ISL55210-ABEVAL1Z Intersil, ISL55210-ABEVAL1Z Datasheet - Page 5

• Manufacturer Part Number: ISL55210-ABEVAL1Z
• Description: Amplifier IC Development Tools ISL55210 Eval Diff. Amp, Active Balun
• Manufacturer: Intersil
• Type: Differential Amplifiersr
• Series: ISL55210r

Specifications of ISL55210-ABEVAL1Z

• Rohs: yes
• Product: Evaluation Board
• Tool Is For Evaluation Of: ISL55210
• Operating Supply Voltage: 3.3 V
• Description/function: Evaluation board to test the performance of the active balun
• Maximum Operating Temperature: + 85 C
• Minimum Operating Temperature: - 40 C
• Operating Supply Current: 34 mA
• For Use With: ISL55210

For this design point, only an 11.8Ω physical input resistor was
required to show a nominal 50Ω input match to the source.
Measuring that input impedance from 1MHz to 500MHz using an
HP4195 analyzer set up for impedance measurement gives the
screen picture of Figure 7.
This is showing <±2Ω deviation from 2MHz to 500MHz (≤28dB
return loss) with nearly exact match over a broad range from
3MHz to 300MHz. The markers at 100MHz are showing an input
magnitude of 50.3Ω with 0.9° phase. This voltage feedback
based FDA is doing a remarkable job of transforming that 11.8Ω
R3 element in Figure 4 into a 50Ω match using its very wideband
common mode feedback loop.
The Noise Figure (NF) in this relatively low gain setting calculates
to approximately 7dB. This low NF is depending on the low
0.85nV/√Hz differential input spot noise for the ISL55210 and
the reduced resistor values enabled by this active input match
capability. Going to higher gain settings will reduce the input NF
where for gains >22dB it drops below 6dB; a remarkable number
for a 115mW amplifier. These numbers are from an analysis
using only the differential path noise numbers. Since this
topology will also have a common mode noise component, the
measured NF will be slightly higher due to component
mismatches converting a small portion of the common mode
noise to differential. Measuring the spot output noise for the
default configuration of the active balun board with a 50Ω
termination on the input and referring that measurement to the
FDA differential output pins gives the plot of Figure 8.
This low output spot noise can be input referred by the 6.6V/V
gain setting and converted to a Noise Figure, as shown in
Figure 9. The Vcm control input has an internal 30MHz filter that
might explain the decrease in noise going up through 40MHz.
That bandlimiting is only on the buffer stage to the internal
reference for the Vcm loop. The internal loop bandwidth is the
much higher >1.5GHz required for this circuit to operate
successfully. The midrange 7.9dB noise figure exceeds the
expected 7dB noise figure probably due to common mode to
differential conversion and more precise matching in the
external resistors might move this closer to theoretical. No effort
here was made for extremely precise matching and simple 1%
resistors were loaded.
FIGURE 7. MEASURED INPUT IMPEDANCE FOR THE CIRCUIT OF
FIGURE 1
5
Application Note 1831
The output 3rd order intercept for the circuit of Figure 1 may also
be measured. The typical definition for the OIP3 is from a
matched source to a matched load. With the 6dB loss inserted by
the matching elements that drops the reported OIP3 from the
FDA output pins to the matched load by 6dBm. Using two 25Ω
series outputs to a 50Ω load shows a total 100Ω differential load
across the FDA output pins. Emulating that with the ADT1-1WT
output option of Figure 4 by putting another 200Ω load directly
across the FDA output pins (the ADT1-1WT path is set up to look
like a 200Ω load) gives the higher frequency OIP3
measurements of Figure 10 with the straight line a
superimposed line fit
application of this net gain of 10.4dB circuit as if it were driving
through 2 - 25Ω series output to a single differential 50Ω load of
some following element. Driving into an ADC or interstage filter
to an ADC, taking less insertion loss, and/or a lighter load will
give higher OIP3 than shown in Figure 10.
FIGURE 8. MEASURED OUTPUT NOISE FOR THE CIRCUIT OF FIGURE 4
FIGURE 9. MEASURED SPOT NOISE CONVERTED TO A NOISE FIGURE
8.5E-09
8.0E-09
7.5E-09
7.0E-09
6.5E-09
9.0
8.8
8.6
8.4
8.2
8.0
7.8
7.6
7.4
7.2
7.0
10M
10M
[3]
. Again, this is a worst case for an
DIFFERENTIAL NOISE AT FDA OUTPUTS
FREQUENCY (Hz)
FREQUENCY (Hz)
NOISE
April 5, 2013
AN1831.0
100M
100M