TPS2514: issue with the long detection time

Vincenzo Pizzolante

Part Number: TPS2514

hi,

We have some problems with TPS2514DBVR during USB Battery Charging 1.2 certification.

Our implementation is very straightforward (“TPS2514 schematic diagram.jpg”).

We are facing 2 major problems:

A) “8.3 DCP Handshaking test” from USB Battery Charging 1.2 Compliance Plan fails. According to waveforms form Certification Unit (file “8.3 DCP handshaking test (from CU).jpg”) and bench tests we see that controller takes too much time between detecting 0.6V on D+ line and driving 0.6V on D- line. As demanded in Battery Charging Specification this time (TVDMSRC_EN) cannot be longer than 20ms. We measure approximately 137ms. Customer believes that the problem is in TPS2514’s auto-detect block, which takes too much time to switch from divider mode to short D+/D- mode.

B) “8.4 DCP Resistance and capacitance tests” also fails for the same reason (file “8.4 DCP resistance and capacitance tests (from CU).jpg”). Test procedure gives only 21 ms (TVDMSRC_EN) to emulate attaching power device. So customer believes that TPS2514 is still in divider mode and it prevents further measurements (resistance between D+/D-, leakage and capacitance of data lines).

May you give me a hand with this?

thansk a lot in advance

Vincenzo

USB Switch.zip

over 4 years ago

0 Emma Hoying over 4 years ago

TI__Genius 9630 points

Vincenzo,

Can you provide more zoomed in graphs and also annotate where the testing begins and where the detection time fails compliance? For 8.3, I can't tell where you are getting the 137ms and am having trouble seeing where each test begins and fails.

Thanks,

Emma

0 Vincenzo Pizzolante over 4 years ago in reply to Emma Hoying

TI__Mastermind 22776 points

sure:

It shows an event when we make TPS2514 to change its mode from divider mode (default on startup) to short mode.

It begins when a D+ line is attached to 0.6V and ends when voltage on D- drops to same value.

This period is equal to 137ms. It is called as TVDMSRC_EN and should be maximum 20ms according to Battery Charging Specification.

thanks a lot for your help

Vincenzo

0 Emma Hoying over 4 years ago in reply to Vincenzo Pizzolante

TI__Genius 9630 points

Vincenzo,

Have you been able to replicate the result on an EVM?

Thanks,

Emma

0 Vincenzo Pizzolante over 4 years ago in reply to Emma Hoying

TI__Mastermind 22776 points

no, but implementation was attached, can't see any specific difference...

plus we don't have neither the BW nor the lab to test the EVM...

on top, I'm sitting in Germany, customer in Poland

looking fwd to hear from you soon

Vincenzo

0 Emma Hoying over 4 years ago in reply to Vincenzo Pizzolante

TI__Genius 9630 points

Vincenzo,

Taking another look at your schematic, you have the Dp and Dm wired to the wrong pins in the P501, which will cause problems with the detection times.

DP should be wired to D+

DM should be wired to D-

I think that this should solve your timing problem.

Thanks,

Emma

0 Vincenzo Pizzolante over 4 years ago in reply to Emma Hoying

TI__Mastermind 22776 points

hi Emma,

design has been made accordingly to datasheet (see fig. 16), as this is a 5W USB charger.

nevertheless we tested the suggested connection (10W, figu.17): unfortunately outcome is the same.

any advice?

thansk a lot in advance

Vincenzo

0 Emma Hoying over 4 years ago in reply to Vincenzo Pizzolante

TI__Genius 9630 points

Vincenzo,

Have you had this issue trying multiple devices? Or have you only been testing one chip?

Thanks,

Emma

0 Emma Hoying over 4 years ago in reply to Emma Hoying

TI__Genius 9630 points

Vincenzo,

After looking into it more and talking with one of our designers, the TPS2514's original intent is for wall adapters that can charge a wide range of devices that are not necessarily compliant. This means that doing a compliance test on the part will fail. While it is not a USB compliant charger, it does support compliant charging schemes. I hope this clears up your question.

Thanks,

Emma

0 Vincenzo Pizzolante over 4 years ago in reply to Emma Hoying

TI__Mastermind 22776 points

hi Emma,

thanks, it helped.

a very last question: would the TPS254x do the job (i.e. meet the USB Battery Charging 1.2 compliance)?

if not, what is the right part (pretty trivial with current filtering options available on the CPC portal)?

thansk a lot

Vincenzo

0 Emma Hoying over 4 years ago in reply to Vincenzo Pizzolante

TI__Genius 9630 points

Vincenzo,

TPS254x does have compliant modes. However, it also has modes that that TPS2514 has, DCP auto and DCP divider 1, that fail compliance. If you keep the TPS254x in a mode other than DCP auto or divider 1, it will pass compliance. DCP auto and divider 1 support noncompliant charging schemes and therefore will not pass compliance.

Thanks,

Emma

Power management

Power management forum

TPS2514: issue with the long detection time