I am using the LTC5536 rf detector at 5 GHz to generate start and end of transmit packets for 802.11n to 802.11ac (MCS0 to MCS9). In the datasheet it states the impedance for the Vm threshold voltage should be low and also a shunt decoupling capacitor may be required. The application circuit shows a 0.22uF to ground which I have designed in. The supply current into Vm is specified as -0.5 to +0.5uA but no indication of the required "low" supply impedance is stated. I plan to use 300mV to 500mV as my Vm voltage and have found only one LDO to date able to supply this low voltage of 500mV but the graphs and data are for 1mA output current and not less than 1uA. The Vm current is specified as an Input AND Output current so is an LDO appropriate for the supply of Vm? . Supplying Vm with a resistor divider then for a low current through the resistor divider of the order of 100 times the Vm current for voltage stability, and lets say the voltage at the top of the resistor divider is from a 1.4V zener, the total resistance is 1.4/50uA=2.8 kohm, so the two resistances are 0.9/1.4 x 2.8 kohm (1.8 kohm) and 1.0 kohm to give the 500mV. The Thevenin equivalent source impedance is then the parallel of the two resistors = 643 ohms, which I do not consider low. I would normally assume "low" is less than 1ohm. As there is a shunt capacitor of 0.22uF is the "low" impedance referring to the ac impedance and the resistor bias network simply needs to provide the current that is being supplied to the Vm port ie 0.5uA? ie "low" is not the dc source impedance? which can be in the 10 kohm or higher value?
Any feedback appreciated, Charles_L
Using resistor divider with your values of 1.8kohm and 1kohm would work. the thevenin equivalent is a lot lower than the input impedance of the pin. as long as the current setup by the resistor network is more than 10x of the input into the Vm, it will be okay. the low impedance source should be low relative to the input of the Vm pin, but using too low of resistor is burning extra current which can be a unnecessary. shunt capacitor is needed to minimize the input noise.
Thanks very much for your answer. As an RF engineer I guess I overlooked the block diagram and the narrative stating the Vm port is tied directly to the negative terminal of the comparator which is an op-amp.