e.g.
Airspy NOS - For safety stay below 0dBm (0.2236 Volt RMS)
HackRF One - Below -5 dBm would avoid the risk of permanent damage (This could be increased to 10 dBm with the front-end RX amplifier disabled).
There are two sets of three U.fl RX connections for each of the two RX channels. So would it be the same input power for each RX connection.
It was sent as a private email through crowdsupply during the initial funding, when I asked the question there. But just after I asked the question on crowdsupply this forum was setup so I cross posted the question here and then when I received the email reply I posted the reply here, because I was sure that others would be interested as well.
If you look at the ADC specification on datasheet for the LMS7002M.
The (maximum) input amplitude is specified at 0.8 volts peak to peak.
And 0.8 volts on a 50 ohm input would be 12.5 mW (+11dBm). And if you assume that there is probably in and around ~3dB of attenuation from the “matching network” (what I would call a front end filter) and if you also add in a miniscule safety factor that would produce the +12dBm (0.90 volts; 16 mW).
The maximum (before permanent damage) of +12dBm still looks good to me.
thank you so much. will it be same in the case of limesdr mini?? I am using a limesdr mini v1.2. I wanted to transmit a qpsk signal through wired connection. I don’t know what value of attenuator should i use for the connection.
It would be the same for all the current LimeSDR hardware that use the LMS7002M chip, the matching networks would provide some additional attenuation.
The people building Vector Network Analysers (VNA’s) using a LimeSDR used a minimum of 10dB of attenuation. I’d probably be a bit more conservative and use 20 dB of attenuation because I want to be standing well back from the cliff edge. For me the goal is to be sitting inside the safe zone and not teetering on the edge of disaster. You could always use something like a (JFW industries) rotary attenuator or a step attenuator and fine tune the attenuation.