Problem output AD9833

Good Morning
a few days ago I asked a question if it would be possible to generate two sine signals in quadrature with two AD9833. Well, I did it. But I have a problem with the output signal from AD9833. The signal does not come out with many problems that I will describe in below topics and I hope that you of Analog Devices can help me at that moment, because I am already hopeless.

1. The sine signal of AD9833 came out very distorted, at first I thought it was normal because there is not a low pass filter on its output, but after searching the internet I saw that the signal from my AD9833 module was totally out of others. As can be seen in picture 1 that accompanies the message the signal is very noisy.

2. When the sampling period on the oscilloscope screen increases, it is possible to see that there is a signal offset oscillation. I do not know why this occurs but this is making me so disturbed. I recorded a video. The quality is not very good however you can see the signal and some oscilloscope settings.

3. In the video I tried to show how the signal gets varying some settings. It is possible to see that when the oscilloscope has a bandwidth of 20 MHz (image 2) there is a significant reduction of the noise and the signal approaches a well-constructed senoid, it is also worth remembering that the trigger coupling is to reject high frequencies and when I put on CC edge the signal becomes unrecognizable.

4. I have also tested with several AD9850 modules and the same thing happens. I keep trying to understand that it is not a defect in DDS devices. But at the same time I think it can be, because when I take the function generator and connect to the oscilloscope input the wave comes out clean and without any noise. Also when doing the tip test the signal comes out perfect.

5. The crystal oscillator that I use is 25MHz which is the limit of the AD9833. On occasion I went to check the signal and I came across a clock signal that had a frequency of 25MHz but did not seem close to a square wave or a sine wave. The signal can be seen in image 3. So I do not know what else to do.

6. The programming is right because I have tested several libraries and they have shown me the same result. I used AC coupling.

[edited by: matlaier at 2:22 AM (GMT 0) on 29 Jun 2019]