TPS65982: Actual EXT, HV and 5V current measurement with ADCs read operation.

Yevhen Horbenko

Intellectual 510 points

Part Number: TPS65982
Other Parts Discussed in Thread: BQ25713

Hi guys,

thanx in advance for assistance.

I need measuring an actual current flows through TPS65982.

I've programed an I2C read\write algorithm:

4CC ADCs command to Cmd1 (0x08) + 1 byte according ADC channel (0:4 bits) to Data1 (0x09) register
Cmd1 (0x08) reading to be compared with following things: whether result is equal to "ADCs" the command is processing now, if "!CMD" so that I have an error and whether it's equal to "\0\0\0\0" the ADCs command has been processed
reading the 2 bytes result from Data1 (0x09) register
if bit 15 (ADCBusy) is one the step #3 is performing again
if bit 15 is zero the ADC measurement result is getting masking to 0:9 bits

An important point: I didn't find any manual descripting ADCs read operation's channels enumeration list. I've used "TPS6598x Application Customization Tool" taking there a drop-down list of ADCs channels.

According the list's items order the channels enums are:

PP_EXT_CURRENT = 3

PP_HV_CURRENT = 5

PP_5V_CURRENT = 7

I've passed them in step #1.

Having no USB devices plugged into USB-C port I was getting results kind like these:

PP_EXT 0

PP_EXT 256

PP_EXT 512

PP_5V 512

PP_5V 256

PP_5V 0

PP_HV 0

PP_HV 512

PP_HV 256

I'm confused of these two fixed 512 and 256 values.

So my questions:

1) Did I use the ADCs channels' enums right? Whether I didn't please point me to a needed article of TI document?

(slvuan1a.pdf doesn't disclose this info)

2) Why the ADC results are those fixed ones (256 and 512)?

over 5 years ago

0 Adam Mc Gaffin over 5 years ago

TI__Mastermind 30000 points

Hello Yevhen,

The expert covering this device is located in India so please expect a delay in response. Thanks in advance for your patience.

0 Prajith Cheerakkoda over 5 years ago in reply to Adam Mc Gaffin

TI__Expert 8680 points

The channels (3, 5, 7) which you are reading are correct.

When the switches are not enabled, ADC returns garbage values. So, please ignore ADC values when the switch is not enabled.

Thanks

Prajith

0 Yevhen Horbenko over 5 years ago in reply to Prajith Cheerakkoda

Intellectual 510 points

Prajith Cheerakkoda

Thank you for the reply.

I'll approve my issue as fixed after some tests.

Can you tell what a document did you use getting know the information about "ADCs" ADC channels? Or is it only of personal practice experience?

0 Yevhen Horbenko over 5 years ago in reply to Prajith Cheerakkoda

Intellectual 510 points

Prajith Cheerakkoda

I've tested "ADCs" reading with SNK device plugged in and SRC charging device plugged in.

The same results and "768" is getting sometimes.

"When the switches are not enabled..."

1) Should I enable the switches in any way? Or would it be enough if PD contract is active and a current is flowing?

Red underline: 2 ^ 10 = 1024 is too small to display all kinds of current could be.

2) Should I use any coeff.? Or would any V(ref) be used in calculation of the result?

Maroon underline of the 1st and the 2nd screenshots:

3) Can 0x6A be read properly at all?

So, these 3 qeustions...

Thanx in advance.

0 Yevhen Horbenko over 5 years ago in reply to Yevhen Horbenko

Intellectual 510 points

Update:

Question 2 is answered:

https://e2e.ti.com/support/interface/f/138/t/635918

"13.1 ADC" provides the equation.

----------------------------------------------

In my project I have BQ25713 and TPS65982 on a bus. I had plugged MacBook into my testing PCB setup. I had measured IADPT (BQ25713 ADC value) current and I had got "ADCs" read result of HV_EXT channel. I was comparing these two measurement values at about the same times. I have to say that results have been disparates, e.g.:

TPS65982 ADC(dec)	TPS65982 ADC EXT channel result (A)	BQ25713 IADPT (mA)
0	0	0
0	0	42
768	4,50439882697947	996
768	4,50439882697947	1488
512	3,00293255131965	767
512	3,00293255131965	849
512	3,00293255131965	1036
512	3,00293255131965	946
256	1,50146627565982	1027
256	1,50146627565982	1199

So, questions 1 and 3 remain unanswered for me. I would only add to the first one "What could I do wrong with my "ADCs" measurement described in post#1?"

Thanks in advance.

