Hello,
We are running an open loop example DPWM code. Here's our setup: We have three DPWM modules active: 0, 1 and 2. All three have CLA_EN=0. Modules 0 and 1 output a PWM signal from both A and B channels, interleaved, each having a period of about P/8. So the waveforms look like:
|----_____________________|----_____________________|
|___----__________________|___----__________________|
|______----_______________|______----_______________|
|_________----____________|_________----____________|
Module 2 is setup such that it outputs another signal, which we intend to use as SYNC, in order to test external sync functionality:
|____________________-____|____________________-___|
Modules 0 and 1 have EXT_SYNC_EN = 1, Module 2 has that bit turned off.
With this configuration, when we connect the DPWM2 output to the SYNC_IN pin, we expect that the period of the DPWM0 and DPWM1 signals to be shorter, since we apply a sync pulse before the counter reaches the programmed period end. However, we see no difference with or without the SYNC signal.
Any ideas what we might be doing wrong?
By the way, what is the actual SYNC mechanism? Does the IC look at the rising edge of SYNC, falling edge of SYNC, or something else entirely? In the "data manual", SYNC pulse is defined to have a typical 256ns width, but does that apply to the case when SYNC is an output? What about when it is an input? Is there a requirement?