frontend.cpp & Si2168/9 Tuners

[EMJB]

There seem to be problems with the signal quality, and SNR values reported by frontend.cpp for the Si2168/9 tuners that is causing the ABM frequency finder to operate incorrectly, and the Signal Finder plug-in to display wrong values. In particular:

readFrontendData(iFrontendInformation.signalQuality) always seem to return 0xFFFF for the Si2169, and zero for the Si2168

readFrontendData(iFrontendInformation.snrValue) returns a sensible value for the Si2169, but not the Si2168

readFrontendData(iFrontendInformation.signalQualitydB) returns a sensible value for the Si2168

A useful stepforward would be a clear definition of what is actually returned - for example the Last seems to return 100 x SNR in dBs.

I will try to clarify the full scope of these issues over the next 1-2 weeks.

[EMJB]

The following log entries indicate the returned values for the 4 data types for two muxes for the two tuner types:
Si2168:
<00-40-24.520> [ABM-FrequencyFinder][EBx] readFrontendData(iFrontendInformation...) for channel 29 and tuner type Tuner D: Silicon Labs Si2168 (DVB-T2)
<00-40-24.520> [ABM-FrequencyFinder][EBx] signalQuality: 65535
<00-40-24.521> [ABM-FrequencyFinder][EBx] snrValue: 0
<00-40-24.521> [ABM-FrequencyFinder][EBx] signalQualitydB: 3575
<00-40-24.522> [ABM-FrequencyFinder][EBx] signalPower: 0
<00-41-12.442> [ABM-FrequencyFinder][EBx] readFrontendData(iFrontendInformation...) for channel 31 and tuner type Tuner D: Silicon Labs Si2168 (DVB-T2)
<00-41-12.443> [ABM-FrequencyFinder][EBx] signalQuality: 65535
<00-41-12.443> [ABM-FrequencyFinder][EBx] snrValue: 0
<00-41-12.444> [ABM-FrequencyFinder][EBx] signalQualitydB: 3300
<00-41-12.444> [ABM-FrequencyFinder][EBx] signalPower: 0

Si2169:
<00-55-19.996> [ABM-FrequencyFinder][EBx] readFrontendData(iFrontendInformation...) for channel 29 and tuner type Tuner C: Si2169 (DVB-T2)
<00-55-19.996> [ABM-FrequencyFinder][EBx] signalQuality: 65535
<00-55-19.997> [ABM-FrequencyFinder][EBx] snrValue: 36044
<00-55-19.997> [ABM-FrequencyFinder][EBx] signalQualitydB: 36044
<00-55-19.998> [ABM-FrequencyFinder][EBx] signalPower: 255
<00-55-32.831> [ABM-FrequencyFinder][EBx] readFrontendData(iFrontendInformation...) for channel 31 and tuner type Tuner C: Si2169 (DVB-T2)
<00-55-32.831> [ABM-FrequencyFinder][EBx] signalQuality: 65535
<00-55-32.832> [ABM-FrequencyFinder][EBx] snrValue: 34734
<00-55-32.832> [ABM-FrequencyFinder][EBx] signalQualitydB: 34734
<00-55-32.833> [ABM-FrequencyFinder][EBx] signalPower: 255

These measurements were done with a high-gain aerial near the towards the edge of the transmitter coverage, so SNR values in the 30s dB seems reasonable.

Hope these clarify the issues.

EMJB

[EMJB]

Oops - sorry I closed this by mistake.

[EMJB]

Expanding on my concerns about lack of definition of what frontend.cpp is supposed to return, I have the following questions/comments:

(1) Do the returned values of 0 and 0xFFFF have a defined special significance? The results posted above seem to indicate that they are both used to indicate absence of data.

(2) Huevos has stated that a "signalQuality" of 0xFFFF indicates a "maxed-out" tuner (which I believe is an indication that it is at or near saturation so in danger of loss of lock etc in the event of further increases in signal) . Is this correct?

(3) What exactly does "signalQuality" provide a measure of? I have seen the term refer to the quality of the decoded data stream, i.e. the opposite of Bit Error Rate. that measure provides a very poor basis for signal comparisons as a few dB in signal strength takes you from awful to perfect, and thereafter signal increases have no effect.

(4) What does "snrValue" mean? A linear version of SNR rather than in dB? What is the scaling?

(5) What does signalQualitydB provide a measure of? SNR in dB perhaps? What is the scaling?

(6) Are "signalPower" measurement values the same for all tuners, or just an arbitrary indication specific to each tuner. Does the value of 255 (0xFF) have any specific meaning such as no data available?

EMJB

[AbuBaniaz]

The SifTeam is defunct