Bob Shoemaker Systems Engineering Manager Automotive Battery Management Overview Automotive Battery Design Challenges Measurement Accuracy High voltages from groups of 616 series cells per module up to 256pack ID: 667262
Download Presentation The PPT/PDF document "1 Special Considerations in Automotive B..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
1
Special Considerations in Automotive Battery Systems
Bob ShoemakerSystems Engineering ManagerAutomotive Battery ManagementSlide2
Overview - Automotive Battery Design Challenges
Measurement Accuracy“High” voltages from groups of 6-16 series cells per module, up to 256/packWide temperature range -40 to105⁰C
TI offers ±2mV standard accuracy in automotive BMS partsCommunicationsExtremely noisy environmentEMC: Bulk-Current Injection (BCI) susceptibility requirementsEMI: Radiated energy must stay out of common radio entertainment bandsHot Plug – random connection of cells to measurement circuitsWildly fluctuating voltages during connectionExposure to many hundreds of voltsInductive ringing during connection may produce >80V on a 30V IC
Safety – ISO 26262
2Slide3
Communications
Driving Through Noise…3Slide4
Communications
Most interference is inductively coupled from the bus bar which carries very high currents up to 100’s of amps – ~ forming a one turn transformer with the communications cableNoise from the inverter is typically a 1V/cell - 3µs pulse at a 30kHz rate, with rise/fall times in the 2-5V/ns range!The noise is wideband with lots of harmonic energyStray capacitance to (isolated) chassis is also a big contributor
4Slide5
BCI Testing
5
Standardized test (IEC, others) that mimics the noise environment
Injects noise current into communications cables with toroid
xfmr
1MHz – 1GHz injected, 50-200mA
Effect on communications measured – CRC errors, dropped packets, etc.
Toroid
to Inject Current into 1M
Comm
Line
Cu TableSlide6
BCI Testing
6
test IC
Effect of BCI Test on communications signals.Slide7
Communications “Cures”
Good PCB layout practices, ground planes are very importantShielding may not buy much in this environmentImproves VHF, but may worsen low end resultsFerrites, very small inductors, small capacitors may help, especially with VHF-UHF
Partial impedance matching can helpSmall capacitors (AC coupling) between ground planes will make a difference if wire runs are short - typical values 3.3nF – 10nFThe best solution against high common-mode noise is differential communications links
7Slide8
Hot Plug
Random Cell Connections8Slide9
Hot Plug – Random Cell Connection Order
Connector pins mate in random orderWild voltage swings across cell input pinsMay stress ICWire length + capacitance around IC causes large overshoots
9Slide10
Hot Plug Exposure Can Damage Devices
3 stacked bq76PL536 devices, 4.7µF cap and 36V zener across each from BAT to VSSWires are ~2m long, ~2.5µHDevice 1 is exposed to ~55V at connection
Zener dynamic impedance is too high to prevent failure10
This circuit causes dramatic voltage overshoot at connection time, a result of natural & parasitic inductance and board (device) capacitance.Slide11
Hot Plug Exposure
With 0.1uF cap, exposure is reduced to <40VIC must have good Power Supply Rejection Ratio (PSRR)11
3 bq76PL536 devices, 4.7µF cap and
no
36V zener across each from BAT to VSS
Wires are ~2m long, ~2.5µH
Device 3 is exposed to 49V at connectionSlide12
Minimize Hot Plug Problems
Keep wiring to cells as short as possible to minimize inductanceKeep capacitance near IC to a minimum – this includes caps for filters, LDO’s, communications, etc.The 36V zener does not help, and may cause harmInstead use 5.1 or 5.6 zener diodes across each cell connection
A series resistor may reduce overshoot, but must be kept to a minimumToo large and it causes a significant voltage dropCauses the VBRICK measurement error due to I·R dropMay cause ESD diode to turn on if BAT goes below TOP cell (i.e. VC6) due to effects of L-R-C
12Slide13
ISO 26262 Introduction
Stefano ZanellaSlide14
Who is working on ISO26262 ?
