Looking for K-factor Override Flag in bq3055

Richard,
I am looking into this for you. How much are the cell voltage measurements off? Can you export a senc file from the pack and attach it? I will check a few parameters to see if I can find the problem.

Tom

Hi Tom

The reported pack voltage (4 cells) is off by about 300-350mV.

I'll try and get a senc file in here shortly; in the meantime can you explain how to view the  "K-factor Override Flags"?

Thanks for your help.

Richard

Hi Tom 

I've not had much success with the bqEVAL "software" when trying to obtain a .senc file.

- after putting the pack in to ROM mode it has become unresponsive:

- the PRO screen offers no option to read FLASH memory and shows "COMMUNICATION ERROR" in the bottom left 

I attach a screenshot:

Detailed instructions on how to get round this might help us provide the .senc you requested.

In lieu of the .senc have you managed to find an answer to my original question about where to find the  K-factor Override Flag that the BQ30555 documentation discusses?

Thanks for your help.

Richard

Put the gauge into ROM mode and then go to the Flash Memory pull down menu and export the senc file. I will check on the K-factor override flag.

Hi Tom

the screeenshot I posted was of the bqEVAL software after I had put the gauge in ROM mode.

The software we have version 0.9.75 & I cannot see a FlashMemory drop down menu: I would be grateful if you could provide instructions (a screenshot) to help.

Thanks for your help.

Richard

The Flash Memory drop down menu is shown in the screenshot that you provided in the earlier posting. It is located between the Options and Window drop down menus.

Richard
I found that the K factor voltage calibration does not apply to the bq3055 and it is a mistake in the TRM. The bq3055 uses the normal cell voltage calibration scheme. Can you calibrate the pack using bqEVSW? If you can export the senc file, then I can review some hidden parameters to see if something may have been corrupted. I will be away from the office for the rest of the week, but I can review the file on Monday.
Tom

I've tried to attach an example .senc file here

On all our gauges that I've looked at the "calibration" registers are at the default 20500 value

I take it from what you are saying that all the documentation referring to factory calibration is bogus and does not apply to the bq3055?

Can you give an indication of the expected accuracy if these are left at the defaults without calibration?

Thanks for your help.

Richard

bq3055.zip

The voltage measurement should be very accurate without calibration. I will program your senc file and check to if I can reproduce your results and to see whether there is any data flash corruption. I will be away from the office until Monday, but I will check it after I return.

Hi Tom

- have you managed to get a chance to look at the .senc file?

- can you qualify "very accurate" with respect to the voltage measurements? 

Thanks four your help.

Best regards,

Richard

 

RIchard,
I loaded your senc file into a pack and the cell voltages were off by 20 to 30mV. I checked the calibration parameters and they were set to the default values and I tried calibrating the pack. The cell voltages were within 1mV of the reported values after calibration and the pack voltage was within 1mV as well. Are you able to calibrate your pack? That should solve the problem. I did not find any corruption in the hidden parameters.

Tom

Hi Tom

If this would be via the calibration scheme described in the TRM, then no we cannot calibrate our packs?

We have many packs in the wild (thousands) and would not be able to recall these to calibrate them.

We might be able to issue SMB commands if there was a process to calibrate remotely but we had been under them impression from the TRM that these chips had been factory calibrated.

In order to work around this I would like to get an understanding of how innaccurate we can expect the voltage readings to be? and what we might have to do to cope with this.

I presume there was nothing special about the pack you were testing with, and that we might consider it "typical"
- given this a 30mV innaccuracy would probably in the middle of the range we would expect?
- or is there reason to expect this to be wider?

Do you think that the inaccuracies in the documentation for the bq3055 might be corrected any time soon?
- or is there a list of the parts known to be wrong
- that at least would stop us from being misled towards avoidable issues in the future

Thanks for your help.

Best regards,

Richard

Richard
The device is programmed with a set of default calibration values during production and final calibration should be performed at the pack level. We do calibrate the bandgap, frequencies, etc. I am not sure how much the voltage readings will be off, but it will depend on external components. I mentioned that I measured about 20 to 30mV error in my case. We do correct errors in the documentation and please let us now if you find something that needs to be corrected. The bq3055 is in the process of moving to NRND status, so we may not provide additional updates to this device. It is being replaced by the bq4050, which was just released, but the bq3055 will still be in production as long as there is a demand for it.

Tom

Hi Tom

Please understand that we are really grateful for your assistance: it has been invaluable given that the documentation has been misleading

It would be best all round if we could rely on the information we have and not have to resort to asking questions in the first place
- is there an official channel for raising awareness of the errors
- if there were a repository of known errata for the documentation it would save us and your other customers time and effort
- I imagine it would also cut down on the number of forum posts you need to address

For the record the TRM we are using is SLUU440 dated Jan 2012, this is the same version offered on your website currently

You might remember we  started using the chip in 2014 on your suggestion to avoid the lockup that could occur with the bq3055.
Specifically you suggested that we use this chip because the firmware had been written to prevent this lockup.

Since then we had several queries about information in the documentation for this chip
A very quick search for my records shows questions about:
- the location of the "cal enable" bit and when it gets set
- the requirement to be in "FAS" to access Dataflash
- and this about the section on the K-factor flag and calibration

In each case you have confirmed that the documentation has been wrong
- but there have been no changes to the docs available
- and there is no visibility of any other issues discovered by any other users

Thankfully you have been able to help us through these mistakes we have found (both in this forum and via email)
It would be great if we could share this information with other users of the device

Thanks for your help.

Best regards,

Richard

Apologies, I incorrectly mentioned the lock up of the bq3055 this was incorrect.

It was the bq3060 that would lock up and you recommended we switch to the bq3055 where this had been addressed in the firmware (because there was no intention to fix the firmware in the bq306).

Sorry for any confusion.
All the best,
Richard

Hi again

I wondered if there was any new info about if (or when) we might see a revision of the TRM for the BQ3055?

- or indeed just if there was anywhere that we could get access to known eratta?

We have been fortunate that the latest discrepancy we have identified in the TRM is to our advantage; this was that SystemData(DeviceType(0x0001)) returns 0x3055 and not the 0x900 listed. However it would be greatly beneficial if we had accurate and up to date documentation.

Thanks for your help.

All the best,

Richard