Recently I got the following answers to my questions I sent to Herman Valkenaers (Field Application Engineer Benelux). Is 27MHz not supported by the ADV7188 (while the default frequency mentioned in the datasheet is still 27MHz)? Is this causing my problems with NTSC?
By the way: I am using a Xtal oscillator and not a Xtal. So I don't use the ELPF caps.
Answers to questions sent to Herman Valkenaers:
• When we replaced the component by the ADV7188, we received a list of recommended settings. Some of these registers were not mentioned in the datasheet.
- Is there one list of recommended settings that is suitable for the ADV7189, ADV7189B and ADV7188?
>> There is no common set of recommended scripts for the ADV7188/9/9B. The “ADV7184/8 Changeover Document” describes the register differences between the ADV7188 and ADV7189B.
- Can we have an extended datasheet with an explanation of the ‘hidden’ registers?
>> Hidden registers are for internal debug/test use and are not disclosed to customers.
• We are reading 0x10 bit 1 to check if an input signal is present (to create a panic freeze option). We made it faster by setting register 0x51 to value 0x5F. When using the ADV7188, it looks like it takes more time for register 0x51 to change when input is lost. Is this right? Can we fix this by changing register settings?
>> Both the ADV7189B and ADV7188 have similar circuitry for the lock/unlock detect block. Could the delay be due to SW drive differences? What is the measured difference?
• Can the ADV7189 and ADV7189B be used on a pcb which has been redesigned for the ADV7188 (to make it backwards compatible)? We followed the ADV7184/88 changeover document (pins TEST6 & TEST7 tied to AGND and TEST8 tied to DVDDIO).
>> The ADV7188 is not backward compatible with the ADV7189B. The “ADV7184/8 Changeover Document” highlights that three of the NC pins on the ADV7189B have become TEST pins on the ADV7188 which are required to be tied to PWR or GND. For the board to be backward compatible I recommend placing series 0R resistors on these pins to allow the connection to be broken.
· Only the ADV7188 seems to be critical in combination with the crystal oscillator we used (25ppm). When applying an NTSC input signal and cooling/heating the oscillator, it causes a bad picture. The behavior changes when using other I2C settings. Can we solve this entirely by using the right settings?
>> If you use the proper scripts, xtal and hardware config as shown on EZ you will not have a problem. Note 27Mhz Xtal is not supported.Just a thought, the ELPF caps should be C0G (NPO) material. If they’re X7R or X5R the PLL can detune at the temperature extremes causing image issues. This is due to poor temperature coefficients of the X7R material.