ucc28810EVM-003 related questions

Jun,

Let me try to answer your questions.

1. Current regulation is done in the buck stage. This is a transition mode buck. the peak current in Q5 is sensed by R35 and R38 and is controlled by the UCC28811 at pin 4 where it is compared to a reference inside the IC. Once the peak current is reached the Q5 turns off. Once the current decays to zero Q5 is turned on. So the current in L3 is a saw tooth with constant amplitude and so constant average. This current then feeds the LLC section where it is transformed through T2 and T3 to the correct level for the LEDs

2. The voltage between PFC+ and VBUCK- varies depending on the number of transformers, their turns ratio and the LED voltage. If the turns ratio is n, number of transformers is t and the voltage on one LED string is VLED then the voltage between PFC+ and VBUCK- is 2 x t x n x VLED. The factor 2 is because this is a half bridge topology the transformers only see half the input voltage.

3. In theory this circuit can support any number of transformers in series at the primary. You cannot allow the transformer see an open circuit, hence the crowbar circuit on each output (R68, R69, D34, D38). This circuit works quite well with the EVM design. The voltage between PFC+ and VBUCK- is really limited by the AC input voltage and the output voltage of the PFC (not more than 30V greater than the peak of the AC). Once you know the minimum value of the PFC output, you know the voltage across the transformers, it will be half the minimum PFC output times the maximum duty cycle of the buck. Of course you know the maximum voltage for the LED string as you set this. With this information you calculate the transformer turns ratio. Where Vprimary is the buck output divide by 2 and divided by the number of transformers. The secondary is VLED max + Vf .

3. The Open Circuit protection is a simple over voltage crow bar circuit. Once the zener (D34) breaks down it turns on SCR D38. This diverts what ever current was in the LED string through R69 and D38. Power dissipation is R69 is ILED squared times R69. Power dissipation in D38 is ILED times Vf of D38. One thing to watch out for in R69 is the surge rating. This resistor discharges C40 and C49 from the LED string voltage (say 50V) to a couple of volts very quickly. Once in crowbar the other outputs maintain regulation. Not sure I would recommend operating for extended periods of time in crowbar mode.

Hope this helps.

Regards,

Richard.

Hi Richard,

Thank you so much for your explanation.

Based on your description, the average current in L3 decides the LED current, i.e.  I_L3 = I_LED / n, n is the T2 ration. and the peak current on R38 is 2 times of average current in L3.(correct?)

According to the Ucc28811 datasheet, the current sense threshold is decided by V_VINS and  V_EAOUT. In this design, V_VINS is a constant, V_EAOUT is integral of V_VSENSE error (2.5V - V_VSENSE), My guess is V_VSENSE of U3 should be about 2.5V, so V_EAOUT can keep stable, then the current sense threshold can be constant.  In the schematics, there is a SYMBOL - VOSNS_BK - at pin-1 of U3 (ucc28811D), It looks like a feedback from somewhere of an output, but I cannot find the same SYMBOL anywhere else in this schematic.  Could you explain this to me?

Best Regards,

Jun

Hi Richard,

Thank you very much for your description. I think I got it. It uses the "Maximum current sense threshold voltage" to limit the current, so both voltages on VINS and VSENSE are set to fixed level, and V_EAOUT can be high enough to generate enough current reference. It makes sense now.

Best Regards,

Jun