Joe, I wrote ASM code for the 16F1508 that uses CLC1 to "AND" the NCO1 output an= d a 32kHz PWM4 output. (Again, the purpose is to pulse all high-going puls= es on/off so as to decrease the volume at the speaker.) I then send the CL= C1 output to the CWG to create the complementary dual waveforms with deadba= nd. But the inverted result of the CWG, while accurately inverted, is unde= sirable for the reason shown below. NCO generated waveform *before* CLC ANDing: +-------------------+ | | -------------+ +---------------------- CWG1A, *after* ANDing the NCO waveform with 32kHz PWM: +---+ +---+ +---+ | | | | | | -------------+ +---+ +---+ +---------------------- CWG1B (inverse of CWG1A) -- BAD!: -------------+ +---+ +---+ +---------------------- | | | | | | +---+ +---+ +---+ But CWG1B is not what is needed. Here is what the Normal and Inverted need= to be: Normal (same as CWG1A above): +---+ +---+ +---+ | | | | | | -------------+ +---+ +---+ +---------------------- Inverted -- GOOD! (I need to generate this!): +---+ +---+ +---+ +---+ +---+ | | | | | | | | | | + +---+ +----------------------+ +---+ +---+ +=20 Using the CWG with the NCO (no CLC ANDing) to create dual complementary out= puts (i.e., a clone of the NCO output and its inverse) works fine. But whe= n using CLC to AND my NCO waveform with a 32kHz PWM squarewave, we have the= problem with the complementary waveform, as shown above. =20 Again, the reason I am using CLC to AND-gate the NCO output with a 32kHz sq= uarewave (PWM output) is because I need to pulse each high-going pulse of t= he NCO to reduce volume at the speaker. As shown above, simply using CWG to= invert the main waveform on CWG1B doesn't create the needed inverse wavefo= rm because CWG1B is just the inverse of CWG1A. But what's needed in the co= mplementary waveform is pulsing of only the HI-GOING pulses of the original= waveform, leaving LO to be LO. I just don't know how to accomplish that. I thought about putting the two CWG outputs into the CLC, since CWG creates= deadband and synchronizes both waveforms, important for my H-BRIDGE. Ther= e are 4 CLC modules I can use in the 16F1508 too. However, the 16F1508's C= LC does not accept CWG as a gated input! Whatever solution is used to create the needed complementary waveform needs= to take careful consideration of the H-BRIDGE. That is why I use CWG to k= eep the Normal and Complementary outputs in synch and to generate deadband = on the complementary waveform. Deadband ensures that neither the Normal no= r the Inverted waveform will ever be HI at the same time. And since the CW= G is used to create both waveforms, there is no drift-separation of the wav= eforms over time, which may result if I use multiple CLC modules and experi= ence propagation delays. Any ideas? Thank you, James Fri, 29 Jul 2016 13:12:22 +1200, "IVP