• State TI Thinks Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 2 answers
  • Subscribers 64 subscribers
  • Views 308 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Original question:

BQ40Z80: BQ40z80 gauge

BQ40Z80: Some question about BQ40Z80,BQ4050,BQ78350

Gabriel Wang1
Genius 4715 points
Part Number: BQ40Z80
Other Parts Discussed in Thread: GPCCHEM, , BQ4050

Dear Expert

As for the questions I told you over the phone, please help me confirm them:

1. The minimum current of full charge is 80 seconds, can it be shortened?Could it be 20 seconds or 10 seconds?

2. Is continuous discharge required during cyclic learning?Does charging for a short time during discharge affect cyclic learning?

3. Will the SOC deviation of coulometer jump from 2% to 8% or 100% after the factory cycle learning and shipment?Does the user need to repeat the loop learning process?

  • 0
    TI__Guru** 105760 points

    Hello Gabriel,

    Is the someone previously supporting this case I can loop in from a phone conversation?

    The minimum time for VCT is 2 40 second periods, if they want to just set FC then they can use RSOC or voltage thresholds.

    The must have continuous discharge during learning, you cannot charge during the learning cycle because it will either disrupt the Ra table learning or disqualify the Qmax update.

    I'm not sure about the final question, you only need to perform one learning cycle on the pack, you don't need to perform it on every pack. Many things could cause an SOC jumps.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    Genius 4715 points

    Dear Wyatt

    I am nor sure this case support by other colleagues.

    I can arrange a meetting with customer through webex if u have time.

  • 0 in reply to Gabriel Wang1
    TI__Guru** 105760 points

    Hello Gabriel,

    Were the questions answered or did they have more follow up questions we can answer? I don't see the need for a call from the information provided.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    Genius 4715 points

    Dear Wyatt

     This is customer's reply for your reference.

    1. about "80 seconds". Thank you so much for your reply, even I am not 100% understand. But let's check and confirm by our-self first. I will feedback to you if we have any other issues.
    2. about "continuous discharge during learning". Thanks!
    3. about "learning cycle". Yes, we know that we need one golden sample one for the learning cycle during design phase. But when we mass produce the packs, we have to do the fully charge and fully discharge at least one cycle, so that the Error can down to 2% from 100%. Right? But my original question is that, after shipping, and during user using, the max error might jump from 2% up to 8% even 100%? 

    Thank you for your support.  Thanks!

    -- Sam CHEN, CTO from Ctechi

  • 0 in reply to Gabriel Wang1
    TI__Guru** 105760 points

    Hello Gabriel,

    The MaxError should not jump up from 2% to 8%. When you upload the packs with the golden image they should have maxerror of 1% and it will take many cycles to increase it. Do they have a .gg file from before the pack is sent out and after the maxerror increases? What's the cycle count for the used pack with higher max error?

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    Prodigy 200 points

    Hi Wyatt,

    This is the stupid guy Sam who popped questions to Gabriel. Ha...Thank you for support! Sincerely appreciate! I just registered my own account so that I can contact with you directly to speed up the process.

    Yes, we get the *.gg file after the golden learning cycle. With the *.gg file, we can do the mass production.But after programming the *.gg during mass production, the MaxError is 100% as default instead of the 1% as you writing. We have to do the fully charge and fully discharge to decrease the MaxError down to 2% for each pack. This is what we have doing for many years. Do we doing correct?

     After shipping out, end user might do many cycle like 1000. I am worrying about that the MaxError might increase up to higher. That is why I would like to check with.

    Thanks Wyatt for your support

    Sam CHEN -- CTO from Ctechi

  • 0 in reply to CHEN Sam
    TI__Guru** 105760 points

    Hello Same,

    Haha they're all very good questions.

    Are you using the .gg file for production or the srec? The .gg file won't contain the chemID you found during the GPCCHEM process. If you use the .gg file you would also need to program the chem ID separately. I would recommend programming with the srec for production.

    The MaxError should be 1% after the learning cycle completes and you upload the .gg file to the gauge, MaxError of 100% usually indicates the gauge has not learned the Qmax or Ra table.

    Sincerely,

    Wyatt Keller 

  • 0 in reply to Wyatt Keller
    Prodigy 200 points

    Hi Wyatt,

    We are using the *.gg for the LifePO4 because of the CEDV, and *srec for NCM. We are using 4050 & 78350 for LifePO4, and 40Z50 for NCM. One project with 78350 for 8 cells LifePO4, we have issues as Gabriel popped at the beginning.

    Thanks

  • 0 in reply to CHEN Sam
    TI__Guru** 105760 points

    Hello Sam,

    That should be fine, .gg should have all the CEDV parameters you need. I guess I'm not sure which parts are the ones in question regarding the MaxError increasing, The title has multiple different parts and the BQ40Z80 is the main part linked to the thread. Is it actually the bq4050 that is showing this MaxError behavior?

    The BQ4050 MaxError will behave differently than the impedance track gauges, it will basically increment every 4 cycles depending on if the FCC is within or outside an expected range when it does get updated. If the FCC were to change significantly, then the max error would increase to 100%, indicating a learning cycle is needed. If it's increasing normally it will be capped at 8% maxerror.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    Prodigy 200 points

    Hi Wyatt,

    As for the MaxError question, I am talking about 4050, 78350，40Z50，and 40Z80. I want to understand the MaxError behaviors, so that we can design better projects. I think I have all three answers now from your information now. Thanks for your support.

    BTW. If you have chance to go to Shenzhen, please contact with me via Gabriel (because I can not list my contact information here). Sincerely hope you can visit our factory.

    Thanks.

    Sam CHEN

    CTO from Ctechi

  • 0 in reply to CHEN Sam
    TI__Guru** 105760 points

    Hello Sam,

    Ahh I see, yes the impedance track increments Max Error a little differently. I'm glad it's all understood now!

    I'm sure we will visit, not sure when travel will open up again for TI.

    Sincerely,

    Wyatt Keller