# What happens with LNA in bidirectional communication? [closed]

I want to ask that if you have a cable that has started with 40dbm source signal and has reached -10dbm then you put LNA (the end of the cable will be plugged to RFIN), then attach the RFOUT to another long cable what happens with:

1. Even though the LNA has 27dbm RFIN limit, I am afraid that even that the signal is attenuated down to -10dbm, the cable is low ohm (resistant) and the remaining wattage should be like 5 watts getting inside RFIN. Wouldn't that blow up the LNA ? If no, will the LNA then bypass these 5 watts to continue along the line (makes sense otherwise they have to dissipate in the LNA, which means it will blow).

2. The LNA is rated 10db gain at the needed frequency and at the same time specifies maximum RF output of 20mW (13dbm). If the output is indeed maximum 20mW, where did the 5000mW (5 watts) in go ?

3. The LNA amplifies one direction only RFIN to RFOUT. If small signal is sent from the distant end from the cable (as in bidirectional communication) it will reach the RFOUT to of the LNA then it has to go trough the RFIN and reach the destination. Now these bidirectional communications happen at < 1us , which is terribly fast to turn of TX & RX. As the LNA is DC powered is it prepared to bypass the signal from RFOUT to RFIN and continue along the line or/and at what cost ?

Thank you!

• This bit doesn't make sense: the signal is attenuated down to -10dbm, the cable is low ohm (resistant) and the remaining wattage should be like 5 watts - 5 watts is 37 dBm, but -10 dBm is 100 microwatts. Which one is it? Perhaps edit the question to stick to one set of units, just dBm and dB? Jun 17, 2023 at 20:14
• @tomnexus Please read this as I am in puzzle exactly for this as well ham.stackexchange.com/questions/21961/… (and the comments more importantly) I trust the ohms law that if you put 10Wof power it will be reduced by the specified cable resistance. Let's say that after 100 meters it is down to 5watts. As I understand this is applicable as Radio Signals are Direct Current at given frequency. Now the puzzling part is that the cable chart for attenuation is eg 50dB for 100m, which kills the signal before ohmLaw Jun 17, 2023 at 21:13
• Ohms law is for resistors, volts and amps, but here we're using a different set of abstractions. The figures you are quoting are Radio Frequency power, not really related to DC power. Power is expressed either in dBm or in watts: 27 dBm is 0.5 W. -10 dBm is 100 uW. Nothing more to it. Jun 17, 2023 at 22:47
• On your third comment, there are many good answers about the difference between dB and dBm, look at these: 1 or 2 to start. Jun 17, 2023 at 22:49
• Coax isn't a regular wire, it is a transmission line. It has not just resistance, but also capacitance and inductance. Regular ohms law doesn't take that into account, you need a more complex equation that includes not just these but also the frequency of the signal and impedance of the source and load. Jun 20, 2023 at 0:49

## 1 Answer

1 and 2: we've addressed this before. No 5 watts gets anywhere near the LNA's input.

3: No. It won't work at all. The LNA might or might not be damaged by the power coming in on its output, but either way it won't pass that power back to the input side. It's strictly a one-way device. I think I have a vague idea of how you could make a version of this that would actually work bidirectionally by using circulators, but it's more difficult than anything I would want to try (I'm not very experienced with microwave) and I think it would be quite expensive to get enough isolation.

• 2 way splitter with some isolation will work ? Like how much isolation is needed ? Jul 2, 2023 at 17:18
• @Svetoslav No, a 2-way splitter won't work, you need circulators. And the isolation needed is whatever it would take to bring the TX power plus the LNA gain down below the RX damage threshold, so something like 50dB. But circulators above 20-25dB isolation are rare and expensive. 50dB ones are in the "call for quote" range but they're at least several hundred dollars. Jul 2, 2023 at 18:05
• Ok I want to ask. You say that the output of LNA is unable to bypass the signal to input. If so what happens with the signal ? Is it reflected back to the source as oscillation with the same power, minus the cable loss ? I need to know what happens. Thanks Jul 6, 2023 at 18:22
• @Svetoslav I think most likely it would be absorbed and dissipated in the LNA's output transistor, but it could also be reflected. Hard to say without detailed knowledge of the specific LNA. Jul 6, 2023 at 18:40