Viper16LD problem

[GArna]

Hello,

We are facing a problem with the use of Viper16LD.

My schemtic is like following:

 

We already produce over 100 board without problem.

Now, on a new batch, we have a problem.

The input vlotage named "DCBus" start at 220*1.41 = 325V.

With that, Viper output is correct (about 16V)

After that, we start a PFC which increase DCBus up to 400V.

But when DCBus over pass about 360V, the Viper output 16V fall!

PLease find a Scope screen:

 

Could you have a check or Ideai why problem occurs?

Regards

[Peter BENSCH]

Welcome, @GArna​, to the community!

Due to the high input voltage and the relatively small output voltage, the duty cycle must be very small. Probably your load current is also a little too small, so that the already very short switching pulse is no longer sufficient to supply the VIPER with the necessary energy.

Please try to increase the load current a little bit to get more stability. At 400V and 16V you should be able to go to about 200mA.

Regards

/Peter

[GArna]

HEllo

I tried to load the Viper with only a 110 ohm (16V/110ohm = 145mA)

Now, the viper not start even at 325V.

It start only when input fall under 300V...

[Peter BENSCH]

Did you run a simulation of the circuit?

The layout also plays an important role.

If necessary, contact the Online Support OLS for personal support with the problem.

Regards

/Peter

[GArna]

Yes we run a simulation.

I note some little differences:

Rc/Cc. Simuation is 6k8/1µF. My board is 48KHz/1nF. But with simulation value, the viper not work aith lower DCBus Value (about 320V) where work up to 360V with my board Value.

Rh/Rl value (to FB pin). THe ratio is the same, but the range are two time lower on my side: 94K/24.3K for the simulation, 47K/11K on my side. There is an importance on the current into this circuit?

ILim: Try with Not connected pin: Problem occurs wathever the input voltage. Set only a 330nF this work up to 360V. Add 47K Rlim. Same behavior: Not work over 360V input voltage.

WHat could be the reason the circuit "Fault" is triggered? What signal Can I read?

I will try to contact personal support also.

For the PCB, the first batch of 150 board work without any problem. THe only difference now is the PCB provider. THe gerber are the same.

[GArna]

Hello

We solve the problem By replace C48 (FB decoupling capa) from 100pF to 1nF.

Is this modification seems acceptale without cause any other problems?

Regards

[Peter BENSCH]

I would rather expect that C48 would have to be reduced to zero to improve phase margin, but if that works for you, it shouldn't be a problem.

Good luck!

/Peter