Hybrid battery ground isolation question (re P0AA6 DTC)

Discussion in 'Gen 2 Prius Care, Maintenance and Troubleshooting' started by simonfunk, Oct 17, 2021.

Tags:
  1. simonfunk

    simonfunk New Member

    Joined:
    Oct 16, 2021
    10
    3
    0
    Location:
    Vancouver, Washington, USA
    Vehicle:
    2005 Prius
    Model:
    Two
    Hi all,

    I mainly have a technical question about the hybrid battery ground isolation, but here's a paragraph of context first:

    I had a P0A80 ("Replace Hybrid Battery Pack") on my 2005 ~250k mile Prius with original battery so went ahead and had Greenbean batteries swap in a refurb. The first(!) time I tried to drive it it would not Ready, and my code reader says P0AA6 (which as many here may know, would have allowed the installer to test-drive it just fine Once). Greenbean now wants me to take it to a dealer to read the sub codes, even though the car hasn't worked since they left. (I can't find a mobile app that can do it. Even though mine does find the P0AA6 and can read freeze frames.)

    To see if I could establish for sure that it was the battery, I tried this (measuring the voltage from the service plug fuse to the battery chassis):


    But I get a steady 144 volts, which I *think* is about where the fuse splits the pack, so it seems like I am just measuring the correct HV against chassis ground.

    My question, then, is: When should the HV battery be grounded to chassis like that? My impression is: Never. But it occurs to me a leak elsewhere in the system could be shorting the grounds together, so this doesn't necessarily mean an issue with the battery? On the other hand, aren't there supposed to be main contactors inside the battery that isolate both sides of the battery entirely when it's not actively in use?

    In short: I see 144V from the service plug fuse to the battery chassis, steady. Does this *definitely* mean there's a problem inside the battery, or could it be a HV->ground leak somewhere else? (The car was of course off at this time, but the 12V may have been live.)
     
    #1 simonfunk, Oct 17, 2021
  2. SFO

    SFO Senior Member

    Joined:
    Feb 7, 2017
    5,302
    4,243
    0
    Location:
    Northern California
    Vehicle:
    2007 Prius
    Model:
    N/A
    Where are you in the usa or which metroplex is nearest, as the country if fairly large still. Maybe another member could assist or help you.

    Don't believe there is an App that can read the INF(s) / subcodes / detailed codes. You can use a mini-vci cable, and techstream for such.

    Sounds like greenbean has installed a failing battery, most likely knowingly. Next time test drive it twice to make sure it starts a second time.

    Have you tried disconnecting the 12v battery, which may allow the vehicle to drive until you power it off again? (rinse and repeat as needed)

    If the above works (resetting the 12v, then 'power cycle' twice), and the cars drives, then take it to greenbean for a warranty claim (INF 612).

    FYI : your posts are being moderated until you've posted 5 times.
     
    #2 SFO, Oct 17, 2021
  3. simonfunk

    simonfunk New Member

    Joined:
    Oct 16, 2021
    10
    3
    0
    Location:
    Vancouver, Washington, USA
    Vehicle:
    2005 Prius
    Model:
    Two
    I'm in Vancouver, Washington. The car drives once after a 12v reset but I'm not keen to take it far that way if avoidable. GreenBean won't diagnose--they require a dealer to do that and they just swap batteries when the dealer declares battery failure with screen cap of failure code to back it up... (They only do mobile installs.) It is my hope they might accept a video of my DVM showing 144 volts where it should show 0, but I want to make sure my logic is sound before pushing that idea... Hence my question here.
     
    #3 simonfunk, Oct 18, 2021
  4. SFO

    SFO Senior Member

    Joined:
    Feb 7, 2017
    5,302
    4,243
    0
    Location:
    Northern California
    Vehicle:
    2007 Prius
    Model:
    N/A
    See if @PriusCamper is driving in your direction, as he can scan with techstream for a screen shot that you can use.

    There are some handheld obd2 scanners that offer INFs or subcodes/detailed codes, an advanced search should find them. A standalone scanner may cost about the same as a visit to the dealership and add another tool to your arsenal. A mini-vci cable / techstream costs $20.

