On the latest revision of our board using the OMAP3530, where previous versions had apparently worked OK, this revision is exhibiting what looks like output amplifier instability on the composite video output. This came to light because the board failed pretty badly at the EMC test lab on noise at around 30MHz that disappeared when the video was disabled. I saw on another forum post where the output amplifier of the OMAP3530 can exhibit instability if there is excess cable capacitance. So I disconnected the output cable (which had not been connected to a TV) and placed a 75 ohm resistor right at the junction of L9, R75 and TV_out1 to ground. The only effect this had was to slightly increase the frequency of the instability. Is there something else I need to be doing with the filter circuit below? Interesting enough, when I load an OpenEmbedded image, the frequency of the noise is around 19.23MHz; loading and Android image drops it to about 16.13MHz. Also the video content seems to make a difference. The instability only appears on the vertical and horizontal blanking intervals of the OpenEmbedded image but is on most all of the lines of the Android image. Ideas anyone??? -Jeff
so...
Horizontal interval when running OpenEmbedded- noise appears only in the blanking interval and in the vertical sync interval.
On Android the noise is present for the entire field.
Answering my own post since no one commented here since I posted this originally. I just had to take the bull by the tail and face the situation myself it seems. For the full resolution to this please see http://www.fomsystems.com/case_studies/omap-noisy-ntscpal-video-issues/
Regards,
-Jeff