Technically:
ISO-TC22-SC3-WG13
TI’s Karl
Greb
,
a co-author of parts 5 and 10Slide15
IEC 61508 vs. ISO 26262
ISO 26262
(Road vehicles)
RTCA/DO
17B
(Aerospace)
IEC 60601
(Medical
)
EN
50128
(Railways)
IEC 62061
(Machinery)
IEC 60880
(Nuclear
power plants)
IEC 61511
(Process)
IEC 50156
(Furnaces)
IEC 61508
(The grand-daddy)
Slide16
SIL vs. ASIL
ASIL -
Automotive Safety Integrity LevelSlide17
ISO 26262 is a Process and a Mindset
System
Risks
Which ASIL level?
ISO26262 is all about managing risk by reducing the number of
undetected faults to an appropriate level for the applicationSlide18
Information Exchange
Collaboration
Specs
Specs
Hazard
& Risk
Analysis
Safety Manual
Safety Manual
Audits
FMEA/FTA
FMEA/FTA
Other documentation
Other documentation
OEM
Tier 1
Semiconductor Supplier
AuditsSlide19
Item Definition
Which one??
Cell, battery, and vehicle pictures from teslamotors.comSlide20
Systematic Faults
Design Errors and EMCReduced through compliant development processLots of checklistsRedundancy helpsDecompositionSlide21
Hazard &Risk Analysis (OEM -> Tier 1)
Example: safety item – battery management system
Functional safety goals:
Over-voltage (OV) detection: ASIL D
Over-temperature (OT) detection: ASIL CSlide22
Supplier Deliverables
FMEA bq76PL455 - 1388 lines
Fault Tree Analysisbq76PL455 – 100 pages
Safety Manual
bq76PL455 – 85 pagesSlide23
What’s in the Safety Manual
Practical guide to functional safety using a component/systemMetrics (example later)Safety mechanismsAssumptions of useAll you need to know to integrate a system with the higher level systemSlide24
Random Faults Classification
All FaultsSlide25
“
Failure Diagnosis and Prognosis for Automotive Systems”, by Tom Fuhrman, General Motors R&D, IFIP Workshop, June 25-27, 2010
SPFMLFM
SPFM
LFMSlide26
Probabilistic Metric for Random Hardware Faults (PHMF)
ASIL
Random hardware failure target values
B
< 10
-7
h
-1
(100 FIT)
C
< 10
-7
h
-1
(100 FIT)
D
< 10
-8
h
-1
(10 FIT)
Sub-system
FIT Budget
Host Controller
2.2 FIT
Supervisory board
1.3 FIT x 6
Total
10 FIT
Budgeting
0.5 FIT
0.3 FIT
0.1 FIT
0.2 FIT
0.15 FIT
0.05 FIT
This is an example using the PL455 Evaluation Board. The FIT rate budget may vary by application.
BudgetingSlide27
“
Failure Diagnosis and Prognosis for Automotive Systems”, by Tom Fuhrman, General Motors R&D, IFIP Workshop, June 25-27, 2010Slide28
Summary
Texas Instruments has launched the SafeTI™ Program, a unified corporate approach to safetyIncludes ISO26262, IEC61508, IEC60730 qualified productsAll new TI automotive battery management products developed according to ISO 26262
ISO 26262 includes prescriptions on the development process for both hardware and softwareISO 26262 is about mitigating risks by reducing the number of undetected faults to an appropriate level for the applicationISO 26262 defines metrics (SPFM, LFM, PHFM) that allow a quantitative assessment of the riskSlide29
29
Bob Shoemaker
Systems Engineering Manager
Automotive Battery Management
Battery Management Solutions
Texas Instruments, Inc.
bobs@ti.com
301-407-9599
Thank You!
Automotive Battery Management
Special thanks to: Karl
Greb
, Stefano Zanella, and others for their contributions of slides.