TPS5430 buck convertor - oscillates & wouldn't start

[zserg]

Hi all!

I have a buck regulator based on TPS5430 from TI, which wouldn't start properly.
The design has been done using TI WEBENCH tool & all the components have been chosen from their recommended BoM.
However, when I supply 12V in, the output voltage gets stuck at 1.255 V and the PH pin scope trace just shows weird oscillations.
What is also weird, the Vsense pin also shows 1.255V, while according to the datasheet it should be 1.221V (internal REF).
I tried playing with input/output capacitance - no luck. Again, the input and output caps were chosen according to WEBENCH recommendations, I've checked them against the datasheet and they fall within specs.
What I've found is that if I briefly short the bootstrap capacitor, the circuit would start, though the PH trace still looks weird on the scope.
If I switch it down and let Cout fully discharge, it wouldn't start again without kicking Cboot.
Now, the datasheet explicitly says to use Cboot of 10n, which is exactly what I've done. I've tried 4.7n and 2x10n in parallel instead - that didn't help.

Any ideas what the reason could be?

[Colt45]

Try putting an electrolytic across the input cap, see if it starts.

Sometimes with a long lead from the wall wart, and only ceramic input caps, and no TVS, the input will ring. Seems to sometimes confuse the buck chips.

[zserg]

Tried already, unfortunately no alterations to input/output capacitors seem to help. :/

[SiliconWizard]

However, when I supply 12V in, the output voltage gets stuck at 1.255 V and the PH pin scope trace just shows weird oscillations.
What is also weird, the Vsense pin also shows 1.255V, while according to the datasheet it should be 1.221V (internal REF).

Vsense is the feedback input of the converter, which in normal regulating conditions will "servo" it to Vref. Not exactly the same thing. It's clear here that the converter is not in normal regulating conditions.

IMO, the second part should be a big hint: how can Vsense, which is the output of a resistive divider, be equal to the input voltage of said divider? Can you check both resistors and your wiring first?

If all is well with this, we could then suspect the Vsense pad is fried. But check all your wiring first.

[zserg]

Thanks a lot for the reply!

Vsense is the feedback input of the converter, which in normal regulating conditions will "servo" it to Vref. Not exactly the same thing. It's clear here that the converter is not in normal regulating conditions.

Yup, that's what I meant, was just trying to be brief.

IMO, the second part should be a big hint: how can Vsense, which is the output of a resistive divider, be equal to the input voltage of said divider? Can you check both resistors and your wiring first?

Fair point! I've checked the wiring & resistor values first thing - it's all OK. Will check it one more time tomorrow though.
For now, it can be seen from the PCB layout that the voltage divider is where it should be - those two 0805 resistors right between the IC and the inductor.

[thinkfat]

I remember a similar question here where the voltage divider was mixed up during assembly. Low- and high-side resistors were flipped.

Now, if you recalculate the output voltage with the resistors flipped you get around 1.6V. What's the minimum duty cycle of the chip? Can it actually reach 1.6V with an input of 12V?

[zserg]

I remember a similar question here where the voltage divider was mixed up during assembly. Low- and high-side resistors were flipped.

Now, if you recalculate the output voltage with the resistors flipped you get around 1.6V. What's the minimum duty cycle of the chip? Can it actually reach 1.6V with an input of 12V?

It's about 8%, so, yes, it can. Checked that already, but will re-check it tomorrow. If this is messed up the resistors, would be quite a shame. 

[Niklas]

I cannot see the 220uF output capacitor and possibly only one of the 1uF on the output in the layout overview. The datasheet mentions low ESR type capacitors, so those capacitors should be located somewhere close. During the fast switching it might be possible that the regulator only sees the 1uF ceramic capacitor and that voltage is also used for the feedback.

TI has a demo board for this IC and it might be a good idea to check the manual how they arranged the components on that board. Especially the 0V connection between the input caps, switcher IC, diode and output caps is done on the top side layer.

[thinkfat]

Hm. If Vsense == Vout, the top reason I can think of is that the low-side resistor of the voltage divider is broken. Vsense > Vref would mean the controller will reduce the duty cycle. And if the difference at the error amplifier input is that large (Vsense >> Vref), it's no wonder you don't get any switching action. The controller has basically shut down.

[OwO]

The fact that the circuit does start in certain conditions (shorting the bootstrap cap), if true, suggests that the feedback divider might not be the culprit. The VSENSE pin is not necessarily a voltage mode (high impedance) input and it could be pulling the voltage to near VREF at all times (just like in a current feedback amplifier). VSENSE being slightly higher than VREF suggests the divider is trying to pull it up. TI makes good ADCs but their power management ICs have a reputation of being easy to "lock up" or get into a bad "state" because they are usually just a MCU with integrated power transistors: https://www.eevblog.com/forum/projects/psa-do-not-use-the-tps61099-boost-reg-in-your-designs/msg2141038/#msg2141038
I would check the power input slew rate and any overshoots, or try another part.

[Warhawk]

Could you share a picture of your board, please?

