Post Go back to editing

Low frequency spurs out of LTC5586 IQ Demod board

Hello,

I am trying to use the DC2349A eval board for LTC5586 I/Q demod chip in a zero IF configuration. My LO is set at 1900 MHz and I have an RF signal at the same frequency. When I have nothing connected to the board and just powered on, I see a lot of low frequency spurs being output by the chip. This is destroying the EVM for my LTE signal that i am trying to sample in using an ADC. 

How do I combat this? The chip should not produce these spurs. I have to tried to check if its coming from the powersupplies or the USB configuration board but both of them dont seem to be contributing. Please see attached pictures. 

Parents
  • You need to do 2 things:

    1. Connect and apply 0dBm LO input

    2. Measure the output differentially through a combiner

    See demo board manual for proper connections.

    The LTC5586's LO signal chain consists of a programmable input matching network, a precision quadrature phase shifter, and buffer amplifiers. The limiting amplifiers have significant gain. When no LO signal is applied, all the noise in LO chain is gain up and propagates to the IF output. When an external LO signal is applied, the amplifiers are driven into hard compression turning the sinusoidal LO input signal into a square wave signal. The noise is only amplified during the very short periods of transition when the amplifier is linear.

    The IF outputs of the LTC5586 are differential. Common-mode noise is present on each output pins. Since the noise on the + output is 1080 degrees out of phase from the - output, when measured differentially the common-mode noise cancels.

  • Also I should have mentioned in my original email. We did use a differential to single ended opamp configuration (AD4857-2) to convert the differential to singled ended. we still saw the noise at the output.

    Sai.

Reply Children