Technical Assistance required for a SMPS flyback using LNK6763K IC

Hi Brosnan Gomes,

As I was checking your PI Expert design, there are some optimization errors to be fixed.

1. Peak primary current exceeds device current limit.
2. Maximum flux density limit exceeded.

What device are you using?
You might want to increase the VOR of the transformer by clicking the design results. Then transformer construction parameters, the gray highlighted can be configured.

Regards,
PI-Pike

Thank you for replying Pike.
We are working with LNK6763K
We have made the circuit as described in the Application Note RDR321 using this IC and a transformer with PQ2620 core, the winding is done as described in the AP just the core is changed from EF25 to the one mentioned before.
When we power the circuit we get a fluctuating voltage output at no load, also this fluctuation was around 8V whereas the desired output was 18V. Upon checking the transformer secondary waveform on the oscilloscope we notice the IC auto restart since the pulses seen were flickering on the screen at intervals of around 1.7s.
But we notice a strange behavior on, as the AC supply to the board was turned off the flyback input bulk capacitor would discharge rapidly from 300Vdc to 200Vdc and after that form that point the required waveform were seen clearly at the output of transformer and thus the output capacitor stays at a stable voltage around 21V at no load and eventually drops as the bulk keeps on discharging slowly.
Can you help us with whats going on here?

Thanks and Regards,
Brosnan

Hi Brosnan Gomes,

Have you check the voltage on the FB pin? It might be dropping below the threshold.

Check also if you have enough voltage on the BP capacitor.

Regards,
PI-Pike

Regards,
PI-Pike

Yes I checked the voltage on the BP capacitor and it was at around 5.7V. Have used a 4.7uF over there and so the IC goes into auto restart as the voltage had dropped below the minimum requirement of 6.1V on it.
But after the AC supply is turned off and the bulk cap starts discharging as mentioned earlier this BP cap attains a voltage of 6.3V and the system gives a stable output.

What's the voltage across C10 - bias cap? Try increasing the R8 - bias resistor. It might be that the current is to high triggering the shutdown threshold current on the BP.

Regards,
PI-Pike

Hi Pike,

We have made a circuit which is given in rdr321 AP. We have not made the one given in the PI designer software since we wanted the circuit to work as given in the AP and afters change it to our required specs.
The C10 cap you asked about had a voltage of 5.75V which is same as the BP cap, but again after AC is cut off the C10 cap attains a voltage of around 11.2V.
The R8 resistor value we used is 6.8K instead of the 7.5K recommended in the AP.

Thanks

Hi Brosnan Gomes,

How about the layout? Is it the same with the rdr321?

I think the problem is on the feedback side of the circuit.

Could you upload a picture of your board, top & bottom side? Thank you!

Regards,
PI-Pike

Hi PIke,
I'm attaching the pics here.

Hi Brosnan Gomes,

Could you give me the AR_ON & AR_OFF Times when the fault is occuring?

You could adjust the resistor R8 - Bypass resistor. I think it triggers the AR Fault due to BP Pin shutdown threshold current.

Regards,
PI-Pike

Hi Pike,

We had tried changing that resistor to 10K, but the circuit was behaving the same.
AR_OFF is around 1.7s and AR_ON is around the time it takes to blink an eye.
I'm including here a drive link for your refernece to check the AR timings as well.

https://drive.google.com/file/d/1kh2-EfD3AXCY1vN-MAvEp3x9xKjJft8s/view
The waveform is of the secondary side of the transformer.

I want to mention another thing that we tried to change the resistor divider voltage by changing the R9 resistor (as shown in rdr321 schematic) to a potentiometer to change the resistance dynamically while the circuit is powered. We saw that a particular resistance value the flyback starts operating fine and we get a stable 5V at the first output (where we required 18V actually) even under loaded condition.

Thanks

You increased/decreased R9? when you get the stable 5V?

I think you should increase/decrease both the feedback resistors, keeping its ratio.

Thank you!

Regards,
PI-Pike

Hi Pike,
Really sorry for the late reply.
We tested the resister divider values change influence keeping the same ratio but were getting the same output. But in that process we noticed the higher voltage secondary schottkey diode was getting short, it happened multiple times. The diode we were using was DSS13UTR with 30V reverse voltage rating. We then changed this diode to BAS21GW which is a 200V rated diode, its not a schottkey but is fast switching. After this change the circuit was giving stable output and seemed to be working. Can help us here with what might have gone wrong there?

Also another thing about the output voltage is that we didn't insert a loading resister in the circuit as you might have seen in our layout pics. So at no load the circuit shows a higher voltage but when we load it the voltage level reduces to a particular value, and this voltage remains constant at any load condition. What we inferred from this was that this circuit might require a loading resistor to work accurately, we had also seen the Application Note having loading resistors in its schematic. So are they absolutely required if yes then what kind of resistor values should we put there or can we make the circuit give constant voltage at no load too.

Thanks

Hi Brosnan Gomes,

Could be that voltage rating damage the schottky?

Thank you,
PI-Pike

Hi Pike,

Seems like that since we were changing the o/p voltage by changing the feedback resistor ratios and at some point the output must have reached a voltage level which shorted the diode. But how does a lower voltage rating schottkey affect the normal operation of the LNK IC. As it started to work after the voltage rating of the output diode was changed to 200V. Or could it be the forward voltage? Since that too is different for both diodes.

Thanks

Hi Brosnan Gomes,

It should not affect the LNK IC, it depends also on the transformer primary to secondary ratio. At first when using the lower voltage rated diode it doesn't work, AR occured, thus at some point in time this will damage the schottky, depends on the voltage/current AR_ON pulse.

For the voltage rating of the schottky diode you should also consider the leading edge spike of the drain voltage, for it not to get damaged.

Regards,
PI-Pike