TPS5410: TPS5410 output waveform problem

what is  "tip adn barrel method"?  , i only use oscilloscope to measure the voltage , then get this wave.

so if is the high frequency noise , how can i fix that ??? please tell me thx

Hello,

Please see the image below of tip and barrel:

This is a way to measure with the scope probe.

If it is high frequency ringing, you want to optimize your input capacitor loop.  You can also add a high frequency bypass capacitor directly next to the input and/or at the output as well.

Best Regards,

Katelyn

hello,dear Katelyn ,i use tip to measure the waveform,
but some question i need to ask , i don't the guide of tps5410 with pole .
i don't know what frequent cal on the user guide of tps5410,
the frequent mean what? i only know the resistor and cap will effect the frequency,
but user guide didn't tell us what frequency mean.
so, i don't know how to choice the cap with input cap.
such like u say if i change the input cap, i will change the cap C11, and i don't know what value to use.
because i design the this circuit i will use for electric tool circuit.so it maybe will connect the mcu.
so the cap value will not to high or low it will effective the electric tool.
maybe u can give me the suggestion .because i trust u r the expert of power .

and this problem is confuse me many day , could u reply me immediately .

because i need to solved  this problem in short time.

my big problem is i need to solved the noise .

and i need to reduce the ripple of voltage to 3~5% because i need to past the emi/emc standard.

so this power circuit is very important.if it could't ,we will give up to use for tps5410.

Hello,

This document covers this in detail.  

Measuring_Output_Noise.pdf

Please review this document about high frequency noise.  If you have any questions, please let me know.

The bypass capacitor can be about .1uF.

Best Regards,

Katelyn

Dear Katelyn Wiggenhorn , first thx u so much

as u say use another measure could reduce the Output Noise.
but i have some problem need to ask.
if i change the measure method that mean like i am test the feature of Oscilloscope.
not like to solved the high frequency problem.I need to resolved by hardware circuit
and as u supply file of document.It also helpful but it look like that meany problem in layout
build the high frequency and noise. such like u say, i need to reduce the HF problem.
if I could , i think eliminate the oscillator of high frequency is the almost method to solved.
so can u give me the more method to solved?

and I am compare the EVM board and my design board. when I measure and i don't use
tip and barrel to measure and I didn't see high frequency like my post first picture.like
green part 1.8Mhz.

but i design board if i don't add R-load on 3V3D and 3v3Gnd the high frequency dose not appear.
but if i and the low R-load (maybe 10ohm~30ohm) it will appear the 1.8mHZ.
i have test for 10ohm,30ohm,and 330ohm, because my circuit R-load near to 330ohm
so I add the 330ohm on my circuit.but it still can't resolved this problem . it appear on 330ohm for high frequency.
as the same value of resistance on EVM . it also not appear.

so this is layout problem?

this is my layout picture, as u document file say that the input cap and inductor distance maybe cause HF noise.

it is 1.8mhz build reason???  So Other than this it only has improve layout to reduce noise  or it

maybe has another method to  reduce such change or add some circuit componet to solved?

Hello,

In regards to the measurement method, you want to understand the true level of noise.  The noise may be there, but it is important to understand the true level that you are trying to filter out.  With a bad measurement technique, noise can couple to the probe and amplify the noise of the design.

Also, how are you adding the resistor?  Is this directly on the board or through long wires?

I think it is very possible this is a layout problem.

Please see my recommendations for your layout:

1. Move C11 as close as possible to the device.

2. Add a ground plane across the entire bottom layer of the board and entire top layer of your board.  This will improve the return path of all signals.  And will also provide an improved design for dissipating heat.

3.The inductor, boot capacitor, and diode need to be in as small as loop as possible.

4. Move the output capacitors directly next to the inductor.

5.  Follow the example below for routing your resistor divider.

Best Regards,

Katelyn

Hello,

I want to follow up with you on this thread. Do you have any additional questions for me?

Thank you,
Katelyn

Hello , recently I am busy , so i didn't reply at first timed.

before u reply me from Nov 16, 2018 1:32 AM.

i have test the 2 vesion of circuit.

like the picture

the v1 is input cap and inductor close to tps5410

the v2 is input cap , output cap inductor close to tps5410

then i compare this two board. the wave is similar.and the high frequency noise is reduce.

so i think that the almost problem is like u say it maybe is inductor and input cap position build.

it it not related for output cap. so maybe i design the circuit i will ignore the output cap position.

such like eliminate. it almost about 70~80mV.

then i compare evm board that i find the evm board is same too. 

so i think the high frequency is resolved a part of but not perfect.

then i will test the circuit by add R-load .

then the result is like this.

if add the Light load resistance (330ohm,the power is output 0.019A) like this , the evm board and my v1 board waveform is like this

that seems it is perfect and OK.

but if add the Over load(which R is i think high load, high current 30ohm,the power is output 0.79A)

please ignore the measure method,but the evm result is the same result so i think my circuit is almost perfect for compare to evm board.

this result is normally?

the problem is confuse me,and seems high challenge to eliminate the high frequency noise,such like top question for yellow part and red part. 

so  i need u help. 

by the way. maybe we can make friend.

my presonal e-mail is Cnyntg927@gmail.com.

Sorry,I forget post schmeatic

please look Component J5 that is where i add load

Hello,

Thank you for sharing this additonal information. Can you please clarify which board/design you are using for both scope waveforms.

I think that is important that you place the load as close as possible to the output (eliminate the wires as much as possible), and measure the output ripple at the output capacitor with the tip and barrel method. If you use the ground loop, this can pick up different noise at different times. In order to get a consistent measurement for accurate comparison, I recommend using the tip and barrel method at the output capacitor.

Also, when you spin this board, you will definitely want to pay attention to the input capacitor and output inductor. I would recommend that the output capacitor can be third priority, but you do not want to let this capacitor move very far in distance from the inductor.

Best Regards,
Katelyn

OK,

Because i think the problem is a part of resolved .

so,i didn't track to find the perfect resolved method.

but Dear Katelyn Wiggenhorn, u help me much better.

thx very much.