    Which DTCs will greenbean consider a "battery failure", and if DTC P0AA6 is one of them, then which INF needs to accompany said DTC?
     
    #4 SFO, Oct 18, 2021
  5. ChapmanF

    ChapmanF Senior Member

    Joined:
    Mar 30, 2008
    25,692
    16,781
    0
    Location:
    Indiana, USA
    Vehicle:
    2010 Prius
    Model:
    IV
    What's the impedance of your meter?

    Do you still see 144 V if you load down the reading some more, say with a 220 kΩ ¼W resistor from your probe to the chassis?
     
    #5 ChapmanF, Oct 18, 2021
  6. TMR-JWAP

    TMR-JWAP Senior Member

    Joined:
    Aug 12, 2016
    6,404
    6,073
    0
    Location:
    Columbia, SC
    Vehicle:
    2007 Prius
    Model:
    Touring
    I think, with the battery case mounted to the car chassis, a steady reading of 140+ volts is a bad thing...certainly means the 'leak' is from the HV battery, since the main relays are open....
     
    #6 TMR-JWAP, Oct 18, 2021
  7. ChapmanF

    ChapmanF Senior Member

    Joined:
    Mar 30, 2008
    25,692
    16,781
    0
    Location:
    Indiana, USA
    Vehicle:
    2010 Prius
    Model:
    IV
    Yeah, I just tried the same thing with my meter, which has an input impedance spec'd at 10 MΩ. I will see an initial voltage between the service plug terminals and the body, but the 10 MΩ load will bleed it down by around a volt per second.
     
    #7 ChapmanF, Oct 18, 2021
  8. simonfunk

    simonfunk New Member

    Joined:
    Oct 16, 2021
    10
    3
    0
    Location:
    Vancouver, Washington, USA
    Vehicle:
    2005 Prius
    Model:
    Two
    My meter is 10Mohm (coincidentally the same meter as in the video above).

    I just jumped my meter leads right at the meter with a 330Kohm resistor (what I had on hand..) so that the meter read such when dialed to Ohms, and ran the test again. This time I got 133 volts (not sure if that's from the higher draw of the resistor, or just because it's been draining for another day or two...).
     
    #8 simonfunk, Oct 18, 2021
  9. simonfunk

    simonfunk New Member

    Joined:
    Oct 16, 2021
    10
    3
    0
    Location:
    Vancouver, Washington, USA
    Vehicle:
    2005 Prius
    Model:
    Two
    Unfortunately don't have access to any windows machines (linux and mobile only) so I assume I can't run touchstream?

    And I'm not sure what they'll accept, but I assume a 612.

    After emailing them an update about my test, I got a call from their warranty dept (haven't connected yet -- phone tag) so it's possible they'll honor it based on this much. Fingers crossed.

    Thanks all for your help here! Good to confirm that it's the battery -- would hate to talk them into replacing it without the codes only to find out it was something else......
     
    #9 simonfunk, Oct 18, 2021
  10. SFO

    SFO Senior Member

    Joined:
    Feb 7, 2017
    5,302
    4,243
    0
    Location:
    Northern California
    Vehicle:
    2007 Prius
    Model:
    N/A
    You would need a mini-vci cable (or better) to run Techstream. You can use virtualbox.org/ to run winblows.

    Do you need a preinstalled winblows image that already has techstream installed, or will you be rolling your own?
     
    #10 SFO, Oct 18, 2021
  11. ChapmanF

    ChapmanF Senior Member

    Joined:
    Mar 30, 2008
    25,692
    16,781
    0
    Location:
    Indiana, USA
    Vehicle:
    2010 Prius
    Model:
    IV
    If your 10 MΩ meter shows you a 144 volt reading, and your (10 MΩ ∥ 330 kΩ) ≈ 320 kΩ load just pulls it down to 133 volts, it's like you're looking at a 144.4 volt source with about 27 kΩ impedance. That's a serious leak; the isolation ought to be up in the megohms at least, and if I remember right, the threshold to trip a leak code is still in the several hundred kΩ range.

