TPS62868: some questions

Jeff YANG

Mastermind 7835 points

Part Number: TPS62868
Other Parts Discussed in Thread: USB2ANY

Tool/software:

Hi,

We have some quesiton about TPS628681

What are the recommended values for Phase Margin (PM) and Gain Margin (GM)?
What is the TPS62868 work mode? current mode, voltage mode or D-CAP mode?
During the I2C command pull-up process, we observed that the I2C voltage has two levels: 3.3V and 2.5V. Is this due to an internal mechanism? (The external I2C pull-up is 3.3V.)
(We use our own MCU to config)
Regarding the OTP function of the PS628681A, I used an external heat source along with a loader to draw 4A, but still could not trigger the OTP at 150°C. Even the surrounding solder has melted. Has this feature been modified?

Thanks!

Jeff

2 months ago

0 Jeff YANG 2 months ago

Mastermind 7835 points

Hi,

Add one quesiton,

Is there have excel calculation tool for gain/margin design evaluation?

Thanks!

Jeff

0 Jingyi 2 months ago in reply to Jeff YANG

TI__Intellectual 1610 points

Hi Jeff,

Thanks for usin ge2e.

Please let me check with the team.

BR,

Jingyi

0 Jingyi 2 months ago in reply to Jeff YANG

TI__Intellectual 1610 points

Hi Jeff,

The commonly recommended PM is 45 degree and GM is 10dB. But in the end the customer need to make the decision.
For this chip the work mode is DCS-control. Fixed-frequency DCS-Control: Fast transient response with clock synchronization
For this chip the I2C high level depends on the external pull-up voltage.
Regarding OTP, is it possible to provide a photo of the set up? It might be that the chip hits a thermal shutdown for only a few microseconds before the temperature drops again.
Unfortunately, there's no calculator currently. In datasheet there are some recommended values.

BR,

Jingyi

0 Jeff YANG 2 months ago in reply to Jingyi

Mastermind 7835 points

Hi Jingyi,

The commonly recommended PM is 45 degree and GM is 10dB. But in the end the customer need to make the decision.

Jeff: Are you mean GM -10dB?

Thanks!

Jeff

0 Jingyi 2 months ago in reply to Jeff YANG

TI__Intellectual 1610 points

Hi Jeff,

yes that's right.

BR,

Jingyi

0 Jeff YANG 2 months ago in reply to Jingyi

Mastermind 7835 points

Hi Jingyi,

For OTP test setting, we using a hot air gun set to 200°C to directly heat the IC did not trigger the OVP function. However, it was observed that the voltage initially increased when heating started, then returned to 0.6V afterward.

Or it just EVM GND plan too big cause it hard to reach the OTP point?

Thanks!

Jeff

0 Jingyi 2 months ago in reply to Jeff YANG

TI__Intellectual 1610 points

Hi Jeff,

You're right. It's difficult to heat up the chip to over 150C with the set up. Most heat will dissipate before reaching the EVM and through the EVM.

In our test environment, chips are put into an oven to be heated up.

BR,

Jingyi

0 Jeff YANG 2 months ago in reply to Jingyi

Mastermind 7835 points

Hi Jingyi,

May I ask what temperature the oven was set to? Was there any additional heating applied, or were there any load conditions we should be aware of?

Thanks!

Jeff

0 Jingyi 2 months ago in reply to Jeff YANG

TI__Intellectual 1610 points

Hi Jeff,

Usually it's set at 150C. No additional heating was applied.

You can also check the Thermal Warning register as in page 21 in datasheet.

BR,

Jingyi

0 Jeff YANG 2 months ago in reply to Jingyi

Mastermind 7835 points

Hi Jingyi,

1. Customer said they don't have oven can do the OTP test, do you have any OTP test report or test video can share to me?

2. We found there have some glitch on I2C signal. We also tried only connect only one of I2C SCL or SDA probe to make sure they won't be impact by other probe. It seems not trigger low level to make communication fail, but just want clarify it is also can found on your EVM on hand?

Thanks!

Jeff

0 Andreas Mueller 2 months ago

TI__Mastermind 22210 points

Hello Jeff,

I'm helping out for Jingyi.

1. I found some existing test results of a Thermal Shutdown for Vin=3.3V, Vout=1.15V and Iout=2A resistive load.

Thermal shutdown entry:

Thermal shutdown exit:

2. I will check for I2C spikes tomorrow, as I have to prepare the setup first. Then I will send the results.

Best regards,

Andreas.

0 Andreas Mueller 2 months ago

TI__Mastermind 22210 points

Hello Jeff,

I checked the I2C signals at our EVM, connected to an USB2ANY interface, and I can confirm the spikes.

It seems to be some cross-talk between SCL and SDA.

Then I tried to trace back the source of the spikes and I identified the USB2ANY interface as the source.

When no EVM is connected then the USB2ANY signal is still showing the same spikes:

Best regards,

Andreas.

0 Jeff YANG 2 months ago in reply to Andreas Mueller

Mastermind 7835 points

Hi Andreas,

The cross-talk between SCL and SDA is impacted by EVM board or IC internal?

Is it impact I2C performance?

Thanks!

Jeff

0 Andreas Mueller 2 months ago

TI__Mastermind 22210 points

Hello Jeff,

I did not see an impact from the EVM.

The observed I²C spikes were caused by the I2C controller (USB2ANY).

I did not find any I²C performance issues.

Best regards,

Andreas.

0 Jeff YANG 2 months ago in reply to Andreas Mueller

Mastermind 7835 points

Hi Andreas,

Our customer found this issue by their own MCU, they didn'y used our USB2ANY.

Is It may casued by EVM layout or IC internal?

Thanks!

Jeff

0 Andreas Mueller 2 months ago

TI__Mastermind 22210 points

Jeff,

I think it is not caused by the EVM layout. I think it is caused by the I²C controller device, in the customer case it is most likely caused by their MCU.

The customer could check the I2C signals when no EVM is connected to confirm.

Best regards,

Andreas.