Hai,
Can somebody help us how to improve the carrier and sideband suppression in TRF37017 Modulator. Few days back we have posted for this issue and got no good response. Kindly help us.
Question:
The In-Phase and Quadrature-Phase baseband signal (150MHz) output from DAC are fed to the modulator TRF370317. The LO given to the Modulator is 1.62GHz with 0dBm level from a clean source. The bias given to the modulator is 1.7V and the resistors in the resistor divider, used to provide this bias, are of 0.1% precision. With this the results what we observed are, 1) Sideband suppression of 12dBc Carrier suppression of 31dBm As far as level at modulator input is considered, capacitors in between the DAC and modulator will remove the DC, if any, on the DAC output, before feeding it to the Modulator. At Modulator input we are shifting the level to 1.7V, (as said in the datasheet [page no. 4]). The voltage levels at the input point of modulator, as measured in the oscilloscope, are IVIN = 1.749V IREF = 1.752V, QVIN = 1.748V , QREF = 1.752V. So the difference in the levels is 0.004V, which is negligible. Here i am attaching you my schematics and layout for reference. The output level what we are targeting are 1) Sideband suppression of more than 50dBc. 2) Carrier suppression of more than 50dBm. Let us know how to improve the performance. One Answer:
The In-Phase and Quadrature-Phase baseband signal (150MHz) output from DAC are fed to the modulator TRF370317. The LO given to the Modulator is 1.62GHz with 0dBm level from a clean source. The bias given to the modulator is 1.7V and the resistors in the resistor divider, used to provide this bias, are of 0.1% precision.
With this the results what we observed are,
1) Sideband suppression of 12dBc
Carrier suppression of 31dBm
As far as level at modulator input is considered, capacitors in between the DAC and modulator will remove the DC, if any, on the DAC output, before feeding it to the Modulator. At Modulator input we are shifting the level to 1.7V, (as said in the datasheet [page no. 4]). The voltage levels at the input point of modulator, as measured in the oscilloscope, are IVIN = 1.749V IREF = 1.752V, QVIN = 1.748V , QREF = 1.752V.
So the difference in the levels is 0.004V, which is negligible.
Here i am attaching you my schematics and layout for reference.
The output level what we are targeting are
1) Sideband suppression of more than 50dBc.
2) Carrier suppression of more than 50dBm.
Let us know how to improve the performance.
One Answer:
The sideband suppression performance can be improved by tuning the amplitude and phase between the I and Q ports. The carrier suppression can be improved by adjusting the DC offset between the I and the Q ports.
Result: After doing it improved a little but not showing the best results that we are expecting.
We've developed a SMA-connectorized module of the TRF37017 that includes differential amplifiers and low pass filters. Part number is AM0350A. At 1.62 GHz, the AM0350A has -40 dBc sideband suppression and -38 dBm LO leakage.
If you are seeing -12 dBc sideband then you are either (A) corrupting amplitude balanced between I and Q or (B) you are introducing phase skew between I and Q. Try adjusting the I/Q sinusoid frequency to < 1 MHz. If sideband stays the same then you can rule out phase skew.
Lowell's response is well stated. There may be one additional contributor to the poor sidbeand suppression. High harmoinc content (specifically the 3rd Harmonic) of the LO source can also degrade sideband suppression performance. It is stated that the LO source is from a clean source, but it may be a good idea to check its harmonic performance or add a simple BPF around the LO. In addition, the filter between the DAC and modulator is a common place to introduce phase imbalance in the system where passive component tolerances of the inductors is relatively poor. Lowell's suggestion to use a low frequency signal is a good experiement.
--RJH