This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

UCC28700 - Not getting the maximum output?

Other Parts Discussed in Thread: UCC28700

Hi,

I have a circuit based on TI ref design (UCC28700DBVT)  using custom fly-back transformer.

The transformer was supposed to deliver 5V@4.2A,

Having constructed the prototyping board I could not get the maximum output.

I wonder if someone could point out what I am doing wrong.

Not sure if the circuit or the transformer itself.

I've attached here the schematic and the transformer spec.


Thanks.

DeejayUSB-Charger-012_flyBack.PDF750342786r00 Spec.pdf

  • Deejay,

    I have a few questions.

    When you say that you could not get the maximum output, do you mean that the output voltage started to roll off and lose regulation as the load current increased? If so, where did it start to lose regulation? If this is happening due to Dmax, then Vout will be falling gradually as load current increases.

    Or did the output hit the CC (constant-current) limit early? If so, at what current? In CC mode, the output voltage will fall more rapidly as the load resistance is decreased, but current should be held approx. constant - the load should be in CR (constant-resistance) mode to verify if the CC limit is kicking in.

    What was the input voltage during the test? Can more load current be taken if the input voltage is increased? 

    Looking at your transformer spec, Nps is very high at 20:1 - assuming you have an output diode rectifier (although you probably need an SR at this level of current, which do you use?), the reflected voltage to the primary will be about 20*5.5 = 110 V. So when the input voltage is at 110 V, the duty cycle will be 50% if the load is high enough to try to make the power stage reach the boundary between DCM/CCM. Since the UCC28700 always operates in DCM or at the boundary between DCM/CCM, at input voltage below 110 V, if the load is high enough to drive the power stage to this boundary, it will start to decrease freq to maintain boundary mode, and as a result the output will start to lose regulation.

    From the datasheet eqn. 21, Dmax is approx. 52%, assuming that you want to hit Fmax of 100 kHz and that the resonant ring period tr is 1 us. Typically, the design should aim for about 50% Dmax and min input voltage.

    From eqn. 22, Nps max is then 15:1, assuming Vbulkmin of 70 V (allowing for the bulk cap mains-freq ripple), Dmax of 50%, no cable comp, and output rectifier drop 0.5 V.

    From eqn. 23, Rcs should be approx. 0.47 ohm for Iocc of 4.8 A.

    From eqn. 25, Lp should then be 230 uH, assuming a target Fmax of 100 kHz

    Can you share more details about your spec, input voltage, target Fmax, etc?

    Thanks,

    Bernard

  • Hi Bernard,

    Thanks for the feedback.

    The input voltage is aiming to support 110VAC – 240VAC. During testing was run at 240VAC ~ 339VDC peak

     The prototype board was built as per drawing.

    -  input voltage was connected to 240VAC.

    - at no load the output voltage measured about 5.6VDC

    - loaded with 2 x 4.7Ohm (in parallel) load (effective 2.35ohm) , the output voltage drops to about 3.4VDC

    Notes:

    1. No sign of over-heating on the transformer or any other component except the input primary TVS.
    2. Have made changes to RS1 and RS2, didn’t seem to effect greatly
    3. The current sensing value 0.68 ohm was selected but made it worst, the voltage drops lower than 3VDC. To get current state was used 1 ohm resistor for the current sensing.

    I hope these help

    Thanks.

    Deejay

     

  • Deejay,

    Your schematic does not show the input AC rectification portion, how much bulk filter capacitance is used? This will dictate the minimum instantaneous voltage that the power stage needs to handle, and dictate the transformer turns ratio - 20:1 implies the min bulk cap level is about 100 V.

    The value of R206 127 k, together with the pri-aux transformer ratio, will set the typical turn-on level to 142 V peak (100 Vac), was this the intention? This seems a bit tight to the 110 Vac level, with tolerances it might not start up at 110 Vac.

    The value of R203 for CBC will set about 220 mV of trim up at full load, again I assume that this was the intention?

    The value of R208 for the peak current compensation vs. line seems to be very big. By re-arranging eqn. 17 in the datasheet, this implies that the IC is set to compensate for a turn-off delay of 870 ns - this is not correct, with 10-ohm turn-off resistor it will be much faster than that. I think this resistor R208 is the cause of the issue, since it will create far too much compensating offset, and will greatly reduce the actual peak current vs target. In addition, the CS pin RC filter time constant of 411 ns is way too big.

    At 339 Vdc, during the on-time the voltage on the aux winding will be -67.8 V, with -534 uA flowing out of the VS pin. Internally, this current is mirrored and divided by 25 and sent out on the CS pin - so 21.4 uA will flow out of the CS pin, developing an offset of 399 mV across R208. This will reduce the max peak current sense voltage across R209 to only 350 mV rather than the expected 750 mV.

    So the peak current in the power stage is much less than it should be, and this explains why the output is limited way below the expected level.

    R208 value should be recalculated, for 100 ns turn-off delay it should be approx. 2.2 k, for Rcs of 0.68 ohm. This would then give a filter time const of 48 ns, which will not affect the peak CS pin level. The IC does have internal leading-edge-blanking, so a CS pin RC filter should be small and maybe even not required.


    So I believe that a much smaller value of R208 will fix your problem, please try it and let us know.


    Thanks,
    Bernard
  • Hi Bernard,

    Thanks again. I haven't had a chance to try out your suggestion as I am on business travel and will certainly get back to you once I have more to feedback. Meantime, please see the attached for the input AC circuit to complement the flyback circuit.

    Thanks.

    Regards,

    Deejay

    USB_Charger-01_ACDC.PDF