Hello TI Team,
We have been struggling with inconsistent Category 3 SYNC results for the past several weeks. We have resolved a number clock issues and timing problems, but one issue remains.
We discovered, completely by accident, that register R49 has a major impact on the category 3 SYNC performance. We inadvertently overwrote the default value of 0x4180 with 0x0401; and the new value of 0x0401 yielded a very stable and repeatable phase relationship between the two LM2572 devices. We tried writing other values to R49, and found these results:
R49 Values |
Comment |
|
0x4180 |
Data Sheet Default |
Final Phase relationship after SYNC highly dependent on 100MHz OSCin input level, and the programmed output frequency. (Up to three final phase states are observed) |
0x4280 |
|
stable and repeatable phase relationship after SYNC operation. All frequencies work with no sensitivity to 100 MHz OSCin |
0x4380 |
|
stable and repeatable phase relationship after SYNC operation. All frequencies work with no sensitivity to 100 MHz OSCin |
0x0000 |
|
stable and repeatable phase relationship after SYNC operation. All frequencies work with no sensitivity to 100 MHz OSCin |
0xFFFF |
|
stable and repeatable phase relationship after SYNC operation. All frequencies work with no sensitivity to 100 MHz OSCin |
0x0401 |
|
stable and repeatable phase relationship after SYNC operation. All frequencies work with no sensitivity to 100 MHz OSCin (original accidental write to R49 that corrected phase sync operation) |
Questions:
1) Could you give us any insight into why this works?
2) Is this a legitimate solution?
3) Are we doing something that should be corrected by another more appropriate register write?
Thanks in advance,
Chuck Beam