TPS54260EVM-597: PH=Vmin exceeded on data sheet when in any mode (pulse-skip/DCM/CCM), signal looks clean Vin =12 but Vin=24 (with UVLO update) I see a sawtooh pattern in CCM on the PH pin

jeff meunier

Prodigy 50 points

Part Number: TPS54260EVM-597

On Data Sheet, absolute max rating:

PH = -2volts, 10-ns Transient

Here are scope screen captures from eval kit (no changes):

EVAL Kit PH=-3.4v, Vin=12v, output load 500ma (w/10w power resistor)

Signal looks clean at Vin=12v, except for the undershoot spike that exceeds the data sheet -Vmax value for PH:

Here are scope screen captures from eval kit (Vin max set resistors ):

EVAL Kit PH=-7.1v (way past abs max value and holds below -2v for 500ns), Vin=24v, output load 500ma (w/10w power resistor)

Some sort of pseudo-sawtooth pattern (from additional shunt resistance at Vin=24v? / scope probe acting as current probe?):

Is this behavior ok or will both scenarios blow the chip as the data sheet seems to mention?

Ground lead for scope probe was connected directly to filter cap (GND)

over 7 years ago

0 Frank De Stasi over 7 years ago

TI__Guru 64515 points

I do not see any 'scope shots in this post?

In any case, the switch pin must be measured with a probe that has no ground lead. A very short

bare wire must be soldered to the nearest ground on the PCB and connected directly to the ground

barrel of the 'scope probe.

0 Frank De Stasi over 7 years ago in reply to Frank De Stasi

TI__Guru 64515 points

The wave forms for the 12V look normal, if you have a long ground lead on the scope probe.

The 24V waveform is not correct. It is possible that the thermal pad did not get soldered down

well or that the device and/or the diode has become damaged. You might try another EVM.

0 jeff meunier over 7 years ago in reply to Frank De Stasi

Prodigy 50 points

The eval module looks ok when I go back to Vin=12v (my custom PCBs all look the same too), I am going to swap the parts out on the eval module to match "Design 2, page 16" on this application report (www.ti.com/.../slva412.pdf) tomorrow and re-test. That design has a Vin=24v nominal, I will keep Vout at 3.3v and only move it exactly to 4.2v if I still have a problem (exact match). Switcher pro doesn't throw any errors, does webbench allow you to model the output of the PH pin?

0 jeff meunier over 7 years ago in reply to jeff meunier

Prodigy 50 points

Corrected link: www.ti.com/.../slva412.pdf

0 jeff meunier over 7 years ago in reply to Frank De Stasi

Prodigy 50 points

Everything looks good at 5v/div on my 350Mhz scope and down 1v/div on my 150Mhz scope, but when the relay kicks in on the scope for the lower v/div (probe impedance change) and sawtooth pattern appears. If I drop the Vin voltage from 24v down to 18v or less the sawtooth pattern disappears. There is no documentation on any of the datasheets with <5v/div screen captures so I guess everything is ok?

0 Frank De Stasi over 7 years ago in reply to jeff meunier

TI__Guru 64515 points

Many scopes will give false readings when overloaded.

Power management

Power management forum

TPS54260EVM-597: PH=Vmin exceeded on data sheet when in any mode (pulse-skip/DCM/CCM), signal looks clean Vin =12 but Vin=24 (with UVLO update) I see a sawtooh pattern in CCM on the PH pin