Figures 3 and 4 demonstrate the GreenPAK outline. The wide thought is to drive the SET contribution of FSM1 low when the mains half-cycle begins, and bring it back high when the half-cycle closes. At the point when the half-cycle closes, the rising edge delivered by INV1 nourishes into DFF6/7/8, bolting the DCMPs‘ states into pins 14, 16, and 17 separately, after which the rising edge (postponed somewhat by DLY7) sets FSM1.
To make the above methodology work practically speaking, we have to work around two or three things. To start with, note that the GreenPAK 4’s DCMP works with 8-bit information, which offers a determination of just 1 section in 256. Imagine a scenario where we need better exactness. Second, the interior oscillator is not as precise as a precious stone oscillator. In this manner, on the off chance that we have to keep the outer parts number truly low, we require a technique for adjusting the recurrence screen. This is talked about later in the article.
For the outline, we picked a 2-MHz RC oscillator. The oscillator yield divisor and the FSM1 clock input divisor are set to 2 and 4, individually, so that the counter recurrence is presently 2000/8 = 250 kHz (period = 4 µs).
Things being what they are, what occurs with an ostensible mains recurrence of 50 Hz, with the half-cycle being 10 ms? Assume FSM1 is arranged to number UP with counter information = 0. At that point, toward the finish of the 10-ms half-cycle, the Q yield of FSM1 will be 10 ms/4 µs = 2500 modulo 256 = 196. We should call this the STOP an incentive for further discourse.
In any case, now we have to perceive the blunder in the oscillator recurrence. From the gadget’s datasheet, we see that if the SLG46620 works at a supply voltage of 3.3 V, the recurrence resilience of the 2-MHz RC oscillator at 25°C is –1.74%/+1.55%. Rather than including an outer gem oscillator, however, we can organize an adjustment method to make up for this variety in a genuine execution.
To accomplish this current, how about we begin the outline with an expected oscillator recurrence at the upper end of the range (or a little past, to represent minor temperature varieties). As it were, if the oscillator recurrence blunder was +2%, then the STOP esteem would be 128, which is the midpoint of the conceivable 0 to 255 scope of the STOP esteem. Realizing that the oscillator blunder is in reality under 2% infers that the real STOP esteem will be (somewhat) under 128.