Further BQ20Z75 Golden File Creation Issues

SImon,

1) An Update Status = 0E is a valid value for the bq20z75 v181 FW.. It indicates that a field Qmax update has occurred. I am not sure about your comment that the bqEVSW does not accept the 0E. it works on my system. You should change this value to a 02 before creating the golden image anyway.

The golden image processes that you noted should all yield good results when generating the golden file. As a general rule, you should use the document with the most resent update if you find discrepancies.

2a) The senc file that you noted is the correct version. 

2b) The comment is from an older applications note. The newer devices only have one version of the senc file that covers all ChemID's.

3) You need to issue a Reset command (41) to initialize the data in the RAM. The Reset command (41) is fairly well documented in the TRM.

3) After you have reloaded the senc file, gg file, WRITE ALL, and RESET, then you are ready to create the .dfi file. It is a good practice to make sure that your Chemistry file is still selected, so check that in bqEASY. You can then go to the Setup screen on bqEASY and save the dfi or ROM file to use to load into other packs.

Regards

Tom

Hi Tom,

Thanks for your reply, as you might gather I am finding this process quite frustrating, and I had been at it for 18 hours straight yesterday when I wrote that post so pardon me.

Nevertheless:

1. The bqEVSW version I have is bq20z75v1.18, I downloaded it directly from the web site address in the last post about 10 days ago, is it not the correct version??? Anyway when I click on All Done from 5B.Learning Cycle 2, bqEasy responds with a dialog box. I cannot seem to copy the graphics in here, but the text reads, and I quote verbatim  

"Dataflash.Gas.Gauging.State.UpdateStatus does NOT = 6. Impedance track learning cycle was not completed correctly. Verify procedure and repeat".

The fact that works for you is nice, but you have a customer in the field, me, for whom it is not working. I am not doing this for fun. We have a customer who is waiting on us to complete the design, and they want to produce up to 10,000 units per month using the TI chip.

1a Do you have the exact same version of bqEasy?
1b Is there a special version you have that is not on the web site? 

I have actually been taking screen snapshots of the entire process with the view to writing a more complete application note on how to do this from scratch as I have done, and submitting it to TI to publish. This I cannot complete because I can not get past the above-mentioned dialog box error.

2. Thank you for clarifying the senc file. Put yourself in my shoes and ask whether there is any way I could have known this without incremental experience on the product portfolio. 

3. I challenge what you say there about the reset command being fairly well documented in the TRM. I did a search in SLUU265A for the BQ20Z75, and it definitely is NOT in that TRM. Given what you said I did a google search for the terms "41 Reset BQ20z" and google found it in SLUU310, the TRM for the BQ20Z60, and it is documented there. Again, in my shoes should I have thought to go back through all the past  BQ20Z reference manuals to find information that has been omitted or deleted from the BQ20Z75 manual? Probably not.

Thanks for your second point 3, this is the clearest explanation of what to do next, I wish you had that in an app note somewhere. I will go try to complete the process doing what you suggest and let you know how it goes. By the way, did you get to look at the gg file and is it ok?

Regards

Simon Buchwald

Simon,

I am sorry about all of the problems that you are having. Let's try this again. I did review your .gg file and your parameters look reasonable.

1) Your .gg file indicates that you are using the bq20z75-V180. The firmware is actually v181, but the part number was released as v180. I am not familiar with the bqEVSW version bq20z75v1.18 that you listed. Maybe it was a typo. You should be using this one from this website.  http://www.ti.com/product/bq20z75-v180

The download file is located under Software. (It is the last file.)

I did not try running the learning cycle with bqEASY, but if you got an Update Status = 0E, then the learning cycle updated Qmax and the Ra-table and it is okay to proceed.

3) The Reset (0x0041) command is covered on page 50 of the bq20z75 TRM document that you noted and "full reset' is covered several other times. I am not sure what else you are looking for on the reset command.

I hope that this helps.

