of the dish.

 

[Titanium]

The joke when ordering from a factory is: Buyer asks, "What is the gain, What is the NR? What is the phase?" The factory replies, "What do you want us to print?"

Only sorted DRO LNBFs are 0.4NR or 0.5NR. These will typically cost a few dollars more because the LNBFs are graded and only the units meeting spec are accepted by the customer. If the design is "off the shelf", the sorted units not meeting 1st customer specs are sold to another customer. Proprietary design LNBFs are either rebuilt to meet spec or discarded. In all reality, most KU band LNBFs that sold by generic sellers are graded 0.8dB NR or higher. Only a few LNBFs actually have reached a 0.3dB NR. There are actually more important specifications to consider when testing a LNB, but the specs rarely get the spotlight because they are not sexy marketing keywords. The current trend to print 0.1dB or 0.2dB NR is pure BS and nothing but a marketing untruth.

Only a few distributors print actual specifications on their products. Most print Creative Marketing (BS/Lies) of whatever will sell the product.

 

[USDownlink]

The joke when ordering from a factory is: Buyer asks, "What is the gain, What is the NR? What is the phase?" The factory replies, "What do you want us to print?"

You've taken a slightly more jaded view than I, but I don't doubt you for a moment. When there is still a residual connection between the company that designed the product and the retailer and/or you have a high-quality customer, my experience suggests there is at least a minimal attempt to pull some massaged engineering data that says "yeah, look, something has been measured and here is how we rationalize it". But when everyone in between knows the end customer won't or can't check, anything goes, I guess.

Only sorted DRO LNBFs are 0.4NR or 0.5NR. These will typically cost a few dollars more because the LNBFs are graded and only the units meeting spec are accepted by the customer.

What is the empirical wisdom as to what happens to this after installation? A few dollars more, because the factory spent time on the grading, but what is that really worth after a few day-night cycles? After a month or a year?

 

[Titanium]

A graded product for me means that the product does not only meet, but exceeds all specifications. The electronic industry norm of 1% failure being acceptable...well, not in my world! I expect less than 1/10th of a percent failure and usually experience less than 1/50th of a percent after 3 years on my past LNBF projects. I cut my teeth on equipment and service contracts where a single failed LNBF translates into a $300-500 minimum trip charge fee for a tech to roll under warranty. One distributor had numerous LNBF problems with initial failure rate on a few models exceeding 10%. While a customer may experience an inconvenience of a single failed product, installers and resellers remember the percentages, unhappy customers, lost time and $$$.

A day, week, months, years, a LNBF should have the similar performance for years... Living with a product that was crap before you installed it makes a big difference.

 

[T134]

I've got to fully agree with Brian on nearly every point. A number of years ago I got to talk to one of the engineers that was doing a "0.1dB" LNBF and mentioned that some people thought it was BS. The answer I got back was basically that in a lab under controlled circumstances and by whatever standards you could come up with these results.

I can also make a 747 weigh in at 100 grams......if tested in orbit! LOL

0.1 is all marketing. In Europe it is common place to label everything 0.1dB. Even our Amiko's have this. Unlike some other places it is not in the specifications, we release the real numbers. For example our PLL twin I think is a realistic 0.7dB NF. Strangely enough it performs as good or better than certain brands here with 0.1dB in their specs.

Even if you have real specifications on a product there may not be a real world performance difference between say a 0.6dB NF and 0.7dB. Other numbers can be more important such as stability and noise floor.

The best way to judge a LNBF isn't by any number given by the people trying to sell it but by putting it on your dish and running it awhile. When I test LNBFs here I do "real world" stuff that almost anyone can replicate. The most important things to me and I think most guys is do you get higher performance with XXX LNBF? Is it consistent and is it going to work a year down the road?

Also agree with Brian, 1% defect is totally unacceptable. In any mass manufactured product eventually you are going to run into a dud. The dud rate on a good product should be more like 0.1% or less. Seems like for many 5%, 10% and even higher is the norm and acceptable. I'd freak out if I saw even 1% issues with my products!

 

[ancient]

Thank you for confirming what I suspected all along.