TPS61070 Operation

You are likely using way too much output capacitance.  The datasheet says that the Cout should be around 10uF.  The internal compensation is set for this, so by going 100x this value, you are likely causing instability, etc.

Try testing without the large output cap and see how your results look.

Hi! All

I am using this chip too.

To using such large capacitor, I guess for transience load.

If we can't using large value capacitor, what do we do tp improve transience load respond?

Lewis

you can use feedforward to improve the transient performance

Jim

Could you show me a schematic for your suggestion?

Here is an AN discussing how to add the feedforward cap.

http://www.ti.com/lit/an/slva289/slva289.pdf

Hello Jim

Sorry I am busy on another issue before.

I am using TPS61070 in our many projects.
It is provided 5V(from Li-iron) for small battery charging & buzzer. The total current consumption is below 300mA(@5V).

The problem is that the SW pin will become short circuit while power up.

I have below test.
1) Increase the capacitor of Vin & Vout. - No improve
2) Increase the start up delay(to 2s) after power up - no improve
3) adding small cap.(15pF) between Vout to FB - no improve
4) try the coil 2.2uH & 10uH - no improve * initially using MURATA LQH32CN4R7M53L

Below test try to repeat the defect(SW pin short circuit)
a) enable the chip before power up(Vin) - no damage
b) let Vout short circuit - no damage
c) open resistor(between Vout to FB) - no damage
d) let FB short circuit - no damage

Shall we make it always power on?
Is it any power sequence request?

Our management beginning considers changing another parts to replace it.

Please help to solve it. At least, I want to clear what going on.

Hi Lewis

the only reason of the SW pin damage is the over voltage stress, most of which result from the bad layout. could you share you schematic an PCB layout please? thanks.

Jasper li

Hi Jasper

About the over voltage stress to damage SW pin, the chip is mounted in PCB with another circuit.
And it haven't any connect to let us direct access.
In board level, we have Li-iron battery(3V-4.2V) only and only this chip has boostor function.
Base on data sheet the max. level is 7V.

I don't know why & how the over voltage stress to damage SW pin.

Please find circuit(attached before) & layout.

Lewis

hi Lewis

although the DC voltage is only 5V, but the voltage spike at SW pin will higher. if the voltage spike is higher than 7V, there is a risk to damage to the IC. please refer this AN for the root cause and solution of voltag spike of boost converter

http://www.ti.com/lit/an/slva255/slva255.pdf

the output capacitor in your PCB is too far away from IC. so  i suspect that the voltage spike in your application is much higher than 7V. please place the capacitor as closed to the IC as possible. you can find PCB layout reference for this device in http://www.ti.com/lit/ug/slvu106/slvu106.pdf and page 16.

thanks

Jasper li

Jasper

Thanks for your comment.

For layout issue, I try adding wire to short the capacitors to pin-2 directly. But no improve.

And I try adding capacitor(220pF ~ 47000pF) between SW to GND. But the spike level no any change(still higher than 7V)

BTW, please check with attached picture, there have some abnormal waveform in behind of switching signal.

Please let me know your comment.

BR

Lewis

the reason is that the track from IC GND to output capacitor GND is too long, as the red line. please try to place the ouput capacitor above the IC as shown below to double confirm the problem.

if this work, you can then consider to modify you PCB. you don't need to add snubber if doing good layout. and actually it is very hard to design good snubber, and most of time subber circuit will cause other problem.

Hi Jasper

I have move the output capacitor close to chip(as what you say).
But the problem doesn’t improve.

BTW, I try to add 1nF between Vout(pin-5) to FB(pin-4).

What will be affected of about this change?
Please let me know your comment.

BR

Lewis

Lewis

what is the phenomenon now. does the IC still damage in startup? please share the part number of the cap used. and the waveform of SW pin and output votlage and input voltage.

the cap from Vout to FB is used to improve the stability of the converter. it should not affect the SW spike if the converter is stable. i stongely suggest that select the feedback resistor  and this cap base on datasheet page 14 (R1 and R2 and CparR1).

thanks

Jasper li