UCC28700

Saurabh,

1) UCC28700 can be used with output voltages than the reference if you are using it as a standard PSR approach where you have an aux winding to provide a reflected output feedback voltage to Vs pin. You can set the turns of the aux higher than turns of secondary(output) so that voltage is higher. for example, you may see a transformer with Npri: Naux: Nsec of 20:4:1 for 3.3Vout design. see attached design below. You may also consider using UCC28704 which is newest version of UCC28700 released recently in February. it is improved in may ways. I will get an engineer from my team to comment on questions 2) and 3).

regards,

John

PMP8595_RevA.pdf

Hello John and Saurabh,

The UCC28700 and most of TI's PSR controllers use control law for AM, FM, AM modulation to control the maximum duty cycle.
This control will not work if the output diode/demagnetizing (Dmagcc) ratio is greater than 42.5%.

Vcst is the peak current sense signal that will be modulated from 0.25V to 0.75V
Vccr is the constant current regulation voltage. This is supposed to work with the RLC resistor to provide better constant current regulation this is internal to the IC and they use it to calculate Rcs. This is around 318 mV and is only shown in table 6.

Regards,

Mike

In my design :

Output Voltage is 3.6V@100mA

Primary Inductance : 6mH
Turns Ratio Np/Ns : 17
Np/Na : 4.7
Naux/Ns : 3.6

Rcs : 2R
Rs1 : 150K
Rs2: 47K
Rlc : 800R

The scematic i am using is same as that of evaluation board for 5V@1A.
if i increase the value of RLC to 4.7K, i get oscillation on VDD pin :

I have one more observation, when i monitored the drain current , for 3 cycles it was 110mA, but after three cycles its shape changes, a steep rise in drain current showing as if the transformer is getting saturated. Max current is 340mA, and most probably i think the transformer is getting saturated, but what could be the reason.

I monitored the duty cycle at 230V AC, around 15%. Ton :5.7us, fsw:27KHz
I reverified from design calculator, where in it specifies primary inductance for rated power to be 7mH, so what could be the issue

At start up, the UCC28700 limits the first three current pulses to IPeak(min) (the same peak current level that the device will allow during the light load frequency modulation mode), no matter what the load is at start up.  The controller limits these first pulses in order to monitor for any initial input or output faults at start up so it can limit power delivery.  Once the three pulses are done and no fault is detected, the converter will operate at the mode demanded by the line and load conditions.  Starting up with limited power delivery also helps eliiminate any chance of audible noise at power up. 

The fourth pulse, and all the pulses after that, should not indicate signs of saturation unless your transformer really is saturating (do you have a waveform to share?)

Thanx for the input, i will share the waveform.

Their is one more point i would like to share. I am using the bias winding itself to drive an additional load of 12mA. this load will be active for 5 sec only after every 40 Sec

To save transformer size and cost i dont want to use additional winding. Is this a right approach, if not what are the limitations

There are some considerations when using the bias winding to drive an additional load with primary side controllers. When the main output has no load on it, the switching frequency will decrease and there may not be enough energy per switching cycle to support the additional load on the primary side, so if the load on the primary side bias winding is greater than the secondary side at light load or no load conditions the secondary main output voltage may rise out of regulation. You need to make sure the pre-load resistor on the main secondary side output (Rpl in the data sheet schematics) is sized appropriately. You may need to reduce this resistor value so the main output dictates regulation under light load or no load conditios. This will unfrotunatley affect your no-load power so be aware of that.