LiBSU - Redesigning the "Battery Support Unit" to Support Lithium

Discussion in 'Gen 3 Prius Technical Discussion' started by mudder, Jun 7, 2024.

Tags:
Page 8 of 8
  1. mudder

    mudder Active Member

    Joined:
    Mar 13, 2024
    298
    404
    0
    Location:
    Chattanooga, TN
    Vehicle:
    Other Electric Vehicle
    Model:
    N/A
    Thanks for the feedback. I've clarified my previous post with the following statement:
    Everything I write in this post is in regard to the Gen3 architecture (which my LiBSU product is designed for).

    I had naively assumed this was obvious, given that we're discussing BTH serial bus... but I agree that wasn't clear to the lay reader.
     
    #141 mudder, Jul 4, 2026 at 5:41 PM
  2. howardc64

    howardc64 Member

    Joined:
    Nov 11, 2011
    135
    54
    0
    Location:
    Pacific Northwest
    Vehicle:
    2011 Prius
    Model:
    Two
    @mudder Great Li battery engineering. I have background in Tesla Model S (MS) battery pack HW design issues/repairs and basic understand of its balancing solution. Also have done manufacturing in China for simple stainless steel parts to resolve MS drive unit's coolant cooled rotor leak fiasco (Audi EV also did coolant cooled EV motor rotor and suffered similar leak fiasco)

    Few questions / thoughts below. Curious on your thoughts and design choices on these.

    Souring Questions

    Cell Source

    Probably one of the most critical factor. Lithium cell manufacturing concept isn't difficult. But high reliability at cost effective scale is the art. Panasonic has done well with Prius NiMH and also 18650 for Tesla. GM's experience seems to show LG Chem has had quality problems early on. These are ultra high volume high quality low cost solutions which enables highest feedback loop on production quality improvement. But these are only accessible to ultra high volume users like car manufacturers.

    Capital

    Will need $$$ to pay for for volume production runs to even support a community like gen3 Prius cars in priuschat.

    Design Question

    Cell Safety

    I see several interesting and natural solution on cell safety

    Boeing 787 use low volume Industrial Japanese Li battery supplier AND I heard they run it < 50% SOC to stay away from the higher SOC and volatility. For airplane application, need the safest of safe.

    Tesla MS will reduce charge rates as cells resistance increase with age. More than 2-3x in my early model 2013 MS.

    Cell Longevity

    Tesla cuts down charge AND discharge rates at cold temperature. On cold winter days, my regen is greatly limited going down a big hill. Obviously to protect the battery.

    Balancing

    MS's balancing logic is often somewhat mysterious (keep changing with software updates and no documented logic) But in general, BMS wants to see behavior of all cells under all kinds of conditions (probably a full sweep of charge and discharge rates, and temperature at all SOCs) to decide where and how much to balance. They use bleed resisters per "cell" (I'll loosely call ganged parallel group of cells a "cell" to avoid manufacturer specific terminology)

    ===

    A side note is MS's huge floor based cell pack suffers from moisture ingress (sits outside of the cabin while Prius battery sits in the driest place) MS's huge pack size/volume require air equalization and didn't include air drying mechanisms in the air path, airborne moisture goes into the pack, condensate and corrode the electronics (conformal coating is far from perfect). This is actually the #1 failure on MS battery packs. Its not the cell failure.

    Anyhow Prius pack avoid this issue entirely by sitting inside of the moisture controlled cabin.

    ===

    Again great work so far. Add me to any ramp up testing program. Also LMK if can help in anyway (sourcing research, production funding ideas, any design analysis / comparison with Tesla solutions etc)
     
    #142 howardc64, Jul 6, 2026 at 3:21 PM
    Last edited: Jul 6, 2026 at 3:29 PM
  3. mudder

    mudder Active Member

    Joined:
    Mar 13, 2024
    298
    404
    0
    Location:
    Chattanooga, TN
    Vehicle:
    Other Electric Vehicle
    Model:
    N/A
    Even the big players have had issues, but yes I agree that cell manufacturing quality is very important.
    My cell supplier isn't a fly-by-night operation. They're a public corporation that's been around for thirty years. Their cells passed my brutal testing with flying colors.

    I am very fortunate that money isn't a limiting factor.