0 Prajith Cheerakkoda over 5 years ago in reply to Yevhen Horbenko

TI__Expert 8680 points

Hi,

Based on the internal discussion, ADC use is limited to the PD FW. We suggest you to use an external current sensor for other uses.

Thanks

Prajith

0 Yevhen Horbenko over 5 years ago in reply to Prajith Cheerakkoda

Intellectual 510 points

Guys, an unexpected discovery.

I had put Saleae Logic Analyzer on the I2C bus when I was reading PP_EXT_CURRENT channel with "ADCs" command from "TPS6598x Application Customization Tool". Aardvark device was used.

Setting EXT switch as ADC channel parameter, performing "ADCs" read operation.

Setup Write to [0x7E] + ACK
0x09 + ACK	Data1 register
0x01 + ACK	length
0x03 + ACK	EXT channel
Setup Write to [0x7E] + ACK
0x08 + ACK	Cmd1 register
0x04 + ACK	length
0x41 + ACK	A
0x44 + ACK	D
0x43 + ACK	C
0x73 + ACK	s

Checking "ADCs" command execution state (command processed cause zeroes returned), reading ADC result (Aardvark does it twice one by one).

Setup Write to [0x7E] + ACK
0x08 + ACK	Cmd1 register
Setup Read to [0x7F] + ACK
0x04 + ACK	length
0x00 + ACK	0
0x00 + ACK	0
0x00 + ACK	0
0x00 + NAK	0
Setup Write to [0x7E] + ACK
0x09 + ACK	Data1 register
Setup Read to [0x7F] + ACK
0x40 + ACK	64 bytes len
0x03 + ACK	ADC ch 3
0xC3 + ACK	Data1 result LSB
0x00 + NAK	Data1 result MSB

Please look at red marked line. Data1 register reading result keeps the channel number.

I haven't expected this byte is presented in the message at all. So I've parsed the needed result beginning of it but not beginning of Data1 result LSB.

How could a developer know this not debugging?

You can see bytes count starts with 1st which keeps LSB of ADCResult and not with any zero byte keeping ADC channel number.

Only three dots could implicitly hint that something could be before 1st byte:

OMG :-(

So I have got the answers for my questions.

Data1 result LSB

0 Yevhen Horbenko over 5 years ago in reply to Prajith Cheerakkoda

Intellectual 510 points

The last point I have unclarified is why I could get output current results only while a power consumer is connected to USB-C port. And I can't get input current real values except zeroes when a charging device is connected to USB-C port.

I've checked this fact with Logic Analyzer and yes, I'm getting zero values in LSB and MSB:

Setup Write to [0x7E] + ACK

0x09 + ACK

Setup Read to [0x7F] + ACK

0x40 + ACK

0x03 + ACK

0x00 + ACK

0x00 + NAK

So please tell me what reasons could be to not receive this real values in this case?

0 Yevhen Horbenko over 5 years ago in reply to Prajith Cheerakkoda

Intellectual 510 points

I tested "ADCs" reading with Aardvark when a 87W Apple charging device was supplying a setup test board.

And as in the case with my SW measuring the Aardvark detects a zero current on EXT switch channel after "ADCs" has performed.

So I'm updating my questions to the last ones:

1) Could an input current be measured on TPS65982 at all?

2) Whether it could what mandatory settings have to be applied to its configuration?

Thanks in advance.

0 Prajith Cheerakkoda over 5 years ago in reply to Yevhen Horbenko

TI__Expert 8680 points

Hi,

Please find my response below.

1) Could an input current be measured on TPS65982 at all?

>>> No, current measurement with ADCs is possible only when the Power Path is enabled AND the port sourcing power.

2) Whether it could what mandatory settings have to be applied to its configuration?

>>> Please find response above. Configuration setting cannot change this behavior.

Thanks

Prajith

0 Yevhen Horbenko over 5 years ago in reply to Prajith Cheerakkoda

Intellectual 510 points

Thank you. The issue has been solved.

P.S. I ask TI please add fixes to TPS65982 Reference Manual about returning by "ADCs" via DataX also the measured Channel number before a ADC(dec) value. Cause it's a real issue which wastes a developer time while he\she's debugging to find a reason.

0 Prajith Cheerakkoda over 5 years ago in reply to Yevhen Horbenko

TI__Expert 8680 points

Hi,

As I mentioned in my previous post, ADC use is restricted to the PD FW. We will consider removing ADC related info from TRM in the next revision.

Thanks
Prajith

Interface

Interface forum

TPS65982: Actual EXT, HV and 5V current measurement with ADCs read operation.

Prajith Cheerakkoda

Prajith Cheerakkoda