HMC849ALP4CE problem

Discussion created by Valimor on Oct 14, 2016
Latest reply on Oct 20, 2016 by Valimor

Hello, in my project using the microwave switch HMC849ALP4CE and everything works fine, but there is a problem, after some time, the chip stops working normally. This is expressed by the following features:
1. The measured DC voltage outputs at the RFC, RF1, RF2 to bypass capacitors is equal 2,6V (measured by a multimeter), although earlier it was equal to the voltage 3.3V power supply;
2. If we measure the resistance between the RFC and RF1, or RFC and RF2 it is equal to 7.9 ohms, while at the same working switch is equal to 8.7 ohms;
3. The insertion loss at 1GHz frequency equal to about -5 or -6 dB, although earlier it was equal to -1 dB or less.

After changing to a new chip all returned to normal. But what's the problem? The scheme in accordance with the datasheet, but for some reason chip fail.
Scheme on and topology can be found in the appendix.
According to the scheme:
- 3.3V power supply voltage switch is derived from the linear regulator ADP3335ACP-3.3.
- Bypass capacitors: input RFC GRM1555C1H102JA01, the outputs RF1 and RF2 GRM1555C1H101JA01.
- Filter capacitors to VDD GRM155R71C104KA88 and GRM1555C1H101JA01.
- EN is connected to ground.
- Vctl on SW_OG_EN circuit controlled by a microcontroller PIC18F67J60 with logic voltage level 3.3V(logic 1) and 0 (logic 0).
- When the RFC is connected to the RF2, then the input RFC comes microwave signal 1 GHz -20 dBm power through connector XW1 and RF cable from the generator.
- When we switch to the RF1 port, the situation is reversed, from the port of RF1 in the RFC come in a signal with a frequency of 10 MHz is -10 dBm, which is then arrives through XW1 connector on the spectrum analyzer.
According to the material: as a microwave board material using Rogers 4350B 0.254 with thick metallization 17mkm.

Microwave microstrip width of 0.54 mm (50 ohms).
I am sorry for my english)