bqEASY learning cycle step 1

The intent of the discharge / rest profile is to collect data during discharge and then to allow time for the device to update Qmax. We like to log data during this period to see how battery responds during the rest period. Cell voltages will increase and will help to diagnose problems, if they occur. It would serve no purpose to start with a discharged battery and let it rest for 5 hours.

ThomasCosby said:

The intent of the discharge / rest profile is to collect data during discharge and then to allow time for the device to update Qmax.

I thought QMAX doesn't get updated during the first discharge??? IT ENABLE is step 2.

The reason I ask if step one is necessary comes from slau597 - "Achieving the successful learning cycle"

"The learning cycle starts with a discharged, relaxed battery (battery voltage is stable and at low
(~3.0–3.3V) voltage). Upon running IT_ENABLE, the gauge measures this voltage and identifies what
state of charge this is associated with using the correct chemistry ID.
At this point, RUP_DIS is cleared, VOK is set and QEN is set.
Now, charge the cell to full"

I was referencing a different procedure than you were using. You were using the one where a close Qmax is not known. In that case you do not have to fully charge the pack before discharging it, but you should charge it until the pack voltage is above the termination voltage. Then discharge it at a C/5 rate until the pack voltage drops below the termination voltage. We recommend waiting for 5 hours to allow time for the cell voltages to stabilize to set the initial DOD point. This will allow the device to make Qmax update during the rest period after the pack has been fully charged.

i'm sorry your answer is confusing.

ThomasCosby said:

I was referencing a different procedure than you were using. You were using the one where a close Qmax is not known.

what procedure is that? I do know what Qmax is.

ThomasCosby said:

In that case you do not have to fully charge the pack before discharging it, but you should charge it until the pack voltage is above the termination voltage.

??? so you are saying I do have to have the battery connected to bqEASY and go through a discharge for step 1? when you say the voltage is above the termination voltage - are you talking charge termination or discharge termination?

ThomasCosby said:

Then discharge it at a C/5 rate until the pack voltage drops below the termination voltage. We recommend waiting for 5 hours to allow time for the cell voltages to stabilize to set the initial DOD point. This will allow the device to make Qmax update during the rest period after the pack has been fully charged.

according to "Achieving the successful learning cycle" the Qmax update occurs after you have charged and discharged. You are saying it happens after charge.

This impedance track stuff is really frustrating to use. The documentation is all over the place. I've never heard of this other method you speak of with a known Qmax. I feel like I need more assistance than forum posts.

Dan,

You can run the learning process using bqEASY or you can run it manually. If you use bqEASY, then you have to follow the process that it recommends. If you run it manually, then you can use one of the processes shown in the Application Book (SLUA404) on pages 208 and 209. Either way will work, but you need to follow them very closely.

Qmax updates can occur during the discharge rest period or the charge rest period.

Regards

Tom

Hello Dan,

I've been going through this process recently and was similarly confused at one point regarding the learning process.

Take a look at SLUA334B http://www.ti.com/lit/an/slua334b/slua334b.pdf. This discusses the two learning processes and defines the two states, no close Qmax known, close Qmax known.

If you assume that no close Qmax is known, then you plug the design capacity for Qmax and follow the first procedure on page 2. If you happen to have a close estimate for Qmax, then follow the second procedure on page 2.

From my experience, BQEASY follows the procedure documented in SLUA334B where Qmax is not known. Either way, if you do or don't know it you can just follow the BQEASY procedure and it will work Qmax out. Be aware, it can take several days depending on the battery capacity. In my case for 2300mA it over 14 hours per day, and BQEasy needed to be babysat at the transition points, so the charger and FET resistive load can be switched in manually.

Maxerror is in the SBS data screen, you can see if things are working when this drops below 1% as SLUA334B notes. Check the Keep Scanning checkbox while it is running.

Regards

Simon Buchwald

Ok thanks for the help. Where can I get the .senc file for my chemID? I have a non standard chemID 1107. The application book states -

"a new firmware file (*.senc) containing the OCV table specific for a particular chemistry
has to be programmed into the pack first. Firmware files for selected chemistry IDs can be download from
the Tools and Software folder of a corresponding Impedance Track™-based device."

I can only find the default .senc file on the website. I currently have another thread because I have changed the chem ID in the bqEASY process but the new ID never shoes up in data flash. Maybe if I started with a 1107 .senc that would solve the problem?

http://e2e.ti.com/support/power_management/battery_management/f/180/t/213794.aspx

There is only one senc file for the device. You can load the chemistry data by selecting the proper ChemID using bqEASY. The OCV table will be saved in the .dfi file that you save to program into additional devices.

good morning

i am newbie on fuel gauges, i am working with the bq27500 evm and ev2300.

i have a couple of question:

1)  i was performing the learning cycle for my battery (350 mAh LiPo) with bq Easy. It takes a very long time (discharege + 5hrs)+(charge + 2)+(discharge + 5) and i'm doing it manually with a resistive load in order to have C/5 discharge current. I can't see a way to save the configuration of flash memory during this cycle and resume the work the day after. Am i wrong? or hopefully there is a way to save the work done avoiding an huge waste of time (it's the 2nd time i'm doing it)? 

