ucc28610 less than 12w supply

 The actual limitation of an applicable design isn't really the overall output power, like the "12W" minimum would imply, but actually the peak primary current, the switching frequency range, and the maximum on-time.  12W makes assumptions for universal input, the minimum valley voltage of the bulk voltage, the derating of the HV MOSFET and its derating. 

The power stage is optimized for Max load, min line and must take into account the minimum of the rectified bulk input voltage, which is why the input bulk cap is an important parameter and can make-or-break the design.  At this min bulk voltage, max load, the primary current will ramp up to its peak value (and the resistor on CL is chosen to achieve this peak value needed for appropriate energy transfer at this min input/max load operating point).  Now this peak primary current will be travelling right thru the UCC28610...in the DRV pin and out the GND pin (note that this controller does NOT use an external current sense resistor).  The current required for current mode control is internally sensed (note the sense line shown in Fig 19 in the data sheet that  goes into the amplifier's non-inverting input and the output of the current modulator block goes into the inverting terminal?).  Because the peak primary current is sent thru the IC, a maximum limit is set (4A) that ensures bond wires aren't melted and other considerations so that sets the maximum power (approximately 65W with assumptions based upon the input voltage range, voltage rating of MOSFET, switching frequency clamps, etc). 

Now at the lower end of the applicable power scale, the peak primary current that is sensed inside the device has to be at a certain level that the logic circuitry can work with at light load.  Notice that the peak primary current is reduced by a factor of 3 at light load (FIG 2 data sheet top graph), so a programmed peak current of 1A at full load will be set at 0.33A at light load (GM).  And everything works well at this range of 1A max peak  adjusted down to 0.33A during GM (that's the amplitude modulation control).  So at light load, with a power stage appropriately designed so that the peak current at minimum input max load is at least 1A but not more than 4A the converter will work very well from no-load all the way to full load, even if this particulatr designis less than 12W.

The trouble happens when the power stage is designed to have a peak primary current at min line max load to be less than 1A because when it gets scaled down by a factor of three during GM the peak current goes down to the area circled in Fig 11 of the data sheet.  At currents down below the 0.33A zone, the logic doesn't function the way the power stage demands it should for energy transfer.  Note that the trace in the circled are to avoid changes slope... so the control logic will not be performing as it should down in this area.  But as long as the peak primary current is at least 1A at full load minimum input, the operating range will be satisfied.

The design calculator takes your desired inputs and goes through a couple of iterations to determine the right inductance to meet this peak primary current range, it also takes into consideration the switching frequency range, the minimum on-time that meets the internal timing constraints, and the maximum power constant, Kp, required for proper modulation.  So my calculator doesn't really stick with the generalized 12W to 65W limit but actually works to determine if there is some way to satisfy the User's requirements and meet all of the device constraints (I have a 10W design that works very well)

Hi Lisa, thanks for the complete answer, i will try with the suggested schematics given by the design calculator and will see the results. Is there a spice model of the ucc28610?

thanks again!