TPS25750: Power management forum

Christopher Lim1

Hi Ali,

How are you currently configuring device? Using GUI you are able to configure device for sink only and change negotiation to be able to sink up to 100W (20V, 5A).

Please see this video for how to configure device using GUI: https://www.youtube.com/watch?v=_y61fyUjo9E

Dead battery condition is supported when there is no battery or a dead battery where the PD will recieve power from VBUS in order to operate. The device will present an Rd during dead-battery or no battery condition on the CC pin in order to present as a sink for the system to operate. 

Very Respectfully,

Brandon Beader

Hi Brandon Beader
Thanks for your reply.
Yes i configured everything perfectly using GUI and Flash the EEPROM using AAVARDAK but I never found PD controller to negotiate the voltage at higher level neither current goes to higher value. It's behave like a constant power supply.

Regarding dead battery, the battery always connect with the board but I believe due to some connection, the board behave like in  a dead battery mode and this might be the reason of PD controller always negotiate 5V supply.

I followed EVM schematic. I read in the datasheet that if I have to use always as a sink then I don't need P3v3 and PP5V circuit.
Because of this, I left VIN_3V3 pin of PD controller with no power. Just a 10uF capacitor is connected. 
In conclusion, my board always start in dead battery condition.

1. Will it affect to negotiate higher voltages for PD controller or it is just startup thing?
2. Also, i used 1A setting for BQ25792 using a 17.4K resistor but I also read somewhere that after loading EEPROM configuration, the register setup accordingly and this fixed value resistor setting is only for startup. Can it also cause the current to not move to higher value?

Thanks

...

Hi Ali,

Can you provide schematic and full flash binary to test on our EVM?

As for BQ device, yes this resister is the default setting for configuring the device. This setting can be changed after loading EEPROM depending on your configuration in the GUI.

Very Respectfully,

Brandon Beader

tps25750_122422.zip
Hi Brandon Beader
Plz, see the program file.
I wasn't able to send earlier due to some issue with Ti page.
Can i send you schematic in private message?
Thanks

Hi Ali,

Yes please send me JSON file for configuration as well as schematic through private message.

Very Respectfully,

Brandon Beader

4024.config.zip
Hi Brandon Beader
Plz, check the attached file. I also sent you schematic in DM.
Thanks

Hi Ali,

I was able to test the binary on our EVM and PD was able to successfully negotiate up to 20V. I will review schematic to make sure there is no design issues, please allow some lead time for this. 

My setup: TPS25750EVM, PD analyzer, Ardvark I2C/SPI, PD Source board.

I first powered the TPS25750EVM board using PD source board and flashed EEPROM using Ardvark I2C. I then power cycled the board to make sure EEPROM was correctly flashed. I then hooked up TPS25750EVM board back to the source board and offered up to 20V, 3A. The TPS25750EVM loaded with your binary was able to request and negotiate up to higher voltage.

Very Respectfully,

Brandon Beader

Hi Brandon Beader
Thanks for the update.
I want to let you know that. I am using TPS25750 as a sink only and for 100W.
I believe, it should go for 20V 5A.
I will wait for your feedback about schematic.
Thanks

Hi Ali,

I have partially reviewed your schematic and it seems okay. Please make sure you are flashing to correct EEPROM: CAT24C256. My assumption is the EEPROM is not being correctly flashed, therefore the configuration i not being loaded by the PD. Can you send your steps you used for flashing EEPROM?

Very Respectfully,

Brandon Beader

Hi Brandon Beader

Thanks for further feedback.

I downloaded the full flash binary. Then using flash control, i uploaded the file into flash. It says successful process after pressing program and verify.

I selected the same catalyst as you mentioned above?

Is it possible to flashed it with wrong catalyst name?

Thanks 

Hi Ali,

Is the battery voltage charging at 5V or is the PD negotiating 5V? If PD is successfully negotiating higher voltages, then the configuration to the BQ device is the issue. I want to ensure we are isolating to PD or BQ device.

Very Respectfully,

Brandon Beader

The PD always negotiate 5V with constant current 1.24A.

I charged battery from some 14.8V to 16.2V and it shows me same voltages and current.

I don't think so it is BQ issue because it is PD job to negotiate higher voltages but i am not sure. Maybe BQ also play some role.

But i never saw PD negotiating higher voltages or changing current.

It took few hours to charge battery from Nominal voltages to full.

Brandon Beader

Hi Ali,

You are correct this is PD issue, I just wanted to double check. Can you capture PD logs? From there we can see if PD is attempting to negotiate to higher voltages. 

Very Respectfully,

Brandon Beader

Hi Brandon Beader
Plz,see the PD analyzer results in different conditions.
1: Battery disconnected and PD wall adapter connected only.

2: Battery connected and pd wall adapter connected also.

3: Battery again disconnected

Hi Ali,

It appears there is a problem loading in the FW patch and configuration since the source is transmitting its capabilities and we are not responding. Can you probe the I2C lines and see upon power up if PD is attempting to read the EEPROM? Make sure it is trying to load from serial address 0x50.

Very Respectfully,

Brandon Beader

Hi Brandon Beader

I ordered a new scope that will deliver to me in few days.

I will try to measure the voltages at I2C lines to see if it vary like pwm signal just to make a guess.

How we can read the patch from eeprom using aavardak? I mean what should be the results if i read eeprom?

Hi Ali,

You can use the Ardvark and set the slave address as 0x50 to see if the PD is attempting to read/write to EEPROM. However, this will mean that the PD will not actually be pulling the patch and config from the EEPROM. This is to test if the PD is attempting to read from EEPROM. In order to capture I2C transactions better, I recommend using a Saleae device.

If the PD is trying to read from 0x50, then we can narrow this down to a flashing issue.

Very Respectfully,

Brandon Beader

Hi Brandon Beader

Plz, see the attached video about the process of uploading patch. After uploading, i did power cycle but same results

Hi Ali,

It seems you are flashing the EEPROM correctly, the next step is to probe the I2C lines using a Saleae to see if the PD is loading base image and configuration from the EEPROM. If this PD is successfully loading from the EEPROM, then it becomes a PD issues itself. 

If you do not have a Saleae, you can keep Ardvark connected and configure Ardvark as a slave using Control Center. In control center you will select adapter->connect->Ardvark-slave->ok. Then keep the Ardvark connected during bootup. This will help us see if the PD is attempting to read from slave address 0x50. I still advise using Saleae as this will allow us to confirm the binary that you are flashing is actually being stored in the EEPROM and the PD is loading.

Very Respectfully,

Brandon Beader

Hi Brandon Beader
Plz, see the video. I don't have Saleae for now but I order a cheap logic analyzer for now then i will share detail about it as well.

Hi Ali,

This shows the PD is trying to load from EEPROM. In order to check if the configuration and patch are being loaded correctly, you will need to attach a Saleae or another tool that is able to decode into hex values the I2C commands.

Very Respectfully,

Brandon Beader