How to get the noise figure of an RF Switch?

[maolh]

Among the specification parameters of RF swithes from varous manufacture, there is no noise parameters. Even in simulation tools like Genesys and ADS, the noise performance is not specified for RF switches. I just wonder if the noise isn't important for RF switch?

In my project, transmitter and receiver share antenna through a RF switch. The transceiver works in half-duplex mode. RF input can be as low as -150dBm. When I do the system simulation, should I consider the noise form RF switch? If yes, how can I get the noise figure of RF switch? Thanks.

 

[biff44]

Re: Noise of RF Switch

Normally noise is not an issue. However, if you assumed that was always the case you could get in trouble quickly. Sometimes "smart" companies put switching regulators into their switches (in case the right voltage or polarity is not available). Depending on your application, that could be a big noise source.

Sometimes insufficient reverse bias is applied, so harmonics or 3rd order products are generated.

Sometimes power lines are not adequately filtered, and you end up with a lot of phase noise.

 

[maolh]

Noise of RF Switch

biff44,

First, thank you very much for your quick reply. To be honest, I don't know why the noise is not an issue. If I guess, it's because the on-resistance is very low, thermal noise will be very small. Am I right?

 

[biff44]

Re: Noise of RF Switch

Well, from a system design point of view, no. If I had a system that needed to switch between two antennas, for instance, and noise figure was important, I would put a low noise amplifier on each antenna, and put the SP2T switch after the gain of the LNA's. I might even just turn the DC bias of the amps on/off to do the switching if I was being cheap.

 

[VSWR]

Re: Noise of RF Switch

The noise figure of an RF switch can be simplified to being the insertion loss. So if the insertion loss of a switch is 0.5 dB, the total system noise figure will increase by the same amount of 0.5 dB, when the switch is located in front of the receiver.

When I do the system simulation, should I consider the noise form RF switch? If yes, how can I get the noise figure of RF switch? Thanks.

 

[maolh]

Re: Noise of RF Switch

The noise figure of an RF switch can be simplified to being the insertion loss. So if the insertion loss of a switch is 0.5 dB, the total system noise figure will increase by the same amount of 0.5 dB, when the switch is located in front of the receiver.

I think for passive components, we can say the insertion loss equals to the noise figure. However, RF switch is electronic switch, it's active device. I'd like to know how much noise it will generate as an active device instead of how badly it will decrease the noise performance as a linear two port network. By the way, I want to use AS193-73 from Skyworks. If you want to see how they specify the switch's performance, you can find it at **broken link removed**.

Thanks.

 

[kayaoo]

Re: Noise of RF Switch

I'm considering to build a cheap source tune kit with Gaas RF switch for measuring the optimimal noise impedance of a LNA. When I search infomation about the RF switch noise, I found this really old topic. Is there any conclusion now?

 

[davenn]

Re: Noise of RF Switch

I think for passive components, we can say the insertion loss equals to the noise figure. However, RF switch is electronic switch, it's active device. I'd like to know how much noise it will generate as an active device instead of how badly it will decrease the noise performance as a linear two port network. By the way, I want to use AS193-73 from Skyworks. If you want to see how they specify the switch's performance, you can find it at **broken link removed**.
Thanks.

I suspect for this device you can still take the insertion loss as the device's noise figure, else they would stipulate a given noise figure. Technically the device is still just a passive device that is... its just a switch its NOT an amplifier
(other than a bit of insertion loss it isn't altering the RF signal passing through it

Dave
VK2TDN

 

[volker_muehlhaus]

Re: Noise of RF Switch

I suspect for this device you can still take the insertion loss as the device's noise figure

Yes, exactly. That is valid because the switch is a "passive" component with regards to the signal path.

 

[kayaoo]

Re: Noise of RF Switch

Thanks All. I simulated a GAAS and a Pin Diode SPDT in @DS. Additional NF degradation should be less than 0.00x dB for both cases.

 

[mehmet89]

Re: Noise of RF Switch

Hi.. I need documentation related noise of RF switch with an AWR simulation. I am looking but I dont find anything. If somebody has that please help me...