    That's a DAL-B level application, which requires more 'safety' than hybrid traction batteries (ASIL-C, which is roughly equivalent to DAL-C).
    There's a huge chasm between these two applications.

    I'm using a relatively low energy density lithium cell, which will only cycle SoC from ~20% to ~85% SoC, which is well within its safe operating area.

    Modern lithium equipment in airplanes is almost always DAL-B, which is one level down from DAL-A.

    The cell I'm using is used well below its maximum charge/discharge current.
    The firmware will limit charge/discharge current as the cell ages.

    Yes, limiting/prohibiting regen below freezing is incredibly important for lithium cells.
    The firmware regulates/disables regen when the cells are below freezing.

    My product includes a built-in resistive pack heater, which contacts both sides of every cell. When active, it generates ~4 watts per cell, which quickly brings the pack above freezing.

    My design simultaneously has dedicated cooling channels between every single cell, so OEM fan cooling isn't restricted. In fact, my design has substantially more cross sectional airflow than OEM, which is certainly overkill because the pack self heats substantially less than OEM due to much lower cell resistance.

    Ideally we'd heat and cool with a heat pump and coolant loop, but adding that hardware would make the product unaffordable.

    I use passive cell balancing, too.

    I've personally seen inside several failed Tesla packs, due to moisture ingress. It's not pretty.

    They've improved their design over the years. For example:
    -they now flood the pack with low moisture air while sealing the pack during manufacturing.
    -they've improved their material stackup to reduce galvanic corrosion.

    ...

    Unfortunately, the Prius cells are directly exposed to cabin air. In fact, the cooling fan intentionally blows air directly by the cells. This means that when the cells are below the dew point, moisture from the cabin air will condense on the cell walls. My product reuses the OEM cooling fan, hence my design has to mitigate corrosion caused by condensate formation.

    I ran into condensate issues early in my similar G1 Honda Insight product. The cooling architecture is conceptually identical to the Prius. Over time, some of my early customers in the pacific northwest experienced galvanic corrosion due to condensed water dripping off the cells and onto my circuit board. This caused traces on the PCB to corrode until they opened, at which point my BMS detected the failure and disabled the system (by asserting a CEL).

    I fixed this issue by applying conformal coating to both sides of the PCB; I had previously only coated the component side.

    I've incorporated this knowledge into my Prius design. In addition, when the fan is running I will activate the built-in pack heater as needed to reduce condensation. This will help keep the cell temperature above the dew point. Each cell is also wrapped in an insulator that prevents condensate in the air channels from contacting the outer metal cell wall.

    The cabin is far from moisture controlled, particularly in cool, moist climates. So far the PNW seems the worst based on my ~QTY400 unit Honda Insight test fleet.

    You're on the list!
    Is the 'c64' in your username related to the computer?
     
    #143 mudder, Jul 6, 2026 at 4:37 PM
    Last edited: Jul 6, 2026 at 4:43 PM
    howardc64 likes this.
  4. howardc64

    howardc64 Member

    Joined:
    Nov 11, 2011
    135
    54
    0
    Location:
    Pacific Northwest
    Vehicle:
    2011 Prius
    Model:
    Two
    Very cool. Thanks for the insightful reply :) Condensation / moisture issues has attacked MS at multiple places... even the cabin heater. I guess water and electricity don't mix is true haha... And ICE's waste thermal heat was quite a nice heater/dryer haha. I'll outline more below...

    Wow, thats a nice design addition from Prius original no thermal control. But I guess Prius used a more thermal tolerance NiMH + just depended on human driver wanting a comfortable cabin and inherit its temp. And indeed, coolant thermal loop would be cost prohibitive.

    I have recently discovered 2 more design problems on MS's PTC heaters.

    Cabin PTC heater

    This is a common failure on MS. The electronics for this heater shows signs of condensation. I guess living near warm and super cold HVAC duct in a closed off low air flow chamber must be the cause. Condensation on the control signals of for high current switches probably killed these heaters. My analysis link if curious.

    Interesting! During repair of MS battery pack boards with condensation corrosion. I tested the conformal coating and found the sharp edge / corners (hardest to cover) are all exposed. Conformal coating doesn't want to cover the sharp corner point and leave it exposed after drying. I even hand brushed multiple additional coats... still exposed. So if the shapes nearby is like a baffle and naturally condenses moisture in air, this starts the corrosion process. Not sure how to solve... Analysis link if curious

    Coolant PTC heater

    Since your design has no coolant thermal loop, this isn't an issue. This is another common failure on MS. PTC heating element is submerged in coolant for battery heater.

