• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 62 subscribers
  • Views 599 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.

Concerning bq77pl900 5-10 cell Li-ion Battery protection & AFE...

DrSchoen
Prodigy 40 points
Other Parts Discussed in Thread: BQ77PL900

Hello,

I have an HR space application for which I attempted to post a question on the forum blog  (see e2e.ti.com/.../528567), but never got responses, so I am trying here...

Since making the aforementioned post, I have (almost) settled on the bq77pl900 as a solution.  However, I need some technical guidance to assure I'm not "barking up the wrong tree".

The application is a 10S 18650 LiPO4 pack, with triple redundancy; we are calling it a 10s3x(1P).  

Each pack will be a separate 32V 1400mAh source, and the host controller will be responsible for overview monitoring and system-level safety.  However, in the event of MPU failure, I want to be able to have the packs default to the "Stand Alone" mode.  

All three packs must operate independently, but the output will be "ored" together, and seen by the satellite as a single power source under the control of a single host controller MPU  (I have designed a serial multiplexing arrangement and software to allow the MPU to communicate with each pack's image individually).

I have three questions:

1) Is the concept of automatic fall-back to Stand Alone mode possible to implement "easily"?

2) Is there any particular issues that are immediately evident in the design concept of having three independent 10S packs "diode-ored" onto a common bus, particularly when the packs are operating in Stand Alone mode?

3) Will the bq77pl900 be capable of cell balancing the LiPO4 chemistry rapidly enough to address the divergence issues common in LiPO4 chemistry, or would external bypass FETs be recommended?

Thank you for your time and assistance.

Cordially,

Tom Schoenborne.

(201) 349-5225

  • 0
    TI__Guru 54715 points
    1. A transistion to stand alone may not be easy to implement since it requires clearing the STATE_CONTROL register HOST bit. Without transition the protections selected will still operate, but recovery will wait for the host. With LiFePO4 cells OV protection would be implemented by the host since the minimum OV threshold selectable is 4.15V.
    2. Diode ORing the packs will allow the highest voltage pack to provide the current for the system until its voltage pulls down to that of the other packs. If the packs are switched together with a ideal-diode OR circuit, high currents could flow between the packs until the voltages equalize. At the low end of operation, if the packs are allowed to run to the protection limit, one pack will turn off first due to tolerance and the load will shift to the other packs. If the load current is low the remaining packs should be fine, if current is high the remaining packs may overcurrent. It will depend on your system design.
    3. The maximum internal balance current of the bq77PL900 is low. Depending on your charge current, the battery capacity and how your system implements balancing you may want external balance FETs. With external balancing keeping the input filter resistors at a normal value will help avoid measurement errors. Depending on the FETs used it may be desirable to use FET gate protection diodes as shown in www.ti.com/lit/slua749 paragraph 4, figure 6. Host balancing would be needed for LiFePO4 cells since the voltage is so high.

    Do rememeber the bq77PL900 is standard device, not a high reliability part.
  • 0 in reply to WM5295
    Prodigy 40 points

    Thank you for your response.

    1) I understand the difficulty, and believe I have a method of producing the necessary transitions of the STATE_CONTROL [HOST] and other signals.  It would be triggered, either by forced sequence from ground control, or if a watchdog timer fired, indicating the MPU failed to restart the timer within the window.  This fall back would be another form of redundancy; implemented rarely; in the event of loss of host control/other critical issue.  Concerning the OV threshold, we've spoken to the cell manufacturer; they believe the cells will tolerate the 4.15V level benignly for short periods; although they anticipate the cells suffering some life degradation.  However, as this fallback mode would likely be collateral to much greater problems (e.g., a critical failure of the MPU), some reduced battery life will be the least of our issues.  While in HC mode (the, we hope, typical state of affairs), the host will track and control cell OVs. 

    2) We do intend to use ideal diode ORing, with augmenting control signals; when I used the term "common bus", I was being a bit simplistic.  We have a power supply module capable of treating each pack as an independent source; the module has its own MPU control image and redundancies; it should be capable of dealing with imbalances between the three sources.

    3) We have decided that FETs will definitely be necessary to implement the rapid convergence balancing algorithm we desire to implement (which can have currents approaching .5C).  The Zener protection diodes are a definite must, and the comment about input filter resistor selection is well taken.

    As for the reliability issue, we are looking for such an HR component that meets our needs, but haven't been able to locate one yet.  We are also interested in transitioning to a rad hard design spec in the near future.  Any direction you might be able to offer in those areas would be appreciated greatly.

    Thanks again,

    Tom S.