PIN switch problem as S12 and S21 by VNA are the same

[trail]

PIN switch problem?

I build a switch using PIN diode(Philips BA277) , but when I measure its S12 and S21 by VNA, both of them are same. Furthermore, when added the swithcing voltage, S21 just changes 2.5 dB. I don't know what the resons are, please help me! Thanks!

 

[KMPA]

Re: PIN switch problem?

Please upload your circuit schematics, untill I look at it.

 

[lguancho]

Re: PIN switch problem?

Hi, trail. A PUN Diode switch is just a current controlled resistor. A simple design can be a shunt diode (one end to the ground) connected to a TX line. When forward biased, the switch acts as a short circuit, reflecting any RF signals entering either ends of the TX line.

It seems that you have poor isolation ( as you say you apply the switching voltage).

Make sure you check the whether the diode switch is turn on or off properly. Use a multimeter to check the voltage drop across the diodes. If forward biased, it should register some value (anything from 0.7V to 1.+ volts, depends on components).

If you can, show us a picture of your board, plus a schematic. Then we should be able to pinpoint the problem better.

 

[trail]

Re: PIN switch problem?

The attachment is my schematic , please tell me where are wrong.Thanks!

 

[lguancho]

Re: PIN switch problem?

I saw your schematic. This is a very common switch design for TDD system. D1 and D2 diodes are always oppositely biased to switch either P2 or P3 to the common port P1 (The control voltages V1 and V2 are always complementary, i.e V1 = High, V2 = 0V or vice versa). Insertion loss is affected by the biasing resistor setting, and the DC blocking resistor. Isolation performance is affected by the RF chokes and the biasing resistor.


I have a few questions about it:

[1] What is your operating frequency ? The BA277 diode is good for VHF band (30 -300MHz) as indicated in the datasheet. But your dc blocking caps in the schematic is 1pF, this maybe too low.

[3] Your biasing resistors are 1 ohm, how is this determined? For a low forward resistance (If=10mA and Vf=1V), assuming you are using a control voltage of 3V, the biasing resistor should be (3-1)/10mA, i.e. 200 ohm. This value will also affect the isolation performance of your switch. Could you try to use 200 ohm fixed biasing resistors, and make your RF chokes (1nH) a optimizable variable (say between 1nH and 15nH)? You can set a goal of say S21= -1.5dB (within your operating frequency) and S31 of say -20dB . Remember to set V1 and V2 correctly during your simulation. The common RF ground return choke is also an important factor.


Lastly, this kind of TDD switch design are widely documented. You can search for Infineon or Philips website for relevant application notes.

 

[E-design]

Re: PIN switch problem?

All the "1" values appear to be the default values by the program. This needs to be changed to the correct values for successful simulation.

 

[trail]

PIN switch problem?

Yes, I know the reason of the problem. BA277 just works on VHF(30MHz~300MHz). Between VHF, the diode behaves very well. However, I use it in ISM(2.4GHz), so the problem happens.