We are undertaking an initial investigation of the ADL5511 evaluation
board and have taken a few measurements to check
the linearity of the envelope detection functionality.
We have measured (VENV – EREF) for input signals
at levels from -20 dBm to +20 dBm using a continuous
wave input (CW). We understand that the part will not respond
linearly at the high end of this power sweep.
The board has been modified to allow operation below 2 GHz
as per instruction in the ADL5511 data sheet. The following
values are in-circuit:
FLT1: (C2) 1n5
FLT2: (C6) 8n2
FLT3: (C10) 12n
C1, the RF Input coupling capacitor, has also been increased
in value (less reactive capacitance) to pass on RF frequencies
well below 1 GHz.
Please refer to the attached graph in PDF format showing the deviation
from linearity (in dB) for three frequencies – 100 kHz,
10 MHz, and 1 GHz.
What is odd, however, is the deviation from linearity – referenced
to 0 dB at 0 dBm – exceeds 1 dB at around -15 dBm input for
all 3 frequencies. This is not as would be expected from the published
data sheet. 1 dB error at -30 dBm or better was our impression
from first reading the data sheet.
There seems little difference in behaviour
between the three frequencies, which would seem to rule out
incorrect filter corner frequency configuration (FLT1/2/3) as
a potential candidate.
Could anyone advise us, please, on which aspects of the evaluation
board setup and/or measurement setup should be checked in order
to resolve this anomaly?