Golden image of non rechargeable battery

Roger,

a few tweaks to your methodology:

1) You only need to wait 5 hours after the battery has been discharged to before you click the "discharge/wait done" button.  Therefore, if you have discharged batteries that were used days/weeks/months prior, use these and you won't have to wait 5 hours.

2) There is some learning that occurs when the gauge observes a charge, and you will have to 'fake' this part of the cycle when you replace the discharged battery with the fully charged battery.  This means that you will have to update the Data Flash manually (using the Data Flash window) before you start discharging your battery.  The basic procedure is outlined below:

a) Replace the discharged battery with the new battery.

b) Go to the  Data Flash Screen and update the following:

"QMax Cell 0": Max capacity of your battery in mAH.  You will need to measure this parameter prior to running your learning cycle.  I will describe how below.

"Update Status 0:" Make this parameter  a '1'.  This indicates to the algorithm that the the qmax value is correctly learned and we can learn resistances during the discharge now.

c) Start discharging at C/5 and continue as you had described.

With regard to measuring QMax, you will have to log a discharge using the EVSW.  Basically, you will do a roughly C/5 discharge similar to the learning cycle discharge.  However, what you will need to do is set up a log on the Data RAM screen.  This will record the battery current/voltage during the discharge. 

When the battery terminates, you will have to integrate the current measured by the gauge to measure QMax.

There is one question which I need to investigate about this process.  I am not sure if you can reset the gauge during the learning cycle.  A power cycle during the replacement of the battery will reset the gauge, and I am not sure if this is acceptable.  I will look into it.

Let me know if this helps,

Charles

Roger,

While it looks like you and Charles are on the right track to get one of our Impedance Track gauges working for this primary lithium battery, another alternative is to take a look at the bq2052.  It's an older product but may meet your needs or could be a backup plan.

http://focus.ti.com/docs/prod/folders/print/bq2052.html

Best regards,

David

Hi David

   Thanks very much for your input.

Problem is that it is pretty much too late for us to change the design at this point in time.

The other factor is that we use 2 different battery types in the same system. One rechargeable and the other non rechargeable...

ie. If I did get it to work with the primary batteries, it would be equaly difficult to get it to work with the rechargeable.

Best regards

   Roger

Hi Roger/Charles,

In order to use non-rechargeable primary battery (LiSOCl2) in our system, i have some question about 1) the chemistry select cycling and 2) the final learning cycle.

In particular, from the Roger posts, i suppose you have rightly done the chemistry select cycling: so ... how can i bypass the charge phase of that cycle?

Secondly: have you found the correct step to realize the final learning cycle with non-rechargeable batteries?  Can you post the correct way?

Thank you very much,

Davide Gavioli

Hi Roger/Charles,

With method you suggested for Li-SoCl2 batteries, have you been able to estimate battery capacity?

I am using Li-SoCL2 battery in my system and bq27510 part.

Thanks in advance,

Regards,

Khodidas

Hi Charles,

Could you not uptdate "QMax Cell 0" with the nominal capacity value that the manufacturer provides?

Best regards, and thank you very much,

To all who are trying to use our Impedance Track gauges with a Primary cell:

I think we all acknowledge that the firmware was not designed to work with a primary cell, but even considering the workarounds discussed above there is a major stumbling block.  The issue is that the OCV (open-circuit voltage) curve of the cell needs to be known.  That means you need a new chemID created for your cell.  We have tried to use our typical method to create a chemID with no success.  There are many issues, including the extremely long relaxation times required for most of the chemistries to get a stable OCV, the fact that some of them actually have an INCREASE in voltage at higher depth-of-discharge, and extremely large internal impedances of some primary cells.  The bottom line is that we haven't been able to create a chemID for any primary cells and since their voltage curves are very different from the standard Lithium rechargable cells, you will get wildly inaccurate results if you try to use an IT gauge with a primary cell.  I'd be interested to hear if anyone has been able to make it work on their own, though.

The best solution at the moment for gauging your primary cell is to use a coulomb counting gauge that doesn't rely on knowing a voltage curve.  This app note is one example:

http://www.ti.com/lit/an/slua367/slua367.pdf

Thank you very much for this answer dMax.

My question now is: is there any way to do something like the slua367 proposes  with the bq27501? I haven´t seen the ILMD register on this fuel gauge. Is there any specific way to do something like you are talking about with this particular fuel gauge?

Thank you very much in advance