TPS61028: TPS61028 Stability

Hi Pengfei,

Thanks for reaching out. Please kindly see my reply below.

1. What about the phase and gain margin of the converter’s feedback loop? Not sure if there is stability issue for the current design. could you kindly help to simulate?

[WW]I can simulate with the TPS61020, there is no TPS61028 model available in library. I am also checking if TPS61028 sample is available in lab.

1. If necessary, could you kindly help to optimize the part selection if stability or phase margin is not good?

[WW]I would suggest you use a bulk capacitor with 47-100uF capacitance and >30mOhm ESR. There is a 2.2uF ceramic capacitor recommended place close to IC to minimize switching stress and noise. I do not think 7Ohm is ok here as this is too large and cause the capacitor to be invalid for higher frequency zone.

1. Since Tantalum is proposed to used, what kinds level of ESR must be met? and if Tantalum and MLCC is parallel to guarantee low ripple voltage, is there capacitance requirement for both capacitor?

[WW]30mOhm is ok, I tested with TPS61020 before with 100mohm and it also works. Capacitance range is between 47uF and 100uF.

-Wenhao

Hi,

Could you kindly help to simulate (TINA or PSPICE) with TPS61020? about the phase margin.

I have used 22uF/4ohm ESR tantalum and 32.4uF MLCC together on prototype, it works well for output voltage and ripple.

so could you help to check what is the required ESR range (min to max) for the stability?

as for the tantalum capacitance, understand the recommend value is listed, if tantalum capacitance is less than 47uF, both total capacitance is higher than 47uF with Tantalum + MLCC, what will be the problem?

Thanks a lot.

Hi Pengfei,

The model is PSPICE. The recommend range is 47-100uF/30mOhm. If you insist to use the capacitor configuration on your own, I can do the simulation. Bode plot needs to be done based on bench. Need some time to check sample availiability.

-Wenhao

Appreciate so much, could you help to check stability for the cases below?

1. 10uF C202 T491A106K016ATAUTO, 7 ohm ESR@250C 100kHz) in parallel with some MLCC(32.4uF totally in the picture above) 

2. 47uF MLCC in series with 0.1ohm resistor (or what is the more suitable resistance depending on your simulation?)

Let me correct something. for case 2.

And are there anything need to be optimized for the current circuit? thanks a lot.

Hi Wenhao, could you kindly help to run PSPICE simulation to verify stability firstly?

Regarding the Bode plot test, it can be double verified later.

Thanks a lot.

Hi Pengfei,

I need some time to do simulation, and applying samples for bench test. I cannot give you more information besides what I have posted above. I will give the feedbacks after CNY holidays.

-Wenhao

Hi Wenhao,

Appreciate so much for your kind help.

I have tantalum capacitors in hand, do you think is it necessary to give them to you to run bench test?

It might speed up the test.

Hi Pengfei,

Do you have your local FAE to deliver the capacitor?

-Wenhao

Hi Wenhao, if you are not in China, it seems delivering is not convenient. is it possible to run simulation firstly?

Thanks a lot.

Hi Peifei,

I use TPS61028 using TPS61020 PSPICE model. The model is ok to be used to simulate TPS61020 as they are only current limit threshold different. 

 Could you please help confirm for MLCC shall I use its effective capacitance or just the rated value? Below is the capacitors you proposed above in total. 

-Wenhao

Hi Wenhao,

Appreciate so much.

For MLCC, rated value is OK.

Let me explain clearly, in your circuit, C11 is not necessary for case #1.

case #2, C11 will be in series with 0.1ohm resistor.

Let us see if there will be some problems.

Hi Pengfei,

Both condition is unstable. Please see attached report for reference. If it ok, I will do the simulation using recommended capacitors for your reference. Thanks!

TPS61020 Simulation.pptx

-Wenhao

Hi Wenhao,

Thank you,

If keep all the 32.4uF MLCC capacitors in my circuit, please help to propose Tantalum or others.

