LM51772: Abnormal output voltage

Hi Ishiwata,

This is just pick up of noise due to the measurement setup (probe connection). Here is a better one. 

Best regards,

 Stefan

Hi Stefan,

I apologize for the late reply.
Thank you for your confirmation.

I have additional questions.

(1) Could you please explain the mechanism by which the LO1, LO2, HO1, and HO2 pins exceed their absolute maximum ratings?
I believe the cause of damage is as follows:
When restarting under no load, the inductor current is nearly zero, but a small amount of energy remains at the switching node.
As a result, an overshoot of Vsw= VIN + L (parasitic inductor) x (di/dt).
I believe that this overshoot, plus the remaining voltage at the output, causes a voltage exceeding the maximum rating to be applied to the SWx pin.
Is this theory correct? Any advice would be appreciated?

(2) The LM51772 has a function that allows the output voltage to be dynamically changed during operation.
Is there any concern that using this function could damage VCC2?
Customers may dynamically change Vout within the range of 5V to 48V.

Best Regards,
Ishiwata

Hi Ishiwata,

1) LOx and HOx abs max ratings violation would appear if VCC2 is overcharged.
    HOx abs max ratings violation can appear additionally if SWx has undershoots = overcharge of HBx

2) no, this is the intended function of the device. Can also be checked on the EVM.

Best regards,

 Stefan

Hi Stefan,

Thank you for your support.

Regarding (1)
Is my speculation correct?
Also, please tell me the cause and process of VCC2 being overcharged.

Regarding (2),
I understand that VCC2 will not be damaged.

Best Regards,
Ishiwata

Hi Ishiwata,

for (1) sorry was not accurate enough here with the previous answer.
An overcharge of VCC2 will not happen but and overcharge of HBx can happen due to the undershoots of SWx.

If some circuits attached to HBx are then damaged it can also impact VCC2 which would look like a damage of VCC2.

Note: HBx is supplied out of VCC2 (Boot Circuit)

Best regards,

 Stefan

Hi Stefan,

Thank you for your support.

I understand that the SWx undershoot is caused by parasitic inductance during startup. Is this my undersatand correct?

Also, let's assume that the SWx undershoot occurs when CBootx on the HBx pin is already charged.
Because CBootx cannot be charged, the charge has nowhere to go, which causes damage to VCC2 via the HBx pin. Is this my undersatand correct?

Best Regards,
Ishiwata

Hello Shuji,

I think you need to look even closer on VCC2. and measure at the same time the switch nodes. What you see there looks a lot like noise or the voltage change due to the current change on VCC2.

Best regards,
Brigitte

Hi Brigitte,

Thank you for your reply.

It is important to observe VCC2 closely, but checking will damage the device.
The customer will need to change the device each time. This inquiry is taking a lot of time.
Therefore, it is difficult to monitor the moment of damage.

I plan to use the information we currently have to deduce the cause of the damage and close the issue by explaining it to the customer.
To do this, I would like to confirm the validity of my theory about the cause of the damage.

Best Regards,
Ishiwata

Hi Shuji,

The experts are out of office and will be back on Jan 8th. Please expect a delayed response.

Best Regards,

Feng

Hi Feng,

Thank you for your reply.

I am waiting for your reply.

Thank you very much for this year. I look forward to working with you next year.

Best Regards,
Ishiwata

Hi Shuji,

Looking forward to working with you next year.

Best Regards,

Feng

Hello Shuji,

Can you please share with me what you have as your theory of cause for the damage?

Best regards,
Brigitte

Hi Brigitte,

Thank you for your reply.

In a previous post, it was discussed that one of the factors that could cause VCC2 to break is violation of the ABS maximum ratings LO1, LO2, HO1, and HO2.

I consider the following factors to be factors that could cause violation of the ABS maximum ratings LO1, LO2, HO1, and HO2:

(1)
When restarting under no load, the inductor current is nearly zero, but a small amount of energy remains at the switching node.
As a result, an overshoot of Vsw= VIN + L (parasitic inductor) x (di/dt).
I believe that this overshoot, plus the remaining voltage at the output, causes a voltage exceeding the maximum rating to be applied to the SWx pin.

(2)
This damage to VCC2 occurred when the power was turned on, then turned off, and then restarted.
After that, I investigated further possible causes.
I believe that if the power is restarted while there is still voltage at VOUT, the device will operate like a boost converter on VIN from the VOUT side, causing a backflow and causing LO1, LO2, HO1, and HO2 to exceed their maximum ratings.

Because it is damaged, I think it will be difficult to identify the cause.
I would like your opinion on whether my suspected cause is correct.

Also, please let me know if you have any other suspected causes besides my own.

Best Regards,
Ishiwata

Hello Shuji Ishiwata,

When the device is starting with pre-charged output, there should not be a damage to it. The device typically does not allow current to the input during startup.

I know that parasitics of the MOSFETs and the layout can cause high voltage spikes and therefore I suspect that these parasitics caused the damage.

Best regards,
Brigitte

Hi Brigitte,

Thank you for your reply and insight.
Can I understand that the (1) I mentioned is a possible cause of the damage?

Best Regards,
Ishiwata

Helly Shuji,

I think that this is possible.

Best regards,
Brigitte

Hi Brigitte,

I have an additional question.
If it is damage based on my speculation, under what conditions is it more likely to occur?
For example, does any advice exist, such as the higher the Vout, the greater the likelihood of damage, or whether it is more likely to occur in a step-up or step-down state?
I believe that the higher the Vout, the more likely damage is to occur, and that damage is more likely to occur when the circuit is started up in a no-load state.
Please also give me your advice on my view.

Best Regards,
Ishiwata

Hi Brigitte,

Thank you for your reply.

I have an additional question.

I have confirmed that some samples output the device's initial value of Vout = 12V upon startup, while others remain at Vout = 0V.

The E2E thread below states that if the VFB voltage reaches 2.3V or higher upon startup, Vout = 12V is output. Therefore, I believe that device variations may result in 12V and 0V outputs.

 LM51772: Voltage applied to the FB pin  

Is there a way to prevent the circuit from outputting the device's initial value of Vout = 12V upon restart?

Best Regards,
Ishiwata

Hello Shuji,

To be honest, I do not understand what you want to achieve and how. Are you applying a external voltage on the FB node? Is this voltage different on different boards?

Best regards,
Brigitte