    I think the failure mode is if air enters the coolant system (say a leak somwhere) and the heater is located at a relatively high point where air will trap. Then PTC element will be partially submerged in coolant and partially exposed to air. Just a guess but probably the difference in medium on the PTC element caused stress when turned on.

    Interesting, I haven't heard of Prius having condensation problems around battery electronics but maybe others have. I'm in PNW (Seattle suburb) but car is garage parked so helps mitigate condensation issues.

    Toyota Rav4 EV actually has the most robust EV pack condensation solution. Basically a fan, dryer (desiccant), and a heater (dry out the moist desiccant) in the breather air path. European supplier MANN even provide a unit for this purpose although I'm not aware of any manufacture using it. Most EV packs from traditional manufacturers locate a breath port PTFE patch in a dry place. For example, facing the bottom of the car floor (Chevy Bolt) MS located its breathing port behind front tire spray and windshield water runoff... So basically design team inexperienced with water channeling around the car.

    Thanks, other have thought C64 computer also haha but its last name initial + birth year :)
     
    #144 howardc64, Jul 6, 2026 at 5:15 PM
  5. mudder

    mudder Active Member

    Joined:
    Mar 13, 2024
    298
    404
    0
    Location:
    Chattanooga, TN
    Vehicle:
    Other Electric Vehicle
    Model:
    N/A
    The answer is DFM ("design for manufacturing"):
    Don't put any copper traces near corners on the PCB. The PCB's fiberglass FR4 substrate won't corrode... so just pull the traces away from the edges.

    The OEM cells are extremely tolerant to condensation. They're aluminum and polypropylene. Since the aluminum isn't near any dissimilar metal, it essentially won't oxidize no matter how wet it gets. If enough condensate got into the pack to bridge the cell aluminum to the steel enclosure, you'd have issues there.

    I've certainly seen a few Prius packs that are quite rusted on the bottom. The Gen3 Prius I purchased for this project still had its original pack, and even though the rest of the car is mostly rust free, the bottom steel plate of the traction battery is quite rusty. I attribute this to condensate forming on the cooler cells while warm cabin air blows in, which pools on this steel and rusts. This car spent most of its life in the humid south.

    I haven't seen enough packs (yet) to know whether or not rusty bottom steel plates are common in the Prius traction battery. Hopefully someone who has seen hundreds of packs chimes in with their anecdotal evidence.
     
    #145 mudder, Jul 6, 2026 at 11:08 PM
    howardc64 likes this.
  6. howardc64

    howardc64 Member

    Joined:
    Nov 11, 2011
    135
    54
    0
    Location:
    Pacific Northwest
    Vehicle:
    2011 Prius
    Model:
    Two
    Sorry I wasn't very clear. The sharp corners I tested were top corners of larger and taller surface mounted capacitors : 1206. On Tesla MS, a pair of 1206s were located near condensation formation and corrode. Spent extra time studying how to cover them with conformal coating and discovered the top 4 corners remains exposed (continuity on VOM) after every layer of conformal coating I brushed on (and dried). The drying process seems to just pull the thin coating of conformal away from these shape corners.

    I found this on most of the "taller" components. On relatively flat components, thickness of conformal coating smoothly transition from top of components to the board and cover the corners.

    Anyhow, if you have taller components like 1206, might check their corners with VOM.
     
    #146 howardc64, Jul 6, 2026 at 11:30 PM
  7. mudder

    mudder Active Member

    Joined:
    Mar 13, 2024
    298
    404
    0
    Location:
    Chattanooga, TN
    Vehicle:
    Other Electric Vehicle
    Model:
    N/A
    What's the part number on your conformal coating?

    I've never seen that behavior with MG Chem 419D applied to a properly cleaned board.
    To clean, I run each PCB in an ultrasonic cleaner filled with Branson EC.
    I've never seen any gaps in the UV signature on taller components with this process.
     
    #147 mudder, Jul 7, 2026 at 2:58 AM
(You must log in or sign up to post here.)
Page 8 of 8