Hi Wenhao,

By the way, I have tested T491A226K010AT(3.2 Ohms ESR) under -40C and 85C, the output voltage 5V is always stable and ripple is very small (less than 10mV). could you also help to verify the stability? not sure if simulation will match with the test result.

Hi Pengfei,

After several simulations, I found FB dividers' resistance will influence loop stability. See below description in datasheet. It suggests the resistance will influence internal compensation circuitry. Therefore, would you mind me changing value of FB resistors? Also your resistance is too low and may rise effective quiescent current.

-Wenhao

Hi Wenhao,

I much more agree with changing resistance comparing with changing Tantalum capacitor, for the current design, 32.4uF MLCC and T491A226K010AT(3.2 Ohms ESR) are used together.

could you help to try if change Tantalum capacitor ESR (0.1ohm to 10ohm), but not change tantalum capacitance, and if stability is problem, try to adjust the resistance. let us try to make it stability.

I do not know why the test waveform is stable, but it seems simulation is not.

Appreciate so much.

Hi Pengfei,

I will do the further analysis and simulations after Chinese New year holiday. 

-Wenhao

Hi Wenhao,

Appreciate so much.

I have noticed there is TPS61020 average module (and simulation example) for TI-TINA tool, which is also for AC simulation, not sure if it could be also good reference comparing with PSPICE simulation?

Based on my lab test on prototype, under low and high temperature, my first attached circuit works well, but I just would like to confirm the simulation result, and by the way, it will be helpful to better understand the stability for future design with TI parts?

Let us continue after holiday.

Happy Spring Festival Holiday!

-Pengfei

Hi Wenhao,

Not sure do you have time to deep investigate this issue?

By the way, about the actual test, not sure did you request the sample to plan the test?

Thanks a lot.

Hi Pengfei,

Sorry. I currently do not have chance to do this test. Will do this ASAP. Thank you.

-Wenhao

Hi Wenhao,

Appreciate so much, not sure is it possible to let us know the result in one week?

Hi Wenhao,

Appreciate your test.

Could you help to check what will be the result or issue for this case?

1. Load current range 50mA to 250mA

2. R240 to 100K, R244+R32 to 11.11k

3. Tantalum capacitor: C202, 33uF/0.5ohm ESR (TPSB336K016R0350)

4. Other MLCC no change

If it will be failed, is it possible to change other components but not change resistance?

-Pengfei

Hi Pengfei,

The loop of TPS61020 should be same as TPS61028. I do not see necessity in changing resistor lower, as what I have already simulated. Lower resistance will result in loop instability;

I need to apply this Tantalum capacitor.  You said you can change resistor but you do not want to change capacitor, why you say now you do not want to change resistor? Right now the loop is already stable, could you please tell me what do you want for test?

-Wenhao

Hi Wenhao,

The components have been prepared, if change is must, in general, it will be easier to find alternative resistors.

Thanks a lot for your understanding.

Hi Pengfei,

I have applied the capacitor (TPSB336K016R0350) from mouser. Will update here if results come out.

-Wenhao

Thank you very much, Wenhao.

Hi Wenhao, Not sure have you received parts, and could you help to test it?

Thanks a lot.

Hi Pengfei,

We are waiting for mouser to delivery the package. I am checking the states with them.

-Wenhao

Hi Pengfei,

To help increase efficiency, I would like to do this in parallel. I see the package is pending approval. Could you please help ship the cap to me directly if you are at China?

-Wenhao

Hi Wenhao,

Currently the parts can not be shipped out because of COVID, it might take some days to ship in SH. not sure if you have sample parts received, 

Thanks a lot.

Regards,

Pengfei

Hi Pengfei,

Yes you are right. We are all now in Quarantine and are not allowed to go outdoors. I would expect to go lab after next two weeks.

-Wenhao

Hi Wenhao,

Understand, not sure how can I contact by your mail?

Regards,

Pengfei

Hi Pengfei,

Please contact your FAE first. I see he sent an E-mail to me before.

-Wenhao

Sure, thanks a lot.