Hi Engineerign zone,
I successfully setup my AD9910 few days ago, everything looks ok, PLL locked, chip surface temperature a little bit high (55~60 degree c)running 1GHz internal clock rate, and single tone frequency running smoothly from 1Hz to 19MHz (I only need 10MHz at most). However, today the amplitude became a little bit different, I attached my circuit, and if you refer to the circuit, I measured TP3/TP5 after reconstruction filter for the following waveform. it became about 20% shorter than it should be (0~500mV when RSET=10k, with 25ohm load). And it "was" okay.
after spending a day looking for the cause, I found when I turn off the inverse sinc filter function at CFR1, it become normal again (500mv like it should be), and I tried back and forth several times, it is consistent.
could anyone tell me if I miss understanding the purpose of the inverse sinc filter? the reason I choose AD9910 is because I want have my circuit a flat response the entire frequency span I care, now I am confused. any idea is appreciated.
I use single tone mode and setting registers by SPI, and design PLL loop filter follow exact value the recommended spread sheet has (please ignore the value in circuit, they have been changed), 42ohm series with 0.68uF and parallel with 0.022uF is the value to be used.
and one more question is PLL lock will sometimes become unlock for some reason that I haven't dig into(even if I cool the system with a fan). hard to say if there is any noise or glitch on master reset pin, because I have to do a power on reset to lock the PLL again. Any suggestion in my circuit will be welcome too. Thanks.