UCC28056: Feed back network

Hello Bright, 

Vos_reg is specified as 2.50V (typical) in the table on page 7 of the datasheet:

In the schematic diagram in Figure 26 (page 28 of datasheet), resistors R18||R19 and R20||R21 are connected in parallel in order to obtain fine tuning on the desired values.    R18||R19 = 36.5kR || 120kR = 27.99kR and R20||R21 = 75kR || 390kR = 62.9kR.   62.9kR is not a standard E96 value and the designer chose to parallel 2 standard values to obtain 62.9kR.  27.99kR is very close to the standard E96 value of 28.0kR, so I don't know why R18 and R19 were put in parallel to get a value that a single resistor could provide.  Maybe it was originally a non-standard value, and changed after some design iterations to a value coincidentally close to the standard value.  

I don't think the specific reason is important.  The concept of paralleling resistors is combine standard values to obtain non-standard values.

3x 3.4M and 66.5kR will regulate to 386V, not 390V.  It is possible to add a potentiometer to adjust the output voltage. 
Most designers avoid using potentiometers because:
a) they have high cost;
b) they require manual adjustment of every board in production;
c) their value setting may drift over time;
d) their long-term reliability is lower than fixed resistors. 

Regards,
Ulrich

Hi Bright, 

Yes, that is exactly correct.  You can change Vout following that calculation. 

 Please be aware that boost-PFC is non-isolated, so the AC input will be limited to 180Vac maximum, in order to allow about 15V difference between the peak of the AC line and Vout. 

Regards,
Ulrich

Hi Bright, 

In a non-isolated boost circuit (whether PFC or simply DC/DC), there is no series switch between the input voltage and the output voltage to interrupt and limit current when Vin > Vout.  So the boost inductor current cannot be controlled whenever Vin > Vout. 

Therefore, to maintain control of a boost converter, Vin must always be < Vout.   Or, put another way, Vout must always be higher than the peak of the highest Vin expected to be applied. 

The controller cannot limit the input voltage.  No controller can do this.  The user must enforce this requirement.

Most PFC converters are designed for "universal line" input of 85Vrms to 264Vrms.  The peak of 264Vrms is 373V, and Vout is usually set for 385~390Vdc. 
There needs to be about 10~15V difference between Vout and maximum Vin_peak to allow some demagnetization of the inductor current. 

If you are designing PFC Vout =270Vdc, then you must rate your application for operation at Vin < 180Vrms only.  Your application may not be plugged into any line voltage > 191Vrms because the input current will be uncontrolled and the PFC output voltage will rise to the peak of whatever line voltage is applied. 

Regards,
Ulrich

Hi Bright, 

Do you want 5A output at 270Vdc?  That is 1350W PFC output and waaayyyy outside the application range of the UCC28056.  

This TM_PFC controller is optimized for applications up to around 150W.  Going up to about ~250W is a stretch, although a gate-drive buffer can help drive a larger MOSFET.  The Webench program for UCC28056 is written to support normal applications for this part and is not flexible to accommodate unusual design targets. 

For 800W to 1350W, I recommend using UCC28180 CCM-PFC controller.  This power range is outside the usual range of TM-PFC, although UCC28064A can probably be stretched to deliver ~800W. 

Regards,
Ulrich

Hi Bright, 

Designing for 0.6A output will affect the component values and stress levels of each of the power conversion devices in the boost stage (inductor, output capacitor, MOSFET, diode, etc) as well as the current sense and compensation values around the controller.  Since Webench simulation is restricted to 150W or less,I really think you should use the Excel calculator tool to generate the information needed for a 0.6A, 390V PFC output. 

Regards,
Ulrich