My issue is with the LTC5536 being used to detect 802.11n (2.4 GHz) transmissions from a WiFi radio, and the comparator output is being used to switch RF switches to route the signal to an amplifier, and turn the PA on. The circuit is attached.
- The RFIN pin is coupled to the RF input signal via a 22 pF capacitor. There is no reactive or resistive matching between the 50 ohm transmission line it is sampled from and the RFIN pin.
- The R4/R5 resistor divider provides approximately 140 mV reference to the VM pin.
- Vout is the TX enable signal which is connected to the appropriate RF switch lines & is used to enable the PA.
- Vout is also routed via a 74AHC1GU04GW inverter to produce the inverted signal required for the opposite RF switch pins. There are 100 pF capacitors on these lines close to the RF switch pins.
- TX and RX status LED's are driven from these lines also.
The issue is I'm seeing what looks like rail-to-rail switch bouncing, or indecision on the comparator for a short time during some pulse detections, and sometimes not, and sometimes for only part of the pulse. The net result of this is lots of spurious noise on the spectrum.
Attachment "scope_4.png" shows two different types of Vout result - both with the switching issues and without.
Attachment "scope_1.png" shows Vout exhibiting this switching issue, and then not.
Attachment "scope_3.png" shows a zoomed in look at the switching issue.
Do you have any hints or tips in tracking this one down?