View Single Post
Old 3-Nov-2017, 8:17 PM   #1
rabbit73
Retired A/V Tech
 
Join Date: Aug 2012
Location: S.E. VA
Posts: 2,747
Weak Signal SNR Measurements

Background:

When giving weak signal reception advice, I have been saying:

Quote:
Originally Posted by rabbit73
It is not possible to receive a signal with a NM below -15 dB. If you are able to receive a signal with a NM below -15 dB, either the TVFool report is wrong, or the signal has been enhanced by Tropospheric Propagation.
A forum member questioned my statement, saying that it IS possible to receive signals with a NM less than -15 dB if you use an antenna with more gain. At that time, I had no way to prove it either way, other than to say a signal that weak would be buried below the thermal noise floor at -106 dBm for a 6 MHz bandwidth DTV signal.



Since that time, I have been able to simulate the problem, and prove to myself that it is NOT possible to receive and decode DTV signals below -15 dB NM.

Weak Signal SNR Measurements, Test 1



The purpose of this test is to determine if additional antenna gain can make it possible to receive and decode 8VSB OTA signals with a NM below -15 dB.

Additional antenna gain would be indicated by a signal from the antenna that was more than 15 dB above the noise floor from the antenna. I will call this ratio C/N (carrier to noise) to differentiate it from the SNR for the TV.

Equipment Used:

Sony KDL32BX320
Sadelco DisplayMax 800 Signal Level Meter (SLM), used in Analog and Digital modes
CM7777HD Preamp #1, High Gain 30 dB setting to increase sensitivity of SLM
RCA TVPRAMP1R for tuner in Part B
CM7777HD #2, Low Gain 17 dB setting for tuner in Part C
20 dB 1% Fixed Attenuator
3 dB 1% Fixed Attenuator
Wavetek 7510A 0-10 dB Variable Attenuator

Setup: for Test 1, Part A:

Antenna > variable attenuator > splitter > tuner and SLM, with preamp #1 before the SLM; no preamp before the tuner

A strong local signal on CH33 was used, because it was fairly stable and had a good SNR before attenuation to make it a weak test signal . CH32 was vacant and was used for the noise floor measurement. The signal level meter (SLM) was set to make a single frequency measurement near center channel in the analog mode. (Digital TV signals, 8VSB-AM, are actually analog signals that carry digital information.) In that mode, it is sensitive down to -35 dBmV (-84 dBm), and the measurement bandwidth is 280 kHz.

I added a 30 dB preamp to increase the meter sensitivity to below -106 dBm. The gain of the preamp was verified with the signal level meter and a stable signal from a Blonder Tongue modulator. The linearity of the SLM was checked with fixed attenuators. The SLM has an internal noise standard, traceable to NIST, which the meter uses to check calibration at fixed intervals.

The noise floor of the measurement system was then -112 dBm.

Code:
Weak Signal SNR Measurements, Test 1, Part A:

Attn    Sony    SLM    Conv#1   Correct  CH 32   Conv#2      C/N
 dB   SS  SNR  CH 33 > to dBm > to DTV   Noise > to dBm       dB
          dB    dBmV              dBm     dBmV

 0    76   28   +14    -65 dBm            -29     -108       43 dB
10    56   25    +4    -75                -33     -112       37 dB
20    32   19    -6    -85
21    32   19    -6    -85                -33     -112       27 dB 
22    32   18    -7    -86                -33     -112       26 dB   
23    30   18    -8    -87                -33     -112       25 dB
24    28   17    -9    -88                -33     -112       24 dB
25    28   16   -10    -89        -84     -33     -112       23 dB

26    26   15   -11    -90        -85     -33     -112       22 dB

27    24   15   -12    -91        -86     -33     -112       21 dB
28    22   14E  -13    -92        -87     -33     -112       20 dB
29    20   13P  -14    -93        -88     -33     -112       19 dB
30    18    0F  -15    -94                -33     -112       18 dB

SS is Signal Strength
E is uncorrected errors
P is pixelation
F is picture freeze 

Conv#1  conversion to dBm factor is -79 dB
Correct to DTV dBm factor is +5 dB
Conv#2  conversion to dBm factor is also -79 dB
Conv#1 factor is -79 dB; 30 dB to subtract the preamp gain, and 49 dB for dBmV to dBm conversion.

Correct to DTV factor is +5 dB. According to the designer of the meter, Rob Bredin, it adds a nominal 6.8 dB for the digital single channel scan, but not for the single frequency analog measurement used above. The actual value depends upon the shape of the signal. The analog measurement for channel 33 was -11 dBmV and -6 dBmV for digital, a difference of 5 dB.

Conv#2 factor is also -79 dB to convert from dBmV to dBm for the measuring system noise floor.

C/N is the ratio of the signal from the antenna to the measurement system noise floor to differentiate it from the TV SNR.

SLMs are calibrated in dBmV and used by antenna installers and cable techs. It was necessary to convert to dBm to match a TVFool signal report

With the attenuator at 26 dB, the SNR was 15 dB for a -85 dBm signal, just where you would expect it to be with a tuner NF of 6 dB. A signal power of -85 dBm is 6 dB above the minimum required signal strength of -91 dBm, which is 15 dB above the -106 dBm noise floor for a 6 MHz bandwidth DTV signal.

Part B: I added an RCA TVPRAMP1R preamp before the tuner.

Attenuator at 0 dB:

SS MAX
SNR 26 dB, 2 dB lower than without preamp because of added preamp NF. The tuner NF is traded for the lower preamp NF and the tuner NF is buried in the amplified noise floor, making it irrelevant.

Attenuator at 26 dB:

SS 66, to offset distribution losses
SNR 15 dB at -86 dBm, 5 dB above the -91 dBm minimum required signal; a 1 dB increase in weak signal sensitivity over no preamp

Part C: I substituted the CM7777#2 for the RCA preamp

Attenuator at 31 dB:

SNR 15 dB at -87 dBm, 4 dB above the -91 dBm minimum required signal; a 2 dB increase in weak signal sensitivity over no preamp. A preamp with a lower NF would increase the sensitivity even more.

CONCLUSION:

It is not possible to receive and decode 8VSB OTA signals with an accurate listed NM below -15 dB at -91 dBm. Even though a narrowband measurement shows sufficient C/N for a SNR of the signal to noise floor at the antenna output, the noise floor for a signal with a bandwidth of 6 MHZ is -106 dBm. Even without a signal present, the intrinsic noise power in a 6 MHz bandwidth is -106 dBm. Any digital TV signal that is weaker than -91 dBm, will have an SNR less than 15 dB, and can not be decoded.

As you can see, there was additional SNR available at the antenna that could not be utilized by the tuner because of the fixed noise floor of -106 dBm for a 6 MHz bandwidth signal.

Code:


 Bandwidth     Thermal      Notes
             Noise Power

 1 Hz         -174 dBm
 280 kHz      -119.5 dBm    My SLM in analog mode*
 6 MHz        -106 dBm      Analog or digital TV channel

 * Does not include Noise Figure of preamp used 
   to increase sensitivity of meter
Attached Images
File Type: jpg DisplayMaxImgCH3A.jpg (144.0 KB, 3379 views)
File Type: png NMdiagDXrev2.png (55.1 KB, 3502 views)
__________________
If you can not measure it, you can not improve it.
Lord Kelvin, 1883
http://www.megalithia.com/elect/aeri...ttpoorman.html

Last edited by rabbit73; 16-Nov-2017 at 3:23 PM.
rabbit73 is offline   Reply With Quote