TPS22946 Power switch IC

ZVI PESHKESS

Prodigy 50 points

Other Parts Discussed in Thread: TPS22946

We are using the TPS22946 to switch +5Vdc to optical detector IC ( Iload=10mA )

We found out that the device does not start sometime ( very random ) !

The circuit is connected with 'CL' and 'ON' are tied to Vin (5v),

We observed this problem with NO Load as well as with load !!

We put two channel scope on the Vin and Vout , when the part does not starts,

The waveform looks like Fig-18 on the data sheet (almost like the short circuit ) except there is NO spike at the begining

of the ON time.

NOTE: the problem shows when there is NO load !!! the only load is the output capacitor 0.1uF Ceramic cap.

We built a separate circuit with only the TPS22946 and monitored the VIN and Vout --

Vin 5v comes from the original PCB (with the original TPS22946 on it ) .

results:

1) when the TPS22946 is powered from the main 5vdc ( Initial Main Power = on ; 5Vcc=0 ; Switch is off ==> power is turned on )

the device fails often

2) The main PCb 5Vdc is already 'ON' ; than we connect the TPS22946 to the 5V to turn it ON

The TPS22946 does not fails ( at least not after 100+ trials )

Question: any idea why the part fails often when it is powred directly from 5Vcc ( +5Vcc power is ramped up from 0V to +5V while TPS22946 is following up to vcc) the part fails but when the +5Vcc is alrady stable than we switch the TPS22946 ??

please copy reply to zvipesh@marvintest.com

over 9 years ago

0 Aleksandras_Kaknevicius over 9 years ago

TI__Genius 15780 points

Hi ZVI,

If my understanding is correct, then the part is having trouble turning on when VIN, CL, and ON are all connected together. When power is brought up, VOUT does not come up, or only comes up a little. If VIN and CL are already powered and ON is toggled, then VOUT comes up normally. Is this an accurate summary?

Also, can you please provide the scope shots where the TPS22946 does and doesn't work?

Thanks,

Alek Kaknevicius

0 Aleksandras_Kaknevicius over 9 years ago

TI__Genius 15780 points

Hi ZVI,

I tried sending my response to your email (zvipesh@marvintest.com) but got a send error.

Thanks,

Alek Kaknevicius

0 ZVI PESHKESS over 9 years ago in reply to Aleksandras_Kaknevicius

Prodigy 50 points

Hi:

yes to your answer with one exception all three "Vin"+"CL"+"ON" are tied together.
I will send scope pictures tomorrow morning

0 Aleksandras_Kaknevicius over 9 years ago in reply to ZVI PESHKESS

TI__Genius 15780 points

Hi ZVI,

Sounds good. I have also ordered some TPS22946 EVMs so that we can verify the issue in our lab as well.

Thanks,

Alek Kaknevicius

0 ZVI PESHKESS over 9 years ago in reply to Aleksandras_Kaknevicius

Prodigy 50 points

here are the picture .

First picture is when the device powers up OK.

CH2(Blue Trace)= Input 5Vdc

CH3(Purple Trace)= TPS22946 Output

The failure is very repeatable.

0 Aleksandras_Kaknevicius over 9 years ago in reply to ZVI PESHKESS

TI__Genius 15780 points

Hi ZVI,

Thank you for the scope shots, your problem is very clear. I should have the EVMs by early next week and will try to replicate this failure as a way to begin debugging this issue.

Thanks,

Alek Kaknevicius

0 ZVI PESHKESS over 9 years ago in reply to Aleksandras_Kaknevicius

Prodigy 50 points

The last scope images came from a setup that included a 5V power supply and the TPS22946 on a prototype board.

looking at the waveform, It just occurred to me that the rise time and the slope of the power-on may be the issue.

The slower the rise time the worst is the case.

with the Vin rise time as shown I get a lots of failures, vs, when using the product P.S. and the prototype board.

Again:

When the VIN is on and stabilized ... and later on I connect the TPS22946 to the power it looks like there

is no failure.

Power management

Power management forum

TPS22946 Power switch IC