I was wondering if someone at Analog could advise on the following.
I am currently designing a 24V 0.2A SEPIC supply utilizing an LT8362 for the effect.
The SEPIC forms part of a multi output SMPS design featuring out of phase sync lines feeding every pair of PWM switchers towards achieving lower EMIs.
The out of phase syncs are generated by two LTC6908-1 spread spectrum clocks operating at different frequencies.
When using the int osc, i.e. MODE pin NOT synced to an ext clock, the simulations on LTspice show a flat, well regulated 24V output with good ripple (about 60mV or so) across a voltage range of 20 to 60V. This is my baseline in terms of output quality.
However when syncing the LT8362 directly from the LTC6908-1 device, the sims show the output varying as much as 4V across the 20 to 60V range i.e about 19.5V at 20V in, and 23.7V at 60V in, which from a requirements perspective is not acceptable.
Found a possible workaround which is to feed the ext sync from the LTC6908-1 to the MODE pin via a 10k/100k voltage divider.
I am however uncomfortable with the idea of commiting an "unknown" to a design to be commercialized without fully understanding its operation/implications. The sync is reestablished without the output drop yet I am none the wiser why it seems to work?!
Given that the datasheet doesn't offer much added detail on the internal circuitry and/or examples on how to drive the MODE pin from an external clock source apart from a straight wire between devices, I would like to ask for some more clarity on the best way to interface the two ICs together as they are both Analog products supposedly designed to work well together (I would kindly refer you to the TI TPS54160 datasheet as an example where a similar application is comprehesively illustrated).
Thanks in advance for your help
Senior Hardware Engineer.