TPS25820: TPS25820 does not source power

Hi Sam,

Can you attach a schematic so I can review what is connected to the CC lines?

Thanks,

Emma

CC1 and CC2 connect to the phone/tablet.

Hello,

Can you check what your enable voltage is?

Can you include a scope capture of VBUS, CC1 and CC2 where the device fails to connect and where the device connects?

Thanks,

Emma

Emma,

Ch1 = SINKn

Ch2 = Vbus

Ch3 = CC1

Ch4 = CC2

Note, the scope VBUS voltage measurement is a little off. With a voltmeter, it is 4.8V.

Plug inserted.

CC2 grounded with a clip lead.

Device powered.

Plug removed.

Thanks!

And, the EN voltage is 3.3.

Hi Sam,

Thanks for all the good scope shots. I have another request. Can you monitor V5P0 as well as VBUS, CC1, and CC2 and attach a scope shot. 

Thanks,

Emma

Emma,

Ch1 = V5P0

Ch2 = Vbus

Ch3 = CC1

Ch4 = CC2

Cable insertion, Vbus is never powered...

TPS25820 unpowered

Thanks,
Sam

Emma,

Might there be a reason the TPS25820 is incompatible with a phone that can both source and sink?

When TPS25820 fails to source, it seems CC1 and CC2 follow each other. If I force CC1 or CC2 to ground, TPS25820 does source, and CC1 and CC2 go to opposite levels.

Thanks,

Sam

Hi Sam,

Yes the phone could not be applying the Rd pulldown correctly. Could you try a different sink to see what the results are?

Thanks,

Emma

Emma,

That was the first thing I did.  I’ve tried a Moto X4 phone and Lenovo M10 tablet.  They have slightly different signatures.  Neither one actually sinks, unless I manually tie either CC to ground.

Is the TPS25820 known for compatibility issues?  Is there a different part that is known to be highly compatible?

Thanks,
Sam

Hi Sam,

Please allow me a few days to work with the team on these issues. The TPS25820 does not have any known compatibility issues. 

Thanks,

Emma

Emma,

Ok.  Thanks for letting me know.

I also ordered an EVM board for comparison.

Thanks,

Sam

Great, let me know how that testing looks on the EVM. I will update you in a few days. 

Thanks,

Emma

Emma,

Should both CC lines be connected to the device?  Or, just one?

I tried disconnecting one of the CC lines, and things were much happier.

Thanks.

Hi Sam,

Both of the CC lines should be connected to the TPS25820. 

Thanks,

Emma

Emma,
It appears the cable cuts one of the CC lines.  Both lines are not connected through to the device.
It appears the TPS25820 does not react gracefully and report a fault, if both lines are connected all the way to the device.
Disconnecting one CC line at some point appears to be a requirement.  Correct?
Thanks,
Sam

Hi Sam,

The source connection works by advertising Rp on both of the CC lines and when a sink is attached, it will put an Rd on one of the CC lines in order to indicate attachment. Both CC lines should be present and you should not need to pull either to ground to get the TPS25820 to source power.  

Certain non-compliant cables exist that only have one CC line in them. I would recommend using a different cable with both CC lines intact. 

Thanks,

Emma

Emma,

Now charging happens, however I cannot get a data connection.

If I remove the TPS25820 and replace with a 5.1k resistor to ground, data works fine.

Any suggestions?

Thanks,

Sam

Hi Sam, 

What did you change to get charging to work? Also, what is your sink? 

Can you send what your schematic looks like without the TPS25820 and just an Rd?

Thanks,

Emma 

Emma,

To get charging to work, I had to use a cable with one CC line connected through (as in the drawing you provided).  If the cable has both CC lines connected, the TPS25820 never sources power.

My application requires the phone be a data host, not a data peripheral.  However, the phone thinks it should be in data peripheral mode when it is a power sink. 

If the phone is a power source (TPS25820 removed), it is happy being a data host.  However, simultaneous charging from the no longer happens.

It appears this is possible, but the TPS25820 is the incorrect part.

What would the correct part be?

Thanks,

Sam