AnsweredAssumed Answered

ADV212 Starting Encode

Question asked by jfmorris on Nov 3, 2015
Latest reply on Nov 19, 2015 by DaveD

I've got a 16-bit host interface to a ADV212 which has taken me a few weeks to work through, and I now have the encode application loaded into the ADV212, I load parameters, I see that the firmware is loaded correctly (0xFF82 in SWFLAG) and so on.  So I am now ready to encode. I basically followed the Figure 12 flow chart from the programming guide, to setup 16-bit normal host mode encode with a 64 word threshold.  I know IRQ's are working, as I get an IRQ before reading SWFLAG.


So following this flow chart, if I simply clear the IRQ flags by writing 0xFFFF to the EIRQFLG register, I should start encoding.  The input to the ADV212 is simply the output of a ADV7180, which I have confirmed is outputting a proper VCLK and 8-bit VDATA, which ties to VDATA[11:4] of the ADV212 (VDATA[3:0] of the ADV212 have 10K pull downs to ground).   I am relying on embedded EAV/SAV codes.


I never ever see an IRQ from the ADV212, or see EIRQFLG[1] get set to 1.  Below are the values I am loading for parameters. Am I not setup right for NTSC 4:2:2 data with embedded syncs?   I've used the ADV7180 on many designs, and its always come up out of the box encoding the analog input on AIN1 to produce the BT656 data stream.


Thanks for any help!


uint16_t adv212_parameters_data[] = {

0x0000, // 0x01, 0x00 : PREC, VFORMAT : 0x00 = 8 bit precision, 0x00 = NTSC 4:2:2

0x0301, // 0x03, 0x02 : UNI, XFORMLEV : 0x03 = Unipoloar C Unipolar Y , 0x01 = 1 level of transform

0x0000, // 0x05, 0x04 : WKERNEL, CBSIZE : 0x00 = 9x7 kernel, 0x00 = 32x32

0x0000, // 0x07, 0x06 : ATTRTYPE, STALLPAR : 0x00 = No attribute data, 0x00 = capture all fields

0x0000, // 0x09, 0x08 : RCVAL, RCTYPE : 0x00 = no target size or quality factor, 0x00 = no rate control

0x0000, // 0x0B, 0x0A : RCVAL : 0x0000 - no target size or quality factor

0x0000, // 0x0D, 0x0C : PICFG, J2KPROG : 0x00 = Use embedded SAV/EAV codes, 0x00 = LRCP

0x0000, // 0x0F, 0x0E : COD_STYLE, QFACT L 0x00 = Output code stream format 0x00 = 256/256 factor