MAX2391/MAX2392/MAX2393 Evaluation Kit

6

_______________________________________________________________________________________

3) Calibrate the power meter, with the low-power

head, at 2140MHz. A rough interpolation of the cal

factor does not introduce noticable error, if reading

the cal factor from a table.

4) Set the signal generator for a 2140.18MHz CW

(unmodulated) output at -27dBm, and connect a

3dB pad to the DUT side of the SMA cable. Use the

power meter to set the input power to the DUT at

-30dBm. Use measured attenuators and/or the

signal generator’s internal step attenuators (-40dB)

to reduce the signal to -90dBm.

5) Connect the RF source’s SMA cable and attenua-

tors to the EV kit’s LNAIN SMA input.

6) Connect the BNC cable from either I or Q to the

spectrum analyzer. Connect the other output into

the oscilloscope—be sure to set the oscilloscope’s

inputs to 50Ω, and not 1MΩ. Cable loss at 180kHz

is neglible; as long as cables are about the same

length, no calibration is required at the output to

observe proper signal level as well as proper I/Q

gain-and-phase balance.

7) Set one of the DC supplies to 2.8V and set a current

limit of 100mA (if available). Connect this supply

through the ammeter to “VCC_IC”, and readjust the

supply if necessary to get 2.8V at the IC when pow-

ered up. This supply connection only powers the IC

on the EV kit – read the ammeter to watch IC supply

current for the receiver. Connect another line direct-

ly from the 2.8V supply to “VCC_EXT” to supply the

external logic on the kit. Not having the voltage

drop of the ammeter inline means the voltage is

slightly higher than VCC_IC, but this does not

cause a problem.

8) Set the other supplies for ±5.0V with a current limit

of about 100mA. Connect these supplies to the +5V,

GND, and -5V on the opposite side of the kit. These

are the bipolar supplies for the MAX4444 differential

line drivers that buffer the I/Q outputs. Note that all

GND test points are connected to the same ground

plane—it is only necessary to use one of them.

8) Set the spectrum analyzer to span from DC (mini-

mum sweep) to 2MHz. Set the reference level to

+10dBm.

9) Set the oscilloscope for a sweep rate of about

1µs/div, DC coupling, with an amplitude scale of

about 100mV/div.

Testing the WCDMA Receiver

The power-up default state of the MAX2391 receiver is for:

• LO midband (2140MHz)

• LNA high gain

• Mixer high gain, normal linearity

• Powered on (out of shutdown)

1) Verify that the IC itself is drawing about 32mA (from

VCC_IC). The two MAX4444 differential line drivers

at the baseband outputs should draw about 80mA

from each of their supplies.

2) Use the AGC adjust potentiometer on the board to

set VAGC at +2.2V (maximum gain).

3) Spot-check the VCO tuning voltage (TUNE) to see

that the synthesizer is locked. The voltage should

be about mid-supply with the RFLO running at its

power-up default of 2140MHz. Disconnect any

leads from this before continuing, as the noise pick-

up onto the tuning line directly frequency-modu-

lates the VCO, and degrades LO phase noise.

4) Observe the 180kHz tone on the spectrum analyzer.

Adjust AGC to to achieve a -3.5dBm output level.

5) The on-board TCXO has a fine-tuning control—the

other potentiometer on the EV kit allows for external

temperature compensation of the TCXO to further

decrease frequency error. Adjust the TCXO poten-

tiometer if desired to bring the output tone exactly

to 180kHz.

6) Observe the other output on the oscilloscope. At

these input power levels, the SNR is typically much

too low to see the output tone through the noise. If

available, use the internal low pass filter option

(often 20MHz) and lots of averaging.

7) To make a gain/phase error measurement, connect

both outputs to the scope. Increase the input power

to about -50dBm, and back off the AGC until the

digital averaging to get both I and Q sinusoids visi-

ble on the scope. If automated measurements for

phase and amplitude are not available, use the cur-

sors to make the measurement. Calculate phase

error in degrees, and gain error in dB, and verify

that the results are better than 2 degrees and

0.6dB, respectively.