    If there is just one leak, it has to be pretty much right at the far end of the pack from the service plug (144 volts away) and the leakage path seems to have about 27 kΩ resistance.

    Of course it's possible for there to be more than one leak, and taken together they happen to be Thévenin-equivalent to a circuit with 144.4 volts through 27 kΩ.
     
    #11 ChapmanF, Oct 18, 2021
    simonfunk likes this.
  12. Travis Decker

    Travis Decker Active Member

    Joined:
    Jul 15, 2015
    139
    115
    0
    Location:
    Portland OR
    Vehicle:
    2012 Prius
    Model:
    II
    We could do a health check at my shop.

    You are welcome to email me [email protected]



    Deal with this all the time, you should be able to do the test Carolyn does in that video and that should be enough for green bean....
     
    #12 Travis Decker, Oct 18, 2021
    SFO likes this.
  13. SFO

    SFO Senior Member

    Joined:
    Feb 7, 2017
    5,302
    4,243
    0
    Location:
    Northern California
    Vehicle:
    2007 Prius
    Model:
    N/A
    #13 SFO, Oct 19, 2021
  14. simonfunk

    simonfunk New Member

    Joined:
    Oct 16, 2021
    10
    3
    0
    Location:
    Vancouver, Washington, USA
    Vehicle:
    2005 Prius
    Model:
    Two
    Honestly I haven't used windows since I switched to linux almost 30 years ago, so it would be a big learning curve for me. I would probably first try to write or improve some open source ODB software if the necessary details were documented anywhere, though I suspect they're not (publicly, at least).

    Anyway, thanks for the help! But it sounds like they've accepted my empirical voltage leak test as adequate so hopefully I will not need the sub codes after all.
     
    #14 simonfunk, Oct 19, 2021
    Goneballistic and SFO like this.
  15. simonfunk

    simonfunk New Member

    Joined:
    Oct 16, 2021
    10
    3
    0
    Location:
    Vancouver, Washington, USA
    Vehicle:
    2005 Prius
    Model:
    Two
    Final update: They replaced the battery under warranty and confirmed the one they originally installed had a leaky cell. So, yay.

    One final curiosity: The (brand new) 12v battery drained down to 11 volts in the first three days after the leaky hybrid was installed. I charged it back up (with an AGM-aware charger) and left it on a battery tender until they finally came and replaced the hybrid battery.

    I've been watching the 12v voltage since then, and this time it seems quite stable (12.85v currently).

    Is there any reason a leaky hybrid battery would sap the 12v battery?? (I think I saw one other person comment that this happened to them, but I have seen no explanation for it so I'm wondering if it's coincidence in both cases.)

    Mainly I want to know because if *not* then I have to assume I have some other issue intermittently draining my 12v and need to either track it down or keep it on a tender when parked. Whereas if so, then perhaps I can relax and trust my car again for a while. :)

    (I'll re-post this question to technical Q&A if nobody here knows offhand.)
     
    #15 simonfunk, Oct 27, 2021
    Goneballistic likes this.
  16. mr_guy_mann

    mr_guy_mann Senior Member

    Joined:
    Sep 3, 2020
    3,880
    1,867
    0
    Location:
    NJ-USA
    Vehicle:
    2006 Prius
    Model:
    Four
    Well, when you have an isolation fault in the HV battery, you have voltage potential to the vehicle chassis- which is ground for the 12V system. And the battery ecu is the only other point where the HV and 12V systems come together when the car is powered down. So I am guessing that the leak is somehow allowing a current path from the HV battery to ground, back through the 12V battery and then the battery ecu (reverse flow through some diode maybe?) finally returning to the HV battery through the voltage sense wires.

    You would have to use a low current amp clamp to measure where the juice is flowing when the fault is there.

    Posted via the PriusChat mobile app.
     
    #16 mr_guy_mann, Oct 27, 2021
    simonfunk likes this.
(You must log in or sign up to post here.)