Page 1 of 1

typical digital rx sensitivity

Posted: Sun May 01, 2016 6:15 pm
by G4WIM
I wonder what typical sensitivities people are getting on 24cm for the various data rates and FEC ?

At 4Mbps and an FEC of 3/4 I get a signal threshold of -100dBm

Not sure if this is good / bad or average.

This is after going thru' a duplexer with about 0.7dB of loss into a PGA103 LNA (about 1dF NF and 13dB gain) followed by a 7 pole interdigital with a loss of 2dB into a Comag SL30/12.

Not knowing the NF and bandwidth of the comag makes it hard to calculate the overall system performance.

73 Tim

Re: typical digital rx sensitivity

Posted: Sun May 01, 2016 8:26 pm
by g8gtz
Hi Tim,

That sounds about right - I do all the coverage plots for -95dBm and have seen up to -100 dBm for 4 Ms on rxrs I've tested.

In practice you will find you will need more gain to drive a sat rx - a typical LNB has 55 dB end to end gain so most of us use a 20 dB sat line amp on the input to the receiver.

73

Noel - G8GTZ

Re: typical digital rx sensitivity

Posted: Mon May 02, 2016 12:22 pm
by F6DZP
It is not easy to give an exact answer:
all NIMs we can find inside SetTopBox or used with a MiniTiouner have an already built in LNA before the zero IF tuner.
gain and noise factor are different for each kind of NIM ( Eardatek NIM uses a BFP450 Gain 15.5dB NF=1.7 dB, LG NIM uses BFP420 gain 21 dB NF=1.1 dB, Sharp uses a 2SC5751....)
so we measure the sensitivity of the whole structure inside the NIM: LNA+ zeroIF + demod.

TT S2-1600 has not LNA between the F plug input and the zero IF.

(So as the level value given by Tutioune/Minitioune is taken from the AGC value given by the DVB demod, you will see a difference of value between the 2 software)

Jean-Pierre F6DZP

Re: typical digital rx sensitivity

Posted: Tue May 03, 2016 12:05 pm
by G4WIM
Hi Guys,

thanks for the replies.

Looks as though I'm in the right ballpark and making some best case assumptions for noise figure, BW and qpsk demodulation threshold for a given data rate leads me to conclude that the very best theoretical sensitivity in my case would be about -104dBm.

I'm currently measuring -100.8dBm on a repeatable basis.

So taking into account connector, duplexer and cable losses I'm probably only about 2dB off and part of that could be down to my LNA NF.

I do find that for a given signal level the Comag SL30/12 captures the signal about 1-2dB before the TT3200 I also run. But both drop out at similar levels.

I'm not trying to measure absolute RF signal levels with the rx - just wanting to make sure the rx threshold is where it should be.

73 Tim