TPS62420: Issues with TPS62420

Hi Bernardo,

I am working on your request and will give you a feedback tomorrow.

Regards,

Febin

Thanks Febin,

I will wait for your comments,

Br.

Hi Febin,

Just to give you some more information in case it helps, today I applied a vanish coating to the IC and related componets (Caps and Resitors related to the IC) and it didn´t work. It happens exactly the same.

If I do the same operation but with an isolated element (touching the IC or surroundings with something that its isolated), everything works ok, so I discard a welding or mounting problem. I have made this test in at least three different boards and it is easily reptroducible.

Br

Hi Bernardo,

Thank you for all the information. I did not get a similar issue with this device and I am a bit surprised.

Could you please double check nothing else changed in the measuring environment when you do this test and that this happens only when you bring an isolated material close to the IC. What I understand is that the Vout oscillates in this scenario. Am I correct?

Regards,

Febin

Hi Febin,

It happens when I touch the IC and/or surrounding components with my bare finger. What it happens (in my board) is that the Vout (only the 1V2 (SW2)) oscillates at 100 Hz and reaches up to nearly 3V.

When I do the same (touching with the finger the area of the converter) in the TI evaluation board for the 28379D MCU it happens something similar in the 3V3. The MCU also resets. (See figures in first messages)

If I touch the IC and/or surrounding components with something isolated (such a plastic pen for example) nothing happens, so I´m thinking in a capacitive related phenomena.

Br,

Hi Bernardo,

What I understand is that when you touch the IC or surroundings with your finger, you could be unknowingly touching any of the signal pins or feedback pin as well. If it injects noise into the f/b pin then there is a possibility to see this in the output voltage too.

I do not think there is any issue here. I guess your touching the f/b or signal pins and they are pretty sensitive.

I hope this helps! If you do not have further questions, I would like to close this thread. 

Regards,

Febin

Hi Febin,

I understand that that is what could be happening, but I´m worried that this signals are too sensitive to work in an industrial environment.

I´m going to try to make a radiated immunity test so I confirm that everything works ok under such conditions.

Br,

Hi Febin,

I´ve performed some more tests (for example I lowered the values of R11 / R12...) and happens exactly the same. I don´t even have to touch the IC to make it fail. It seems that with the capacitance you add simply getting close to the IC is enough to make it fail. This is obviuously not acceptable so I´ve decided to change the IC.

Do you have any recommendation of one (or two if necessary) step down converters that would be more robust than this one?,

Br,

Hi Bernardo,

I am not sure of the tests you did so far. But if you inject noise into any of the signal pins especially feedback pin, this will be seen in Vout.

I can propose newer devices, please let me know the complete requirements.

Please be informed that I am on vacation until 11/11, so please expect a delay in my response!

Thanks for your understanding!

Regards,

Febin

Hi Febin,

The thing is that just approaching the finger to the IC makes the IC work bad, we´ve invesitgated and checked routing and everything seems to be ok and according to requirements. I´ve decided to change to 2 x TPS56628 (for the 3V3 and 1V2)since we´ve used them in other designs with better results.

If you have any other recommendations it would be nice to hear them.

Br,

Hi Bernardo,

I was on vacation as I mentioned in the previous post.

As a quick update, did you already finalize your design? 

Regards,

Febin

Hi Febin,

I made several tests (lowering the values of the resistors to set the voltage, changing bypass and output capacitors...) and the problem remains exactly the same.

I´ve decided to change the TPS62420 for two TPS56628DDAR, one for each rail. I´ve used these in some designs with good results. I don´t now if you have a better recommendation.

Br,

Hi Bernardo,

Unfortunately, we do not have any replacement parts for TPS62420 in our current portfolio. 

TPS56628DDAR does not fall in the product line I take care. If you need further support on this device, please open a new thread.

If you are open to single output bucks in the low Vin (<7V), I can propose a few devices based on your requirements.

Regards,

Febin