[ocset]

Make a 5V linear shunt regulator with a big TO247 FET…put it on a big heatsink…..then connect it to your output..then ground the vsense pin, apply vin  and see  what you get at the output…ensure your inductor is rated  for high enough current first.
If you don’t get anything at the output in that case, then I think the TPS5430 has been ESD’d and is dead.

[zserg]

Hi all!

Thanks for all suggestions. It turned out to be dodgy connection on the voltage feedback resitor divider. All measurements were OK, but today I just blew off both resistors with hot air to check them separately, then soldered both of them back and it works. Shame on me.

I cannot see the 220uF output capacitor and possibly only one of the 1uF on the output in the layout overview. The datasheet mentions low ESR type capacitors, so those capacitors should be located somewhere close. During the fast switching it might be possible that the regulator only sees the 1uF ceramic capacitor and that voltage is also used for the feedback.

Yup, the layout without designators look confusing, the guy on the TI forum didn't find the output cap as well. I've attached the version with designators shown.
Regarding the ceramics on the output, strangely enough, TI design tool (WEBENCH) just recommends a single low ESR tantalum cap on the output - I've added 1 uF MLCCs myself.
The output cap is as close as possible to the V sense feedback exactly as you've suggested - the middle point of resistor divider connects right to the cap pin.

Actually, this Panasonic cap raises some questions. They state 35 mOhm max ESR@100 kHz in the   datasheet for 6.3V 220 uF
This value, if put into the formula from the TPS5430 datesheet gives me perfect 18 mV output ripple voltage. However, the measured ESR on those caps I have is more like 120-130 mOhm, which gives much less enjoyable 60 mV ripple, which is exactly what I see in the scope trace. Any ideas why?
Ok, that's me being an idiot, noise was picked up by scope ground wire, measured with the proper probe (with spring clip) and now it's well within 20 mV.

And another question - can someone explain the PWM shape under small loads? I've attached three traces:
12V input, 5V output, 50 mA load
12V input, 5V output, 1.6A load
24V input, 5V output, 1.6 A load

The latter two look perfectly reasonable & the duty cycle changes according to the change in input voltage.
The trace under 50 mA load looks like a strange overlap of the normal rectangular wave with some half-arsed half-sine. Has anyone seen anything like that?

[thinkfat]

Yeah, that waveform is from the converter switching to pulse skipping mode during low load. Check the data sheet, it's typical for TI parts.

[zserg]

Yeah, that waveform is from the converter switching to pulse skipping mode during low load. Check the data sheet, it's typical for TI parts.

You are right, that's due to pulse skipping. That's basically the tail of ringing from the skipped pulse - after a bit of googling, I've found several similar traces:
https://e2e.ti.com/support/power-management/f/196/t/591911?TPS54232-TPS54232-Pulse-Skipping-ECO-MODE-Disable
https://e2e.ti.com/support/power-management/f/196/t/818662
https://www.analog.com/ru/technical-articles/inverting-dc-dc-controller-converts-a-positive-input-to-a-negative-output-with-a-single-inductor.html#

In my case it was just a single half-period of the ringing waveform, so I didn't recognize it.

[Niklas]

What you see at lower load current is discontinuos current (DCM) through the inductor. During the high part of the pwm the inductor is charged and that is followed by a short low part where the inductor is discharged through the load and freewheeling diode. The sinusoidal part is just due to residual charge that make the parasitics oscillate. You can try a 10k resistor from the switch node to GND and see that it will bleed the oscillating charge so the ringing is dampend. You will not need it in your design, but just to see that the energy level is quite low in the sine wave part.
When the load current increases, then there will always flow current through the inductor (CCM). That current is typically sawtooth shaped and centered around the load's average DC current.

[SiliconWizard]

Thanks for all suggestions. It turned out to be dodgy connection on the voltage feedback resitor divider. All measurements were OK, but today I just blew off both resistors with hot air to check them separately, then soldered both of them back and it works. Shame on me.

Oh, that can happen to just anyone!
The important thing here is hopefully you will have learned one lesson or two!
First one, checking assembly after hand-soldering. If you have one, use a stereo microscope, also don't hesitate to check connections with a multimeter when in doubt.
The second one, how to debug hardware using a methodical approach.

[zserg]

Thanks for all suggestions. It turned out to be dodgy connection on the voltage feedback resitor divider. All measurements were OK, but today I just blew off both resistors with hot air to check them separately, then soldered both of them back and it works. Shame on me.

Oh, that can happen to just anyone!
The important thing here is hopefully you will have learned one lesson or two!
First one, checking assembly after hand-soldering. If you have one, use a stereo microscope, also don't hesitate to check connections with a multimeter when in doubt.
The second one, how to debug hardware using a methodical approach.

Well, the point it, everything was absolutely fine according to multimeter readings - I've checked all parts' values as well as all the connections when I figured out it wouldn't turn on. So it's a bit of a mystery why de-soldering and soldering back did fix it.

[SiliconWizard]

Some bad soldering jobs can create poor connections that are hard to spot with a multimeter. Of course, you first have to test from component pin to component pin. But even so, sometimes just pressing on the pins with a probe is enough to make a good connection, so you miss the problem. This is where a microscope will also help.

But as you saw, inspection is only one step. From your symptoms, it was clear there was a problem either with the feedback network, or with the converter itself.