1. What is the function of "PWRDWN" switch on evaluation board?? No matter "PWRDWN" is pull high or pull low, ADRF6820 still working.
2. When using internal LO,
RFIN : 1900MHz, 0dBm, CW tone
REFIN : 153.6MHz, 0dBm, CW tone
LO set to 1890MHz (using customer evaluation software)
=> I+ port (we remove transformer T6 and resistors R38 & R42) see a ~38MHz ~358mVpp sinewave (oscilloscope is set to 1Mohm)..
When using external LO,
RFIN : same as above setting
LOIP : 1890MHz, 0dBm, CW tone
=> I+ port see a ~10MHz ~273mVpp sinewave (oscilloscope is set to 1Mohm).
Our question are:
a. When using internal LO, why the IQ sinewave frequency is not ~10MHz??
b. Why the amplitude level of I+ signal has large difference between using internal LO and external LO??
c. For external LO case, do you have a formula to calculate the I+ Vpp value with RF input power level??
Not sure why you're not seeing the chip power down. What indication are you looking for, loss of output signal ?
It appears the internal LO is not locked in your first case.
From the EVAL ADRF6820 User Guide, pg 4: "For optimum performance using an external LO source, the LO input pins(LOIN− and LOIN+)should be driven differentially. The wide input range of the external LO inputs span from 350 MHz to 6000 MHz. Unless an ac-coupled balun/transformer is used to generate the differential LO, the inputs must be ac-coupled. The input impedance of the differential LO signals is 50 Ω."
Also make sure the LO drive level to the board is at least -6 dBm after any external baluns, cables, etc.
In addition, it's probably best to match the I/Q ports as well:
Again, from pg 4 of User Guide:
"On the evaluation board, the quadrature I/Q outputs are terminated with 200 Ω differential. This is achieved by using a 1:1 transformer and an L-impedance matching network. The 1:1 transformer, TC1-1-13M+ from Mini-Circuits, is a very wide bandwidth (4.5 MHz to 3000 MHz) transformer that allows a flat pass-band response of up to 1 GHz."
We do not advise removing one leg of the I output.
Could you please help to check the attached file about our setting?
Thanks for your inquiry into the ADRF6820. I'll address each of your issues in series.
Pwr down switch not working: This is a known bug that we don't yet have a solution for.
IQ freq offset incorrect: When using the internal PLL, you will need to set the PFD polarity box to “NEG”. This will move the LO frequency to the expected value. You can measure this LO frequency by selecting the LO_Output pull-down to "LO_DRV1X_EN", and connect a spectrum analyzer to the LOOP SMA connector near the center of the board.
The pictures below represent the GUI interface that you should see:
Depending on the load you're presenting to the IQ ports, the IQ signals may vary in amplitude. To assure consistent loading to all ports, you can install a 10 Kohm resistor to ground on R39-R42. Also, please verify C101 and C108 are not installed. If the imbalance remains then please let me know and we can look further.
We did some tests with ADRF6820 EVB and found the IQ amplitude level was not big enough. Could you help to check our test result, as attached file, that it is reasonable or not.
Due to our base-band platform is differential IQ port with 50R impedance, so we plan to connect with ADRF6820 IQ differential ports directly without any transformer. Is that okay?? Do you have any comments or suggestions??
The impedance presented to the IQ output should be 200 ohms differential. This is accomplished with resistive pads on our EVB, to allow connection to a 50 ohm instrument and still present 200 ohms to the part. This padding gives 17 dB of attenuation however, so please take this into account in your circuit. Otherwise, you'll need to use a 4:1 transformer as noted in the user guide, pg 4, to transform from your 50 ohm load to the desired 200 ohm load for the IQ ports.