Post Go back to editing

ADG936 input DC offset

I am planning to use the ADG936 switches for application where input DC offset is 0.8V. The input signal frequency will be less than 400MHz and the AC swing on each input (each RF1X, RF2X input) can be as high as 0.8Vpp (or +- 0.4V). So each input can go as high as 1.2V peak (0.8V DC offset + 0.4V peak AC input). The Switch should be able to pass this DC offset and AC signal to the output.

According to the datasheet it looks encouraging for 0.5V DC offset on inputs, where input can be as high as 16dBm. How about 0.8V DC offset, with 2.5V VCC and input conditions that i mentioned above ? Will there be any bandwidth reduction, lower input power limits or any other issues ? According to AN-952, the ON-Resistance increases with Source-Voltage. So with 1.2V maximum source-voltage (per  in my case, the ON-Resistance is around 9 Ohms.

Please let me know of any issues possible with my application.



  • Hi AB,

    As you noted, the trade off here is with Ron and Vbias, the higher the DC bias the higher Ron goes. Figure 1 in AN-952 shows a typical plot of this relationship.  In your case, using 0.8V offset will result in a higher Ron as you noted.  The switch will work but it will have slightly higher loss.  Since your peak voltage is less than your Vdd supply voltage and you have no negative going voltage cycle I think you should not see signal clipping.  But you should try and look at the insertion loss of your signal as this will have increased.   I can't say by how much as weonly when up to a 0.5Vbias for our insertion loss datasheet measurements.  As long as the Ron doesn't increase significantly then it shouldn't be too much of an increase.  With this in mind, I would recommend using a Vdd as close to 2.75V as you can.  This keeps the Ron Vbias curve flatter for longer, e.g. at 1.65Vdd I have seen Ron to go high,approximately 25ohms (with 0.8Vbias).  A 2.75Vdd would be the best case to give you as low an Ron as possible at 1.2V peak signal.