UCC28061-Q1: Automotive Qualified Analog PFC Controller

Hello Aravind, 

Thank you for your interest in the UCC28061-Q1 interleaving TM-PFC controller. 

From TI's selection list I find that UCC28061-Q1, UCC28070-Q1, and UCC2818A-Q1 are the automotive-rated PFC controllers available from TI.  
The -061 is interleaving Transition Mode (TM), while the -070 is interleaving CCM, and the -18A is single-phase CCM. 

The -061 controller cost is lowest, but uses two boost stages in parallel to spread the losses out and allow for low-profile packaging.  Component count and cost may be high.
The -070 cost is highest, and interleaves for the same reason, but is more suited to much higher power conversion. 
The -18A cost is nearly as high, but is single-phase so number of components and their costs can be lower. 

Thermal design will depend on the ratio of high power time to standby time.  The time interval for peak power will determine if the PFC must be designed for 300W or can be rated for 200W (with 300W peak coming from stored charge in the output capacitor).  These timings have not been provided. 

Your requirements are achievable except #5.  Standby PF will be leading, not lagging, and it may be < 0.85 depending on how much X-capacitance is used in the line filter. 
I do not have a PF vs. load curve for the UCC28061 EVM, however, I have found this curve from the PMP11282 460-W reference design using a similar UCC28063 controller.

As can be seen, PF falls off below 0.85 at ~50W, which is about 11% of rated load. 
In your design, 10W is only 5% of 200W and 3.3% of 300W, so PF might be much lower than 0.85 unless the line filter is designed to use very little X-capacitance.
Light-load distortion will also factor into lower PF.

Regards,
Ulrich

Hello Aravind,

That "Burst Operation" signal in the UCC28061-Q1 (and UCC28061) datasheet is a mistake.  It should not be there. 
The block diagram is in error.  It was carried forward from the UCC28060 datasheet and modified (but apparently not completely) for the -061 datasheet.  

There is no Burst function in the UCC28061 device.  It is present in the UCC28060, but removed from the -061. See this comparison table:
UCC28060 vs UCC28061.pdf

Regards,
Ulrich

Hello Aravind, 

The fact that you application is for Avionics PFC is new information. 
Line frequencies from 350Hz to 800Hz reduce PF especially at light loads mainly due to high displacement current in the EMI filter X-caps. 
This current is 90-degrees out of phase with the line voltage and often exceeds the actual in-phase (resistive) load current. 

THDi still plays a roll in PF, but not as much as cosθ.  

We are not in a position to take measurements on the EVM and provide you with data.  That is not the function of this forum. 
In any case, data at frequencies > 60Hz are invalid since the EVM is not designed to support Avionics.   

Line filter capacitance affects the cosθ factor of PF independently of any PFC controller.
Of the limited automotive-qualified controllers currently available from TI, the UCC28061-Q1 has an advantage of being able to shed a phase under light-load conditions to improve (not only efficiency, but) THDi during such conditions. 

Regards,
Ulrich 

Hi Aravind, 

The UCC28180 is an 8-pin CCM-PFC controller which works by leading-edge PWM modulation to have the switching cycle off duty cycle D' follow the Vin over Vout ratio over the line cycle (which shapes the current).  The control is optimized for "universal line" operation (85~264Vrms) and ~400V output with little tolerance for deviations from this typical application.  It does not include input voltage feedforward.  It has a slightly higher frequency range than the -18A.

The UCC2818A-Q1 is an older 16-pin CCM-PFC controller which uses the traditional 3-input multiplier method for shaping input current. It also uses leading-edge PWM and includes input voltage feed-forward.  Having more pins allows more flexibility in adapting (statically and dynamically) to applications.  I think this device can probably deliver better THDi performance than the UCC28180, especially at lighter loads.  How much better I can't guess.    
But being an older process technology in a larger package makes it more expensive than the -180. 

Both devices are single-phase boost, so do not have any phase-shedding option for light-load efficiency improvement.  Neither have a burst mode feature, but I'm thinking it might be easier to implement one externally with the -18A than with the -180.  

Regards,
Ulrich

Hi Avarind, 

The UCC2818A-Q1 is an automotive-qualified version of the UCC2818A.  

Please see this App-note about the differences between UCC2818A and UCC2818.  By extension, the app-note will also apply to -2818A-Q1 and -2818.
Differences between UCC2817A-18A-19A and UCC2817-18-19 slua294.pdf
Basically, only the effective resistances of the pull-up and pull-down MOSFETs in the gate-drive circuit are different (higher in -2818A-Q1).

All three versions of the UCC2818x device are characterized, and parameters are guaranteed, to -40C.  
-55C is the absolute maximum limit on Tj for UCC2818 and UCC2818A, but parameters are not guaranteed for Tj < -40C. 
UCC2818A-Q1 has been rerated for an Abs Max lower limit on Tj of -40C and well as being characterized to that temperature. 
Basically, the effective lower Tj limit for all three are the same, at -40C.

Regards,
Ulrich  

Hi Avarind, 

Sorry, I missed your last question in my previous reply. 

The UCC2818x does not have any special mode for light load.  It will continue to operate as a fixed-frequency  CCM-PFC using average current-mode control.  At some light load level, inductor current may become discontinuous, so THDi may increase, but this controller will not change its basic cycle-by-cycle operation. 

Ast some very light load level, the VAOUT voltage may fall to less than the 0.33V "zero-power" threshold and the zero-power comparator will block gate-drive pulses (at DRVOUT) until VAOUT rises above 0.33V again (there is no hysteresis specified.)   Then operation may appear to resemble a burst mode, but it is mainly a random burst result, not a pre-programmed pattern.  

Regards,
Ulrich