I'm designing an isolated flyback dc-dc converter.
My specs are 20-70 V input, 24V output @ 3A.
I'm mostly done with schematic capture and layout, but have one issue to address before I finish and build boards.
One requirement I have for my project is that the circuit go into hiccup mode on persistent overcurrent condition. By hiccup mode, I mean that the output drops to zero, and the unit shuts down for a bit before restarting. If the overcurrent condition persists, then the unit shuts down again, comes back, shuts down, etc....
I am using an auxiliary winding on my transformer to produce 12V for the circuit to run off, which shuts down the integrated linear regulator and reduces power dissipation. I think I can accomplish hiccup mode by sizing capacitor on auxiliary winding correctly and capacitor on softstart pin.
Does anyone know how to do hiccup mode? Can you point me in the right direction
David,
That should work, but you may have to experiment with the VIN capacitor size some to get the effect you want. I've had that happen by accident in some past designs and had to play around with cap size to prevent it from happening.
Regards,
Ron Crews
Do you mean VCC capacitor?
Is there a part that would be better suited for this application?
Yes, I meant the VCC capacitor. The safer way to do hiccup involves more parts, but gives you a lot better control.
Another method is to connect the current limit in the standard way, but add a comparator with the positive input going to the sense resistor and the negative input to a reference set slightly higher than the current limit threshold. Connect the output without a pullup to the UVLO pin. When there is a current limit event, especially a hard event the comparator will be triggered and pull the UVLO low, turning off the IC. Add an appropriate capacitor to the UVLO input to ground. Then when the current limit event is over, the upper UVLO resistor will change the capacitor with an appropriate time constant and turn the IC back on.
Great idea. I like direct methods with good control.
I'm running into space constraints. Does TI offer a PWM IC with this function integrated into it that will work for this application? 20-70V in, 24V out, 3A.
No other part comes to mind given the higher input voltage and the topology of a flyback.
Hi Ronald,
LM5020 have pulse by pulse current limit. When over current happened, internal fet (driven by controller logic) discharge soft start cap and controller (LM5020) shut down for at least 50ns and after that soft start and....... circle is close. If you wont longer "off time" simplest way is to increase value of "soft start" cap. "Hiccup " mode (low ON/OFF ratio based on long soft start time) works better if Vcc is present and stable all time.
Dragan
Just be careful that you part will start under full load at worst case VIN with a long soft start time. That can be a problem depending on load and soft start time.
Ron
What do you mean this comment?
Do you mean that the startup regulator will be at full load during softstart and will be dissipative?
What can happen with a long soft start time is between the charging current into the output capacitance and the load requirement is the output will hang and never reach full voltage because the combined load along with the narrow pulses caused by soft start will allow the part to go into current limit before the output can reach regulation. It may not be an issue in your case but be sure and check. I just had that problem on a different part with a customer that had a need for a very long soft start time.