TPS82130: behavior at start up and low lord current

Hi Fujiwara-san,

From our previous thread, the load current is very small, only 50mA or 3.3V/0.05A = 66ohm load, so the additional 1ohm resistor is quite small relative to the load. The second waveform is more what I would expect for the current.

Is it possible for the customer to try lower series resistor values to see impact on the measured current, for example, 0.1ohm and 0.5ohm?

Kris

Hi Kris-san,

Customer took measurement by adding 0.22ohm resistor. Please see attached file, figure 4. From measurement result, it seems low phase margin of TPS82130 is affect. Do you have any idea to solve this phenomenon? Note that this phenomenon has caused when output voltage set 3.3V. Customer never saw this phenomenon at output voltage is 1.8V.

TPS82130_question(02192020).xlsx

Best Regards,
Fujiwara

Fujiwara-san,

I'm assuming EVM is not modified (other than feedback divider for 3.3Vout). What is the 3.3Vout driving on the customer board? What is the capacitance on the board?

In the end application, will the TPS82130 be on a separate board connected by this 5cm cable to another board?

Hi Kris-san,

Yes, your assuming is correct. The load of TPS82130EVM(3.3V output) is customer board. The load capacitance is around 100uF. Customer board load is at same board of TPS82130.

Best Regards,
Fujiwara

Hi Fujiwara-san,

Thanks for the information. 100uF load capacitor should not be an issue, but what else is on the customer board? Is it an LDO or other converter that is powered from the 3.3V?

I will be on business travel until Thursday, but when I return I can try to duplicate the setup and take some measurements as well.

Thanks,
Kris

Hi Kris-san,

The load is FPGA and SSD, so it does not include LDO and DCDC converters. Please see attached file which is showed customer board block diagram. Would you please try to duplicate customer's phenomenon?

Best Regards,
Fujiwara

TPS82130_question(02252020).xlsx

Fujiwara-san, I am just getting back into the office and need some time to look into this. I will give an update early next week. Feel free to email me directly with any other details if needed. Appreciate your patience!

Fujiwara-san,

I took some measurements on the TPS82130EVM modified for 3.3V output today.

Conditions are not exactly same as customer setup, but wanted to share what I observed so far.

Below waveform is with output connected with about 18 inch (45cm) cable to electronic load but no load applied:

I do not see any reverse current flow back to TPS82130 EVM.

As a second test, I connected 3.3V output with same length cable to input of a 2nd TPS82130 which has 68uF + 10uF input capacitor, but with the load EVM disabled (using EN pin header). I also do not see any reverse current.

In the customer's setup, are the 5cm cables twisted together to minimize the loop area?
Are the load circuits on the customer board enabled or disabled?

Kris

Krus-san,

Customer mentioned that your measurement results shows negative current flow as attached file. How do you think?

TPS82130(03042020).xlsx


In the customer's setup, are the 5cm cables twisted together to minimize the loop area?
=> No.

Are the load circuits on the customer board enabled or disabled?
=> The load is enable but load current does not flow so much current.


Best Regards,
Fujiwara

Fujiwara-san,

I do not see an issue in the results. Please help me understand the customer's concerns. Are they seeing unacceptable voltage ripple or other undesirable performance at the TPS82130 output on the EVM or at the load on their board?

Let me email you directly and we can discuss offline as needed.