Regards

Tom

Hi Tom,

Thanks for working back on this and reviewing the gg file, I see its 7:55pm your time. I have a customer demo next Wednesday, and after 2 weeks on this, still no golden file.

So just to clarify versions:

1. The senc file I used is the following: bq20z75_0181_bld_0003.senc

2. The bqEasy software is in a zip file, sluc305.zip, directly from that link you showed, I down loaded it on 21 August 2012. Inside is the bqEVSWSetup00.09.79_bq20z75v1.81.exe install file, which at the time I ran and installed BqEasy. (As an aside, I had some install issues with it on another machine running XP SP3, where it would not recognize the device driver for the EV2300 but it runs on this one. Both machines have XPSP3 and are Dell Dimensions). What I wrote was definitely NOT a typo, so you may choose to familiarise yourself with what is inside sluc305.zip, because its what TI provides to generate the golden file for the BQ20z75 V1.80/V1.81.

So I have reviewed the golden file instructions once again, and explored BQEASY some more to see how the golden dfi file can be written out. I can NOT see any option other than "5C. Update Golden Pack". At this stage I am really stuck, as I cannot get past the error I noted in my previous post where bqEVSW 00.09.79_bq20z75v1.81will not recognize the 0E status.

My point about the Reset command was to re-inforce that the documentation is really woeful for this product range given it is so complex, and that as a customer I am not in a position to second guess where all the relevant information is given it is so disparate on the TI website, various forum posts.

Can you find out what I should do please?

Is it possible to escalate this issue to someone in the software/design team? Where are they all?

Regards

Simon Buchwald

Tom,

Further to the last post, I thought to attach the screen shot log of the process for you to see. From page 35 onwards shows the 0E status and error screens.

Can you find somebody at TI help me please?

Regards

Simon Buchwald

Simon

You are making this more difficult that it needs to be. Since your Update Status = 0E and bqEASY is not accepting it. Please try these steps.

1) Go to the Data Flash screen. Press "Read All". Then File >> Export and save the .gg file to the desktop.

2) Edit the ,gg file with Notepad. Change the Update Status from 0E to 02. Change the Cycle Count to 0. Save the file.

3) Go to the Pro Screen. Use the Write SMB Word section. Set SMB Command to 00. Word = 0f00. Press the Write button to put the device in ROM mode.

4) Use the Srec programming section and find your senc file. Press the Program button and load the firmware.

5) Use the Send SMB Command and set the SMB Command to 08. Press Send to take the device out of ROM mode.

6) Go to the Data Flash screen. File >> Import the .gg file that you modified. Press the Write All button.

7) Go to the SBS screen. Enter a 41 in the Manufacturer Access field and press Enter on the computer.

8) Go to the bqEASY screen. Go to the Chemistry section and make sure that the proper Chemistry is selected.

9) Go back the Setup section. There should be a Read dfi button. Press this button and save the dfi file somewhere. You can do this with the Read ROM button, if you want to save a ROM file. These files are your Golden files.

10) If you want to load a Golden File into a device, then use the bqEASY Setup screen. Select the radio button that you will use a dfi file. (I think that it is the middle one.) Find your Golden file. Press the button that loads the dfi file into the device.

11) Go to the SBS screen. enter a 41 in the Manufacturer Access field and press Enter on the computer to initialize the RAM. Enter a 21 into the Manufacturer Access field and press Enter on the computer to start Impedance tracking.Check to see if the VOK and QEN flags turn red under the Operation Status section.

12) Go to the Data Flash screen. Press the Read All button. Select the Gas Gauging tab and check to see if the Update Status = 06. The pack is running now.

I am listing these steps from memory, so let me know if something does not work and I will correct it.

Good luck.

Tom

Thank you Tom,

The GG file I sent in a previous post include was at step 2.

I will go try the rest of your steps and let you know how it goes.

Again I wish this was documented so clearly in an app note somewhere, I am glad you were able to extract it from your memory to tell me.