2) during configuration wizard with bqEasy is asked the value of the max taper current. I'm not sure if it is the max value of current provided by the battery charger (i put 140 mAh) or it is the value of current at the end of charge. that parameter is one of the 3 necessary to evaluate the SOC for a discharged battery so i would like to be sure about it.

thank you in advance for your help   

Hello Suzumebake,

1. I also could not find a way to save work done one one day for the next, and left my setup running overnight as follows. This process took me about 3 runs over 6 days to figure out. I had various false starts, BQEASY lost SMB bus comms one time, and I had the USB connected on the Dell screen USB hub which I turned off to save power, but lost the days work because BQEASY lost comms to the EV2300, and was unable to restart it because the FET in the EV2300 could not be restarted.

My experience is with the BQ20Z75 and it's version of BQEASY. I found that the  learning cycle could take several days to complete, with the charge, settle, discharge, settle, charge, settle, discharge, settle cycle.

I found that I could do charge, settle, discharge and settle in one day, then did the same on the next day. During the intervening night, I left the PC with BQ Easy on, and the battery system connected through the EV2300, but both the charger and the load disconnected. These were long days, 14 hrs per day.

The setup I used consisted of the following:

1. Target board with EV2300 connected on SMB
2. FET Discharge circuit with C/5 resistor connected to the EV2300 HDQ port, so BQEASY could control the discharge as per Figure 4 of SLUU278.
3. A bench supply set to the charge voltage, and max charge current limit .
4. A DPDT switch, with a center off position, so that at the transition points in the BQEasy control I could quickly and easily switch the load in, or the charger in, or both off for settling. This switch made life easier. The center pole pair was connected to PACK+ and PACK-, left pole pair to the bench supply, VCC, GND, and the right pole pair to the FET load drain and GND.

The BqEasy for the BQ20Z75 which I used had an automated settle and discharge control, so long as the FET circuit  was wired up to the EV2300 HDQ port. 

Early in the morning, 8am I would first use the DPDT to switch the charger in and then I would flick to the SBS screen in BQEASY to watch the charge state. When the TDA alarm was set  I would turn the switch to the center off position, and then go to BQEASY and kick off the the first learning cycle. Just after I kicked this off I would turn the switch to the discharge position. BQEasy then automatically waited 2 hrs for settle and automatically switched the FET on for discharge, and at the set volts for discharge it would switch the FET off and wait the 5 hrs settling time. All I did was monitor this using the SBS screen. At the end of the 5 hrs settling, I would turn the switch to the centre off position, and leave everything on and running for the night, come back the next morning and repeat the process. The fact it had an extra 8 hours settling time didnt seem to bother it.

2. The max taper current I gather is the current at which BQ executes charge termination. For your part it is mentioned in SLUS914A

Table 4.7 - Taper Current Default value of 100 mA

5.6.1 Detecting Charge Termination
For proper bq27500 operation, the cell charging voltage must be specified by the user. The default value
for this variable is Charging Voltage = 4200 mV.
The bq27500 detects charge termination when (1) during two consecutive periods of Current Taper
Window, the AverageCurrent( ) is < Taper Current, (2) during the same periods, the accumulated change
in capacity > 0.25 mAh/Current Taper Window, and (3) Voltage( ) > Charging Voltage – Taper Voltage.
When this occurs, the [CHG] bit of Flags( ) is cleared. Also, if the [RMFCC] bit of Operation
Configuration is set, then RemainingCapacity( ) is set equal to FullChargeCapacity( )

I gather the taper current is actually cell/battery dependant and you can set it depending when you want the BQ to terminate the charge for your specific cell based off the cell charge curve graph. I just let BQEasy pick my taper current itself in the configuration for my cell setup.

Regards

Simon Buchwald

Hello Simon

first of all thank you very much for the answer and for sharing your experience.
So yesterday after a week of work i was able to accomplish a full cycle in the same day (almost) and produce my first Golden file.

1)Well i am not yet completely sure about the procedure but that's what i did:
I discharged the battery the day before. In this way the morning after i was able to start with a dicharged and fully relaxed battery.Indeed in the first step it seems BQeasy just need a battery in that condition so I initialized IT algorithm.
I charged the battery, and after full charge i waited something less than 2 hrs
but i veryfied the status of VOK flag (cleared) and UPDATE_STATUS0=01 before go at step 4.
I started the discharge. Once it is done,i waited (again) this time for 5 hrs and i verified the status of VOK (cleared) and update_status=02
From what i read in (SLUA597) it should be one of the "symptom" of a good cycle.
About my equipment, it is a bit raw, about the load (i used a resistor in order to have about C/5 current) and to charge the battery i used an old board with a BQ24072 (at least i am pretty sure this board works).

2) About taper current, i saw the documentation you suggested, that's more or less what it's written also in (slua432A pg. 5). There is also stated that "Normal Setting: This register depends on battery cell characteristics and charger specifications, but typical values are C/10 to C/20" so i suppose you're right. I put 6.6mA, it is the end current from datasheet.

Now i will proceed with another test with another battery with different capacity. For sure i will have more questions...

Thanks again

Your Update Status to 1 and then 2 indicates that Qmax updates have occurred and your procedures are good.

i was looking in the forum for some support for my project and i just discover i never thanked you for your answer..

I'm really sorry but i missed your post.

thank you.