We just received a new assembly of our pcbs, and a UCD9248 design that previously worked does not. We're trying to determine if there was an assembly error or if we have bad/miscalibrated chips. The symptoms are that with the same circuit, our old board produces a V33A of 3.3V (using a circuit similar to figure 6 in the datasheet). In our new assembly of these boards, the V33A is actually 3.7V-3.8V, even though all of the resistors measure correctly. The other major symptom is that when we try to change the PWM frequency, it does not match the frequency set by the fusion power software. As we increase the frequency, the actual PWM rate increases, monotonically (and possibly linearly), however it is reduced from the set frequency. For example, if we program a freqency of 1000kHz, the measure PWM signal is only about 400kHz, a PWM set of 500 produces an actual rate of about 200. On our old build of this board, the PWM frequency matches what we set it to. Lifting the appropriate pins and providing our own 3.3V does not fix the PWM problem.
Is there anything external to the chip that could cause this problem?
thanks,
Amanda