To Whom it May Concern,
We are having a problem with the dynamic DC compensation of the AD9670. Because we use back to back diodes to clamp the signal after the TR switch, we are left with a diode voltage drop after the transmitter pulses. We are using active termination so there is was no DC path to bleed this voltage and hence was handled completely by the dynamic DC compensation of the AD9670. This resulted in prominent banding in the image. We added a FET to discharge this voltage for a duration of about 1uS after pulsing. This seemed to eliminate the problem. Now however, after improving the signal to noise ratio of the system, we are finding the banding caused by the dynamic DC compensation is still there. We are also seeing banding when using a TGC control waveform which I believe is also due to the dynamic DC compensation.
In our system, we are summing 8 channels of ultrasound together, so that banding which is not detectable on a single channel is easily seen in the final processed image. See attached photo. Concentric bands appear when dynamic DC compensation is tracking a DC offset. For the purpose of the photo, I setup system to maximize visibility of bands.
We are using active termination of 426 ohms, LNAgain of 21.6dB and reg 0x120[4:3] set at 2.0mS setting. Reducing value of CLG and CS seem to help. My conclusion is that if there is any dynamic DC compensation taking place, we will see it in the image as banding. I have the following questions:
- 1. Is it possible to disable dynamic DC compensation shortly after pulsing then re-enable it again after receive cycle?
- 2. Any thoughts or suggestions minimize or eliminate the banding.
Thank you for any help on this issue