USB-PD-CHG-EVM-01: TPS25750 Application Customization Tool generates bad binary

Hello,

We will get back to you by next Monday.

Thanks,

Mike Emanuel

Hello Mike,

Any update?

Thomas,

Hi I'm Chuck and I support the TPS25750D.

We have identified a bug in the configuration setup for the 7.02 revision of the GUI that is causing the failure that you are seeing.  We are actively working on fixing it now.

I will update you once we have a fix in place.

Regards,

Chuck

I think I may be experiencing the same issue. My dev kit worked fine until I re-flashed it using the TPS25750D GUI. USB negotiation works but charging does not. Hoping to hear of a fix soon. If not, do older versions of the GUI work? Prior to 7.02? Thanks

Tony,

We are in the final testing of the GUI update that will resolve this issue.  

I do not have an exact date commitment, because it is related to when the firmware completes the regression cycle which is targeted for 10/8.  

After the firmware is fully blessed, I need a couple of days to validate that everything is configured correctly, so it should be middle of next week if everything goes as planned.

Hey Chuck, I also received this EVM this week and I'm faced with the same Bad Binary from the online GUI 7.0.2 tool.

Do you realistically think you will have a GUI producing good firmware by this Friday?

Looks like this fix has been in the works for months now.

I'm hoping so since were all stuck until you guys get this resolved.

Thanks!

Hi Chuck, if the fix is going to take a few days, would it be possible to create us a binary if we give you our requirements as per the GUI interface? Charge current, USB sink and source power, etc? Unfortunately we relied on the "no firmware required" for this project and it's now holding us up. I have also been speaking to Mike Emanuel about this project. I was going to get him the BQ25792 register contents but I can't find my EV2400.However, it sounds like you already know of the issue.

Ideally I am after the following question configuration:

1. - EVM mode (top left)
2. - 45W source
3. 60W sink
4. Power Sink
5. Host & Device
6. USB 3.2 Gen 2
7. No
8. No
9. BQ25790 or BQ25792
10. 16.8V (4S Li-Ion battery)
11. 2.5A battery charge current
12. 0.08A
13. 0.4A

It would be much appreciated.

Many thanks

Tony

Yes this Binary would be helpful for me also if your able to supply us with Binaries that work.

I will look into generating this later this week,.  Mike and I work closely together.

I need close to the same binary if possible.

Ideally I am after the following configuration:

1. - EVM mode (top left)
2. 60W source
3. 60W sink
4. Power Source
5. Host & Device
6. USB 3.2 Gen 2
7. No
8. No
9. BQ25790 or BQ25792
10. 16.8V (4S Li-Ion battery)
11. 3.5A battery charge current
12. 0.04A
13. 0.52A
    
    
14. Under Advanced Settings can I change the Advertised 20V Source Current to 2.5A max vs 3A? *The 3A @ 20v Output when using a 4 cell 5Ah pack causes the BQ to overheat and cycle on/off. 20v @ 2.5A output works just fine so I'm hoping I can set the output to throttle to 2.5A max. Possible?

It would be much appreciated.

I can match this setup once I get the final firmware from regression testing.

Could I make one correction to my spec, as my cells can only support 2.75A drain. Can I therefore limit to 27W source capability

1. - EVM mode (top left)
2. 27W source - CORRECTION HERE
3. 60W sink
4. Power Sink
5. Host & Device
6. USB 3.2 Gen 2
7. No
8. No
9. BQ25790 or BQ25792
10. 16.8V (4S Li-Ion battery)
11. 2.5A battery charge current
12. 0.08A
13. 0.4A

Look forward to the binary - many thanks.

Tony

Tony,

Based on the current state of regression, I am hopeful to get the release firmware tomorrow so that I can build this configuration for you.

Regrads,

Chuck

That's awesome news!

I spent a few hours playing around with the EVM last night before putting it away on the shelf until the fix is released.

I tried dropping the Advertised OTG current @ 20v output to 2.5A but it does not throttle the current output to 2.5A so I don't think throttling the OTG current output is possible based on the testing I have done so correct me if I am wrong.

