TPS92314A

Hello,

Sorry for the delay,

It sounds like the converter doesn't want to start.  I would look at the bias voltage first, there may not be enough energy storage in the start-up/bias capacitor to get the power supply running at a high enough voltage to run itself.  Look at the bias voltage to the TPS92314A and see if it is dropping to UVLO and it tries to restart.  Another check would be to increase the bias capacitor value and test if it starts at a lower voltage.

Dear Irwin, Need your advise on the following problem. We are trying to parallel 2 TPS92314 based LED drivers each with 80W output to achieve 160w total output load. The LED load is 60V/2.4A. The outputs of the Driver units are connected +ve to +ve and -ve to -ve. Individually each driver works fine outputting 80W each. When connected in parallel one them works fine while the other one goes on and off. ie the waveform at the Mosfet drain is normal for a fraction of a second and goes off  the next second. Obviously the LEDs blink and out put current fluctuates between  1.2 A and 1 .8Amp. The input mains is common for the drivers and the outputs are connected in parallel after the smoothing capacitors. Any suggestions.

Regards,

S.Ramakrishnan

Hello,

I don't know this part too well so I'm taking a guess. The TPS92314 is a primary side regulated supply. Internally it has a start-up sequence and may be detecting something it deems a fault. Are these complete separate supplies generating their own bias voltages etc? What I would try is adding another two diodes after the smoothing capacitors so they both have to start with zero output voltage.

Do they both exhibit this problem. If you start one TPS92314 does the second one powering up have issues. If you switch the start-up order does it switch which converter exhibits this?

Thanks,

Dear Irwin, Thanks for your prompt inputs. You are right. TPS92314 is a PSR Constant current regulator with the starting supply from Tank capacitor after the mains bridge rectifier.  once it starts the supply is taken over by the Aux winding in the SMPS transformer.we have used common mains rectifier and EMC filters  with 2 nos tank capacitors.(0.33uF/630V) across the bridge output. from there the two drivers are independent both adjusted for 60v/1.!Amp. The outputs of the drivers are paired across the  Elco caps at the output so as to deliver 60V/2.4Amp total output. Individually they are working fine. But the moment  you club the outputs one of the drivers start 'blinking'- the output keeps swinging from zero to 1Amp. You suggested to put two diodes--in the starting supply to the controller? Please come back.

We have used this method of clubbing with other solutions like Power Integration, TOREX etc and they are working fine I would be grateful if you can suggest some leads to work on!  I can forward the schematic  if it will help.

Regards,

S.Ramakrishnan

Dear Irwin, Thanks for the clarificstion.we will try out the idea with appropriately Rated diodes and come back.
In the meantime there is another issue bothering us. The TPS 92314A drivers work well between 90V and 300V AC input with excellent line regulation and output regulation better than 3%. The problem is the output is kept constant even when the input voltage drops to 50V. This causes the input current to rise , the efficiency drops and input power rises disproportionately overheating the Power MOSFET and line Fusible resistor. Effectively the ON time exceeds 50% and continuous operation leads to failure. Is there some way to prevent the ON time increasing beyond a point - the output current should be allowed to fall and taper off at voltages lower than 90V AC ? Please suggest.
Regards,
S.Ranakrishnan

Hello,

This is common for a lot of offline flyback converters.  You may be able to increase the current sense resistor value so it enters current limit at higher input voltages.  This may still not be enough to solve the issues.  Or you may want to add an undervoltage lockout circuit that has hysteresis.

Dear Irwin, Thanks. We can not increase the Current sense resistor since it is linked to the LED output current. Current limitation through the current sense resistor happens during short circuit conditions. Ofcourse low input cutoff circuit similar to the one we shared with you can be used but at some additional cost and of course board area. I was wondering if you could come up with some trick/idea to stop the TPS 92314 from regulating the output current below say 90V AC.It has been a wonderful experience sounding you out on such issues and I am grateful for the time you are giving to finding solutions for us.

Regards,

S.Ramakrishnan 

I looked again at the datasheet.  Anything that I would come up with would add parts.  The on time is internal to the IC, even if it were external it would require components to limit it.

Thanks,

Irwin

Dear Irwin, Thanks for the prompt reply. yes you may have to add some external circuitry to kill this problem. This may be a better approach than adding the more elaborate low voltage cutoff circuitry. I will be grateful if you can come up with some idea on how this can be done on the basic application circuit adding a few components. we will take care of the pcb layout once the concept is clear and validated. we are currently using a 4.7uF ceramic cap at the compensation pin. The potential at this pin is normally 2V from 100V to 300V AC input and start to go up as the input dips and reaches around 2.4V when the AC in is around 65V. Apart from this the Min. frequency is around 60Khz at 100V and drops to around 40Khz at 65V AC input.the input power jumps almost by 25 to 35% from 100V to 60V AC input.. Hope all this additional observations help. If you need any specific inputs further please let me know.

Regards,

S.Ramakrishnan