BQ27425 Learning Cycle, Update Status = 0x34?

John Gassel

If I look at the data block summary for the BQ27425 it shows the Update Status register to have a default value of 0x34.

I'm trying to go through a learning cycle of my battery and I'm finding that it's stuck at this value. Should it get reset to 0 before going through the procedure? It's confusing because so much of the learning cycle documentation talks about IT_ENABLE and QEN, which I don't think this gauge has. Is there something else I need to do before I start the charging and relaxing step, waiting for UPDATE_STATUS to equal 1?

I have log files that I can attach of my charging and relaxing steps if that would be helpful. Thanks.

over 7 years ago

0 John Gassel over 7 years ago

Intellectual 365 points

I just tried to doing a full reset of the device and that didn't seem to change the value.

0 Dominik Hartl11 over 7 years ago in reply to John Gassel

TI__Guru 51346 points

The bq27425 will automatically start learning when it detects a battery. You don't have to issue an IT_ENABLE command. See page 23 in the dataheet (SLUSAI6), POR and INITIALIZATION modes.

You can check the CONTROL_STATUS register if the gauge has learned new values, QMAX_UP (=1 when it learned a new Qmax) and RES_IP (=1 when it learned new impedances).

0 John Gassel over 7 years ago in reply to Dominik Hartl11

Intellectual 365 points

Thanks for the reply Dominik That's interesting. Okay, I'll have to check for those two bits specifically. Before leaving today I just did a full reset and started a discharge cycle. Hopefully that'll have good results in the morning.

So I can just ignore the UPDATE_STATUS bit in data flash as it's meaningless on this device?

0 John Gassel over 7 years ago in reply to John Gassel

Intellectual 365 points

Well, the cycle ran overnight and it didn't learn any new values unfortunately. Is there some other step that I'm missing? Should it be in CFGUPMODE? Do I need to do a soft reset to clear ITPOR?

Really not sure what else to try here. Here are my current status and flag registers.

Also attached are logs from the last 24 hours.

FuelGauge.zip

0 John Gassel over 7 years ago in reply to John Gassel

Intellectual 365 points

Issued a soft reset and ran a charge cycle so far today. Found that QMAX_UP has been set and I have a new value for "Qmax Cell 0" in the flash. That's good progress. I'll let it relax for a few hours and then do a discharge. Is there anything else I need to do before discharge to make sure that I get a RES_UP result? Thanks.

0 John Gassel over 7 years ago in reply to Dominik Hartl11

Intellectual 365 points

Dominik, I'm able to get the QMAX to update, but I can't seem to get the RES_UP bit to go high and get a new RA table.

Could you or someone else look at my logs? I'm really not sure what else to try. Thanks!

2016-08-25_learing_cycle.zip

0 John Gassel over 7 years ago in reply to John Gassel

Intellectual 365 points

Reading through the document, "Achieving The Successful Learning Cycle" SLUA597 section titled "What can Go Wrong, Hot to Diagnose Problems, and How to Solve Problems" it seems I'm doing everything right.

Based on (e) in that section:

(e) Discharge C/5 Constant Current: Make sure to use a C/5 current. This is preferred, and there can be some error. However, if the current drifts too high or too low, the cycle can fail. Too low is around C/10, too high is around C/3 to C/2. Smaller cells (<800mAH) are much less forgiving in this regard.

I'm going to try and pick different currents. Although I was already using C/5 discharge twice without success. Any guidance would be greatly appreciated! Cheers.

0 John Gassel over 7 years ago in reply to John Gassel

Intellectual 365 points

The discharge at a different constant current still had no luck. Please help.

0 Dominik Hartl11 over 7 years ago in reply to John Gassel

TI__Guru 51346 points

Your cell is fairly small (note the comment about small cells in the learning cycle app note). What currents did you try? Did you go all the way up to C/3?

0 John Gassel over 7 years ago in reply to Dominik Hartl11

Intellectual 365 points

The highest I've went is ~ C/4. I'll try C/3 tomorrow.

0 John Gassel over 7 years ago in reply to Dominik Hartl11

Intellectual 365 points

That did the trick! Thanks Dominik. I was able to get the learning cycle to work and update the resistance tables using a discharge current of ~ C/3.

Power management

Power management forum

BQ27425 Learning Cycle, Update Status = 0x34?