Regards

Simon Buchwald

Hi Tom,

I managed to follow your steps and save the golden dfi file, and also program up a board and initialise it. One final small problem with the prototype PCB, the charge FET is not actually turning on but its because of a bad solder joint to the gate. Tomorrow I will redo the programming of the golden file in a fresh pre-production board and see if it all works.

A couple of odd things happened during the .senc file that I have not noticed before:

1. Several times the following dialog box text appeared, Error Programming Flash! Error code = 772, No Acknowledge: VB_T2H_NACK. I have had problems with the I2C lines to the BQ on the target board in the past so I shortened the cable to about 10cm. This seemed to make it more reliable. Any comments or experience here?

2. After loading the .senc file, BQEASY says: To avoid Damage to the Gas Guage, remember to restart the EVSW before updating the Data Flash! Can you tell me the significance of this please? Can I just ignore it or not?

Regards

Simon Buchwald

Simon

You may see NACKs, if the setup is not  clean. That is probably the case since it improved with shorter cables. You can also see them when going in and out of ROM mode.

The program recommends closing and reopening  bqEVSW after loading the senc file to make sure that they are synced up. You may have loaded a newer version of FW and addressing would be different. That was not the case with your pack, so I did not have you go through that step.

Tom

Hi Tom,

I ran through the steps, and fixed up the MOSFET issue, issued the 21 and watched what happened. The gas gauge and time to empty/full seems not to be working correctly. Attached is a file with screen shots of progress with explanatory text. I don't understand what's happening, can you help please?

Cheers

Simon

Simon,

Your charging current is displayed as a negative current, where it should be positive. You must not have not attached the proper sign to the current when calibrating the pack. You will need to recalibrate the current. Gas gauging worked when you first started, because the calculation was based on an OCV simulation of the cell voltages. It quit working, because the current measurements confused the gas gauge. It started again after the reset based on another OCV simulation.

Tom

Hi Tom,

Finally it seems that I have the BQ20Z75 fuel gauging. It turns out that on the production prototypes the  P and N to the current sense resistor was back to front, so the chip was reading negative current for positive. The circuit reviewer accidentally flipped the sense during the design review and noone picked it up.   The circuit reviewer accidentally flipped the sense during the design review and noone picked it up. Also had some COV and  CUV faults but tracked that down to hairline shorts between cell pins on the BQ20Z75.

One last question, relating to the INIT flag. The documentation says it is cleared when the part is operating, but it always seems to stay set, even though I get an 06 status in the GG dataflash screen, and the QEN comes up and it seems to be gas gauging.  I attached the relevant SBS screenshot from when it first came up.

Hope you can help again.

Regards

Simon Buchwald

Simon,

in the case of the INIT flag, a red flag means that the part is initialized and ready to run. It should only stay green for about 1 second.

regards

Tom

Hi Tom,

 Thanks for clarifying that. It’s just the documentation says the flag is cleared on successful init, and I see the flag being set.

If the flag is set on successful initializing as you say, then all is good, except for the documentation not matching the device again.

As I mentioned I have almost enough screenshots and text to submit a how to to guide for the BQ20Z75 for newbies with setup diagrams and photos, are you or TI interested in publishing it?

Regards

Simon

Simon,

The TRM does say that the INIT flag is cleared after Initialization, but it also says that 1= Initialization to adhere to the SBS Specification. A "1" is a red flag in the bqEVSW world, so a green flag is a not-cleared state in this case. 

Tom

Tom,

To me generally speaking if a bit is set, it is a one; and BQEasy shows red. If a bit is clear, it is set to zero and it shows green. To me cleared means 0, or green, and set means 1 or red.

If there is another meaning just for this case so be it, but you answered my question that set means cleared in this case. I get what you are saying, but it really sounds like double speak to me.

To whom should I email my BQ20z75 draft app note if you are not interested?

Thanks

Simon