Based on that info I want to update the Binary config request I posted above to this:

• - EVM mode (top left)
• 60W source
• 60W sink
• Power Source
• Host & Device -
• USB 3.2 Gen 2
• No
• No
• BQ25790 or BQ25792
• 16.8V (4S Li-Ion battery)
• 3.5A battery charge current
• 0.04A
• 0.52A

Let me know if you have any questions. Thanks

The OTG current limit is one of the issues that is being addressed in the next release.

Next release meaning the one you are working on having done by tomorrow or a separate release after this current fix is ready?

I tried changing the OTG current settings directly in the BQ25792 charger but they made no difference. I assume the charger will be the chip that needs to throttle current.

Keeping my fingers cross that the new binaries will be ready tomorrow regardless.

Thanks for being responsive!

The release that is scheduled for tomorrow will have this fix, but the release date has a high risk of slipping.

I can help with doing the configuration for you once I get access to the internal code.

Thanks for the clarification.

As long as we can get a new working binary from you then that's all that's needed for us to keep pushing forward with this project.

I assume the high risk of slipping is regarding getting the new firmware integrated into the online GUI?

Not trying to drive you crazy but we have designed a portable battery pack with 2 independent TPS25750 chips and chargers, the firmware is holding us back from moving forward so the sooner the better on a working binary.

Thanks again for your responsiveness.

The GUI integration is the risk.  I am working on it in more detail because of the issues with Charger integration.

I will be doing configurations for customers myself with our developers tool until we get the new GUI operational.

My last report from our regression was that we were at 95% coverage with no failures.  I am still waiting for our internal official release.

Hey Chuck! Hope you had a good weekend!

Did your partners in India get the firmware wrapped up for us so we can get those working binaries today?

I have not gotten the formal release message, so I cannot provide you with a production worthy binary.

I have a call scheduled tonight to get this resolved.

Any good news from your phone call with the firmware development team?

Ryan,

We identified one more fix that was required.  They are retesting tonight and I hope to have a production binary tomorrow.

Great to hear, tomorrow would be awesome! Fingers crossed 

Thanks for the update.

Hey Chuck! Is today the day

Ryan,

Is your release for production or testing?

I can do a configuration with a short term patch tomorrow that is not rated for production or our final release firmware will come out on Monday.

For now just testing.

I have built custom boards based off the EVM design and I have the EVM for testing. 

So non-production Binary is perfectly fine.

I'll test it out and let you know if I find any issues with my testing also. 

So far with the current binary I can power everything but the latest 2020 Macbook Pro M1, but the 2016 Macbook Pro charges just fine. 

I'll test again with the latest Binary you send over today and see if that is resolved. 

So a Binary with the following settings would be ideal:

• - EVM mode (top left)
• 60W source
• 60W sink
• Power Source
• No Data Needed - Portable Battery Pack
• No Data Needed - Portable Battery Pack
• No
• No
• BQ25790 or BQ25792
• 16.9V (4S Li-Ion battery)
• 3.5A battery charge current
• 0.04A
• 0.52A

Let me know if you have any questions. Thanks

Ryan,

I will be off tomorrow morning, but I can generate this configuration fiie for you on Monday when I get back

Man you guys have been working on this for almost a month now

Either the problem is really bad or somebody is slacking on the job

Really excited to see this working so hoping Monday works out for a working binary. 

Thanks for the daily updates Chuck. 

Hi Ryan,

Thank you for your patience. Chuck will provide you with some additional feedback next week as he is currently out of office.

Thank you,

Hari

Hope you had a good weekend Chuck!

Any progress on that Binary for us today?

I will be completing binaries tonight.

I'll be up late. If you can email me the BIN 

Rwb.Ceo@Gmail.com

Hey Chuck!

Those Binaries ready?

Thanks!

RyanBrown.zip

Ryan,

Please let me know if I missed anything in this configuration.  I disabled all GPIO so they do not impact your schematic,.

I have no idea how to use that file and when I look at the text it looks to be refrencing a different chip the TPS65992D.

Here are the settings for creating the Binary for my setup:

Thanks!

