bq 27742 and mamagement studio software

Hi Joe,

Can you confirm the following questions:

-What firmware version of the device are you using?

- what bqstudio version are you using?

You can reach out directly to me via email as i believe we have had correspndence in the past.

 

thanks

Onyx

 

I revisited the management studio software and installed the latest version today (.36).  We have a customer product that uses the bq27742 for fuel-gauging and safety.  I find that after I installed the latest version today, the software works worse than the older version (.24).  It is buggy, the registers don't update, after programming scans do not update register values so I cannot verify quickly whether or not the flash was properly programmed, the register views changed and I cannot change them back.  Even simple voltage and current readings do not update properly.  I tried uninstalling and reinstalling the software with no success.  I am incredibly frustrated using this program to try to create sample product for our customer, that I don't know what we are going to do when we have to build thousands of batteries in production.   It takes forever to program and calibrate devices.  Sometimes the calibrations don't take, which may be the device and not the interface.  I am frustrated to the point where I would consider scrapping the whole 27742 design altogether.

I would appreciate some support on trying to resolve some of the issues I am having with the 27742 and management studio.

--Joe P.

Hi Joe,

I understand your frustration with the tool and I apologize.Can you try disabling the dash board to see if your registers start updating correctly. Another customer reported a similar issue to us and we are working to fix it. In the meanwhile disabling the dashboard which is the window on the left screen should allow your registers to get updated. I also observed the issue with calibration while working on a board yesterday. I saw that bqstudio had updated my cc gain and cc delta to a very high value and wouldn't let me calibrate afterwards. i had to manually change those parameters by hand to a close enough value to my sense resistor and then recalibraed. I have forwarded these complaints to our tools team and they are working to fix this.

thanks

Onyx

Hi Joe,
There is a slight update to the calibration issue. To ensure successful calibration, you can only use a discharge current. A charge current will return cc gain and cc delta values significantly higher than what they ought to be. This is an inherent limitation of the gauge and not the tool. I know our previous gauges allowed either a charge or dishcharge current for calibration, however for this device only a discharge current should be used.
thanksOnyx

Hello Onyx,

Thanks for the quick reply.  I am calibrating with discharge current.  Actually, current is the one that provides the fewest problems.  I am running management studio under Windows 7 Professional with an HP laptop and an EV2300.  I have proper pullups on the I2C lines.  We are using management studio to muddle through until I can complete a new test process.  Here are some observations:

1)  I have to calibrate temperature first, otherwise other calibrations become less accurate after calibrating temperature.  Even though temperature calibration is listed 3rd.  I presume there is a temperature correction on measured parameters in the device firmware.

2)  Sometimes the offset calibration does not properly null the current, and has to be repeated more than once.  Given that board offset takes 30 seconds, having to repeat it is incredibly annoying.  Especially when dealing with a large volume of batteries to be tested.  Any idea why this would happen?

3)  I have to do the calibrations one at a time using management studio.  If I check more than one item, often I have to repeat one or more calibration item, because the calibration was not accurate.  Again, has anyone at TI noticed this also?

4)  Management studio will often crash after programming a device for no understandable reason, and the program has to be re-opened.  After re-opening, I am unable to notice any issue with device programming when I read the data flash. 

5) I will occasionally have a problem where, for no discernible reason, the program refuses to calibrate and I get errors.  Restarting studio resolves the problem.

6)  Sometimes, when refreshing or scanning, register and parameter values do not update and I have to restart the program.  With the latest version of management studio, this problem happens often.  When I re-open the program, calibration and register values are as I expect.   Again, this is incredibly annoying.  A minor complaint is that sometimes the display of parameters changes and it is not intuitive, or even possible it seems, to change display format even when I select restore defaults.  One time I had to uninstall the program, restart the laptop, and re-install the program.

7) With older devices on the SMBus interface, I could flash them in one second.  Here, it takes 45-60 seconds to write the .srec file.  In production, that's going to be a serious bottleneck even without all of the other issues.   Is there a faster way to flash these devices, and would TI be able to provide code samples like they did for the older devices to assist with the details of programming, as I re-invent our production processes for these new devices?

Thanks,

Joe P.

Hi Joe,

Thank you for enumeratng the issues you are facing clearly. I have passed this info accross to our tools team. Indeed I have been able to notice some of the issues you pointed out. The current work around for most of them seems to be disabling the dashboard. Our recent experiments has been able to narrow this down as the cause of most issues.

All our single cell guages have always operated using I2C communication. Or multi cell gauges use smbus in accodance with SBS specs whichi guess is what you are used to. Programming of our single cell gas guages is often done by customers using a flash stream file and a host controller. This may perharp be faster than programming using the srec.  I have attached sample code and there is a link below for android /linux drivers incase you want to go that route.

 

 

The attached code is for a different device but the procedure is analogous.

 

 FlashStream028_bundle (2).zip

Onyx,

Thanks again for the quick reply, passing on the observations, and for the sample code.  I will let you know how we make out.

Best,

Joe P.