Rx sensitivity measurements for DVB-S2 and DVB-T
Posted: Wed May 19, 2021 7:10 pm
Following my attempts to work Rob M0DTS I thought I'd make some rx system measurements for DVB_S2 and DVB-T on 437 and 1315MHz,
I was using 250kS and FEC 1/2 for both DVB-T and DVB-S2 with a LIme as the source at a level of -31dBm output followed by precision switched attenuators giving me confidence in levels down to at least -127dBm. Running the Lime ast a low level helps signal leakage issues.
Both my 437 and 1315Mhz have high performance helical or interdigital filters directly on the antenna inputs - adding about 0.3dB extra loss.
The 23cm Rx uses a gasfet LNA whereas the 437MHz rx uses a PGA103. The gasfet measures a NF of just under 1dB whereas the PGA103 measures just over 1.1dB.
Taking into account the rest of the rx chains both systems have a noise figure close to 2dB - not brilliant but adequate.
I found the got the same results sensistivity wise on both bands - as might be expected as the noise figures are similar
On both bands I could reliably decode both DVB-S2 and DVB-T down to -117dBm.
DVB-S2 took about about 0.5dB more to intially capture whereas DVB-T took 2dB more (and about 30 seconds) - but once locked both rx's would work down to -117dBm and hold the signals without dropping.
Looking at the signals on SDR Sharp at -117dBm they were about 6dB above noise floor.
I was wondering if anyone else had tried measuring their rx thresholds and how they compare to mine ?
Using Knucker for DBT-T and Winterhill for DVB-S2
FYI I suspect the reason I couldn't rx Rob M0DTS was due to a very strong signal out of band on 443MHz blocking my 437MHz rx - the antenna filter has significantly reduced this blocker hopefully down to a level where it's no longer a problem, in band blockers are something else !!
73 Tim
I was using 250kS and FEC 1/2 for both DVB-T and DVB-S2 with a LIme as the source at a level of -31dBm output followed by precision switched attenuators giving me confidence in levels down to at least -127dBm. Running the Lime ast a low level helps signal leakage issues.
Both my 437 and 1315Mhz have high performance helical or interdigital filters directly on the antenna inputs - adding about 0.3dB extra loss.
The 23cm Rx uses a gasfet LNA whereas the 437MHz rx uses a PGA103. The gasfet measures a NF of just under 1dB whereas the PGA103 measures just over 1.1dB.
Taking into account the rest of the rx chains both systems have a noise figure close to 2dB - not brilliant but adequate.
I found the got the same results sensistivity wise on both bands - as might be expected as the noise figures are similar
On both bands I could reliably decode both DVB-S2 and DVB-T down to -117dBm.
DVB-S2 took about about 0.5dB more to intially capture whereas DVB-T took 2dB more (and about 30 seconds) - but once locked both rx's would work down to -117dBm and hold the signals without dropping.
Looking at the signals on SDR Sharp at -117dBm they were about 6dB above noise floor.
I was wondering if anyone else had tried measuring their rx thresholds and how they compare to mine ?
Using Knucker for DBT-T and Winterhill for DVB-S2
FYI I suspect the reason I couldn't rx Rob M0DTS was due to a very strong signal out of band on 443MHz blocking my 437MHz rx - the antenna filter has significantly reduced this blocker hopefully down to a level where it's no longer a problem, in band blockers are something else !!
73 Tim