UCC28064A: Relationship with VPHB_HR, VPHB and VPHB_LR

Hello Takahara-san,

The UCC28064A transition-mode, interleaved PFC controller can be used in a wide variety of AC and DC applications, but it is optimized for Boost-PFC covering the world-wide AC-mains range (the so-called "universal line range")  from 90Vrms to 264Vrms. Everything following below also applies to the UCC28065 PFC controller. 

Yes, VPHB_LR designates a threshold voltage for the Phase-B turn-off in the case of 90 to 132 VRMS (US mains), considered the "Low-Range" (_LR).
Yes, VPHB_HR designates a threshold voltage for the Phase-B turn-off in the case of 180 to 265 VRMS (EU mains), considered the "High-Range" (_HR).

VPHB = VPHB_LR when targeting a PHB threshold while operating in the Low-Range, typically 90 to 132 VRMS.
VPHB = VPHB_HR when targeting a PHB threshold while operating in the High-Range, typically 180 to 265 VRMS.
So, VPHB is either VPHB_LR or VPHB_HR, depending on which input voltage range is active.

It should be noted that the actual threshold for changing ranges from _LR to _HR is when the 3uA source out of PHB is activated.
The 3uA source is activated when a rising voltage at VINAC crosses above 3.5V, and 3uA is disabled when the peak of VINAC has been detected to fall below 3.15V.  This can be seen in Figure 23 of the UCC28064A datasheet.  
Ultimately, it is the resistor divider network on the VINAC pin that determines the input low-range and high-range for any given application.

The purpose of the 3uA PHB hysteresis current is to provide the user the means to keep light load efficiency higher at high line (_HR).  If Phase-B is shed at the same output power level regardless of low-line or high line, then high-line efficiency would be lower.  This is because the _HR currents are lower and switching frequency is higher, so switching losses are higher.  By shifting the PHB threshold to a higher COMP voltage at high-line (corresponding to a higher output power), a higher light-load efficiency can be maintained when operating at high line.

Note: there is no means to disable the 3uA PBH current, so it is impossible to set VPHB_HR = VPHB_LR.  The ΔVPHB of equation 9 is the result of 3uA acting on the Thevenin equivalent of RU and RD which cannot be reduced to zero.  In fact, RU + RD must be high enough to avoid overloading VREF, which has a 2-mA limit.
VPHB_HR must be chosen > VPHB_LR by some reasonable amount. Since equation 5 shows VPHB_LR to be independent of the 3uA source, it is usually chosen first.  Then VPHB_HR is chosen based on previous experience or an empirical estimate.  Final values of RU and RD can be adjusted based on prototype evaluation.

Using the example PHB resistor values you provided (RU=133kohm, RD=47kohm), eqn 5 shows VPHB_LR = 47/180*5V = 1.306V, and eqn 6 shows VPHB_HR = 1.306V + 0.104V = 1.410V.  Here, ΔVPHB = 0.104V.  During low-line condition, VPHB = VPHB_LR = 1.306V.  During the high-line condition, VPHB = VPHB_HR = 1.410V. 

Using the corrected equation 10 (denominator term "VREF" replaced by "1") and rearranging terms, we can find the load percentages corresponding to VPHB_LR and VPHB_HR: 
%Pout(phb_LR) =  (1.306V - 0.125V)/4.825V = 0.24477 = 24.48% of Pout(max), and
%Pout(phb_HR) =  (1.410V - 0.125V)/4.825V = 0.26632 = 26.63% of Pout(max).

On the other hand, when solving for RU and RD resistor values for Pout_LR = 28% and Pout _HR = 34% (for arbitrary examples), corrected eqn 10 provides:
VPHB_LR = 4.825V*0.28 + 0.125V = 1.476V and VPHB_HR = 4.825V*0.34 + 0.125V = 1.766V.  ΔVPHB = 1.766V - 1.476V = 0.29V.

Eqn 7 shows, RU = (0.29V *5.0V)/(1.476V*3uA) = 327.4kR, and eqn 8 shows RD = (0.29V*5.0V)/((5.0V-1.476V)*3uA) = 137.2kR.    
Using the nearest E24 standard values of 330kR and 130kR the actual PHB thresholds become (by eqns 5 & 6), VPHB_LR = 130/460*5.0V = 1.413V and VPHB_HR = 130/460*5.0V + 93,261R*3uA = 1.413V + 0.280V = 1.693V.  LR and HR PHB-power thresholds are 26.7% and 32.5% respectively, as a result of the choosing the nearest E24 values. 

I hope this clarifies the PHB design process.

Regards,
Ulrich