MAX2870

The 3dB pad (i.e. attenuator) is the pi-circuit consisting of the two 300 ohm resistors to ground with the 18 ohm series resistor between them.  If the RFOUT SMA has a 50 ohm termination such as when it is connected to a spectrum analyzer with 50 ohm input, the MAX2870 RFOUT will see a 50.6 ohm load to ground which can be confirmed by circuit analysis of the resistor network consisting of the pi-circuit and the 50 ohm termination resistor.  However, the voltage at the equipment side of the 18 ohm resistor, call it Vout, will be 0.704*Vin where Vin is the voltage at the RFOUT pin of the IC.  Squaring this to get power ratio gives Pout = 0.5*Pin or 3dB power drop.  To eliminate this 3dB pad, remove the two 300 ohm resistors and replace the series 18 ohm resistor with a 0 ohm resistor.

Regards

Andrew

Thank you so much Andrew! I have some further questions about the maximum output signal level. After removing the attenuation resistor networks, I observed amplitude inconsistency among the four output signals of the eval board. Also the maximum output power varies with the center frequency. As the frequency increases, the output power increases. I wonder what reason causes this phenomenon? To my understanding, in the software, the APWR and BPWR control the output power strength, and the RF_OUTA and RF_OUTB control the center frequency of the signal. Please let me know if there's any parameters I missed.

Hi Alina,

You state that you see power variation after removing the 3dB pads.  This surprises me.  I certainly expect power variation over frequency.  That is common for RF parts, especially ones operating over a 6GHz frequency range.  However, I would not expect the pad to have any effect other than attenuating the power by 3dB.  Your statements imply that you measured power over frequency and saw no variation prior to removing the pads.  Anyhow, that aside, there are reactive components with frequency-dependent impedance in the signal path, for example the 100pF AC coupling capacitors.  The impedance of this capacitor will vary from 64ohms at 25 MHz to 0.3ohms at 6GHz.  This capacitor would form an impedance divider with the termination resistor, and this would affect the measured output power.  At higher frequencies, parasitic inductances and capacitances also start to play a role in affecting the output power.

Your assumptions about APWR and BPWR are correct.  The signal frequency can be changed with the RF_OUTA and RF_OUTB widgets in the GUI.  Although typically RFOUTA and RFOUTB will be at the same frequency, or have an integer relationship with each other.

Regards

Andrew

Hi Alina,

You state that you see power variation after removing the 3dB pads.  This surprises me.  I certainly expect power variation over frequency.  That is common for RF parts, especially ones operating over a 6GHz frequency range.  However, I would not expect the pad to have any effect other than attenuating the power by 3dB.  Your statements imply that you measured power over frequency and saw no variation prior to removing the pads.  Anyhow, that aside, there are reactive components with frequency-dependent impedance in the signal path, for example the 100pF AC coupling capacitors.  The impedance of this capacitor will vary from 64ohms at 25 MHz to 0.3ohms at 6GHz.  This capacitor would form an impedance divider with the termination resistor, and this would affect the measured output power.  At higher frequencies, parasitic inductances and capacitances also start to play a role in affecting the output power.

Your assumptions about APWR and BPWR are correct.  The signal frequency can be changed with the RF_OUTA and RF_OUTB widgets in the GUI.  Although typically RFOUTA and RFOUTB will be at the same frequency, or have an integer relationship with each other.

Regards

Andrew