TIDM-02013: Huge explosion on the PCB

Hello,

I have looped in the expert from the team, and they should get back to you with a response in the next 1-2 days.

Best Regards,

Delaney

Hello Delaney

Thank you very much. I look forward to hearing from you as soon as possible due to the importance of the matter.

Best Regards.

Juan Carlos

Hi Juan,

While waiting for our HW design experts to comment further, could you please tell us when exactly this incident happened? This would help us to understand what went wrong.

The PTC is not to block the AC input voltage and  230V AC input could not be the reason for this incident.

1. May I know at what power did you run this experiment? What's the output voltage and Current?

2. How long after powering up this board did this incident occur?

3. Was the relay K1 connected parallel to this R165, R160 is closed prior to applying any load on the output?

Thanks

Srikanth

Hi Srikanth,

The PTC is acting because, in a transient state, when the board is initially connected, the DCLINK capacitors (C223, C224, and C225) act like a short circuit (low inpedance). The chokes (L16 and L17) also have very low impedance, so almost all input voltage falls across the PTCs until the relay is activated. Here schematic:

I have done a simulation with TINA and it confirms what I am saying:

As you can see in the simulation, until the Relay is activated, in the PTC there is a lot of voltage, more than 24V recommended by the manufacturer. So probably this is the cause of the fire.

To answer your questions:

1. The voltage that we are testing is 230V AC, and the DCLINK voltage was about 400V with a 500R load on the DCLINK.

2. After a few seconds.

3. The relay was probably not activated, since our firmware didn't activate the relay until the TTPLPFC_VAC voltage was above 200VAC.

This is how our FW was when the board exploded.

I await your news as it is very urgent.

Regards.

Juan Carlos

hello Brent,

I don't know if I explained myself correctly, but during the board's startup, there was no charge on the DCLINK. The current that burned the PTCs was the current generated when the DCLINK capacitors were charged through the PTC (no load on DC LINK connector).

We connected the 500R load after the relay had already been activated during normal testing.

I'm attaching a screenshot of the osciloscope before the explosion:

As you can see, the signal on CH1 is the input voltage (230 VAC).

The signal on CH2 is the voltage on the DCLINK that is charging without any load on the DCLINK. If there had been a load, this signal would not be ramping up.

The signal on CH3 is the input current. As you can see, it is very low, with only current spikes of a few mS during charging of the capacitors.

The signal on CH4 corresponds to the relay activation. When this signal drops to a low level, the relay is activated.

It is clearly seen that during the charging of the DCLINK capacitors for a few seconds the voltage on the PTCs is much higher than 24VAC and this clearly damages these components.This is confirmed by both the simulations carried out in Tina Space and the oscilloscope captures.

Please tell us what happens to your board when it's powered by 230V and the relay doesn't activate until the DCLINK capacitors are charged (no load on the DC LINK). Can you please run this test for us please?

Note that previously in the experiments we performed earlier, the relay was activated instantly when the board was connected to 230 VAC and this triggered the magnetothermal and differential switches due to high current peaks, this is why the tests we performed involved delaying the relay activation to gently charge the capacitors through the PTCs but doing these tests the board exploded.

Regards

Juan Carlos

Hi Brent,

Thank you very much for the information. I have a few questions:

* Do you connect the board to some kind of grid simulator or directly to the grid (230 VAC)?

* Do you have a screenshot oscilloscope showing how long it takes the relay to activate once you connect the board to the grid?

* If you don't have a capture, can you tell me how long it takes for the relay to activate once you connect the board to 230 VAC?

Best regards.

Juan Carlos

Hi, Juan Carlos,

During my tests I connected the the input to an AC source.

Sorry, no I don't have this. The last time I checked the relay was activated based on the CCS command to start the PFC. It was not activated based on any voltage monitoring. The code could be modified to do this but I don't believe this was ever done.

Best regards,

Brent

Hi Brent,

Can you try measuring how much time takes between the board is powered up and the relay is activated? Would that be very helpful?

I still think that from the moment we power the board until the relay is on, during that time, the PTCs are energized with 230 Vac, as demonstrated by the attached simulations. Could you please give me your opinion on this?

Best regards,

Juan Carlos

Hi Brent,

Okay, I won't bother you any longer. Last question I just ask you to look at this simulation and give me your opinion. As you can see, the voltage across the PTC at the beginning is very high, as we have almost 600 Vpp when the PTC is actually 24 Vac. This could never work... dont you agree?

Best regards.

Juan Carlos