Ryan,

Attached is the correct binary.  My zip was wrong.

RyanBrown.bin

Please let me know if there are any issues with this file

Hi Chuck, 

It seems you are now able to create binaries for us. Please could I request this one?

1. - EVM mode (top left)
2. 27W source - CORRECTION HERE
3. 60W sink
4. Power Sink
5. Host & Device
6. USB 3.2 Gen 2
7. No
8. No
9. BQ25790 or BQ25792
10. 16.8V (4S Li-Ion battery)
11. 2.5A battery charge current
12. 0.08A
13. 0.4A

Look forward to the binary - many thanks

Tony

Hello all,

I saw a couple "TI Thinks Resolved" marks on the uploads of custom binaries for Ryan. Not sure how "TI Thinks Resolved" works, but I will not consider this problem resolved until the online GUI is fixed and usable.

To work around this problem I've edited my fullFlash binaries, replacing occurances of  0x16,0xC0 (the last command seen in my original I2C dump) with 0x0A,0xA3 to fix REG0A for my configuration. REG16 on the BQ25792 resets to 0xC0, so that register write was unnecessary for me.

It's been over five weeks since I reported this issue... I wonder if this could have been resolved quicker if the GUI were open source.

Hey Chuck,

I loaded the binary and checked the settings and they are not what I had listed above this post.

I had the charge voltage set to 16.9v the file sets it at 16.8V.
I had the charge current set to 3.5A the file sets it to 2.7A

I'm not sure what other settings may be different but it does now charge when a 60w Apple USB C brick is plugged into it automatically which it didn't before.

But now it will not charge my old 2016 or new Apple Macbook Pro M1 laptops, it just constantly cycles ON/OFF and the LED on the EVM flashes at 1hz.

Now when I plug in my Samsung Note 10+ phone it will only charge at 5v and will not negotiate the normal 9v - 14w charging mode.

It looks like now it only advertises 5v and it will not negotiate voltages above 5v.

Can you check the file and make sure it's advertising all the default higher voltages above 5v @3A?

Hopefully the Bin File is just mis-configured.

Thanks!

Ryan,

I am updating your project here.  I have 3 configs in this thread and I crossed up some settings and transcribed one value incorrectly from my settings calculator.

0743.RyanBrown.bin

Tony,

Attached is the binary to match your configuration.  I disabled all GPIO except for the USB connections.

TonyHannallah1.bin

Thomas,

The TI thinks resolved flag is used when the specific question asked as been resolved.  

In this case, there are 3 customers intertwined in the same thread which makes tracking our our system very difficult.

I acknowledge that this has taken a very long time to address.  The release was not just the GUI and involved several teams at TI to get to resolution.  I am doing configurations for customers so that you can make progress even though our GUI is not ready.  

Thanks Chuck,

I can see the charging settings have been updated and charging automatically starts but the EVM will only supply 5v output when plugged into Old & New Macbook Pro's and a Samsung Note 10+.

When plugged into the Macbook Pro M1 it starts sourcing current from the laptops USBC port.

Do you know if it's setup as a Sink or Source device in the settings you used?

All it can do now is recharge the battery and supply 5v output on the USBC port to charge my phone at 500mA.

I'm looking for the settings below:

Ryan,

It should come up as a prefer sink.  I have updated the configuration to attempt to always become a source.

This should allow your battery back to prefer to source power.  We can talk about other possible configurations using an MCU later if you'd like.

RyanBrownS2S.bin

Thanks Chuck!

Unfortunately I get the same results. The battery chargers automatically from a 60w Apple USBC Brick with the correct settings.

USBC power negotiations are not working at all now, it only supplies the 5v output to laptops and phones.

I'm testing this on your TI EVM kit, are these binaries working on your TI EVM's for both power output and charging input?

Ryan,

I will put it onto my EVM tomorrow to verify.  I don't have the same devices, but I will confirm with what I have.

That will be great.

Looking forward to see what you find out.

Is there any expected release date for the official fix or are we still trying to figure out how to make the chip combo work after all this time?

Ryan,

I will have to delay this until Monday.

I do not have a release date on the GUI yet.