Up one level
Back to Home
Can you hear a 3 dB change in power?
By Neil B. Johnson WBØEMU and Roger Grady K9OPO
HTML'd by Mike Morris WA6ILQ
Print this Page

A question posed to the Repeater-Builder email list:

Can you hear a 3 dB difference in transmitter power?
Is 3 dB a significant enough change worthy of making serious changes to produce it?
For us going from 45 watts to 90 watts means a whole new duplexer.

A reply by Neil B. Johnson, WB0EMU:

It is generally accepted industry practice for path reliability projection that 4 dB C/N is the point at which an FM communications receiver delivers 12 dB SINAD and that 7 dB C/N is the point at which such a receiver delivers 20 dB quieting.

Hence, it follows that a 3 dB change in transmitter power [or increase in antenna gain or reduction of path loss], which will result in a corresponding 3 dB improvement of received C/N, would be sufficient to:

A) Take a received signal from 1 dB C/N [unreadable] to 12 dB SINAD [readable with difficulty].


B) Take a received signal from 12 dB SINAD to 20 dB quieting - a very noticeable audible improvement in DAQ [Delivered Audio Quality]

DAQ is a metric that has been used extensively in radio coverage prediction and verification. It is a scale that runs from 1 to 5 with 1 being unusable [speech present but unreadable] to 5, which is perfectly readable with negligible noise.  12 dB SINAD represents a DAQ of 2.0, which is readable with difficulty.  20 dB quieting [about 17 dB SINAD] represents DAQ 3.0, which is commercial quality audio, readable with only a few syllables missing.

OK, so does this mean the noise temperature of a NBFM FM receiver can be approximated by measuring the 12 dB SINAD point & then subtracting 4 dB?

Essentially - yes. The question asks about noise temperature, which is commonly used by our weak signal amateur brethren.  Typically in commercial [two-way]communications we work with Noise Figure.  Noise temperature can be computed from Noise Figure if desired, however, I find Noise Figure to be quite convenient because it is expressed in dB as are receiver sensitivity figures, noise levels, preamp gains and other factors of interest in receiving system calculations.

Given that it is generally accepted that:

N=kTB and that N= -174 dBm/Hz [at room temperature or about 290 degrees Kelvin]

The 'accepted' value for the noise power in your receiver bandwidth will be -133 dBm.

If your receiver has a measured 20 dB quieting of 0.35 microvolts [-116 dBm] and a corresponding 0.25 microvolt 12 dB SINAD [-119 dBm] then it follows that it has a Noise Figure  NF = 10 dB [-119 minus 4 = -123, -133 minus -123 = 10]

Keep this in mind when someone tells you they have a super pre-amp on their repeater receiver and their receiver sensitivity is better than a tenth of a microvolt - they are either using faulty measuring procedures or poorly calibrated instruments because 20 dB quieting would be on the order of 0.11 microvolts if the noise figure was zero.

Roger Grady, K9OPO,  offers one more significant fact on the subject:

The quieting curve of an FM receiver is very non-linear.   3 dB will make a very noticeable difference if the signal is well into the noise, barely any or none at all if it's strong.   The result is that a 3 dB improvement will in most cases approximately double your range.

In closing I would agree 3 dB is not noticeable or barely noticeable if the signal is already well out of the noise. However, in areas where you are trying to push repeater coverage from a given site to the limits, 3 dB is significant on the outer fringes and in the problem areas.   I say this from personal experience; I have seen a 3 dB change in repeater power or antenna gain make a significant and consistent improvement in fringe/weak areas. It can be enough to make the difference between the repeater being usable or not...   I call that significant.

Comment from WA6ILQ:
Note that the above material makes two important assumptions:
1) The repeater receiver is unaffected by the power increase. More output power requires more isolation between transmitter and receiver (see the articles on Isolation, and on horizontal versus vertical separation on repeater antennas).
2) The geography / topography / terrain is having no effect. If you have a repater that is servicing a community in a bowl-shaped valley once the bowl is full (i.e. you have a full quieting signal out to the edges of the bowl) then adding more does no good.

In real life both assumptions can affect your results.

In short, raising the signal to the fringe area user from 7 dB quieting to 10 dB quieting can make the difference from zero copy to usable copy, whereas raising the strong area users signal level from 17 dB to 20 dB is of no benefit.

The sound files in this "Measuring Sensitivity" article may be of interest.

Back to the top of the page
Up one level
Back to Home

Text copyright February 20 2005, Neil B. Johnson WBØEMU and Roger Grady K9OPO
HTML Copyright February 20 2005 Kevin K. Custer W3KKC
All Rights Reserved.
This page last updated 24-Sep-2012.


The information presented in and on these conversion pages is © Copyrighted 1995 and (date of last update) by Kevin Custer W3KKC and multiple originating authors.