HW Integration Meetings

Coulomb counting

• Decided to go with on-vehicle SoC calc with max17261.pdf on current sense

  • CC by itself is not sufficient due to drift, algo needs to integrate vsense to anchor CC mesurements

  • Developing in-house SoC algo is difficult and unsure of what accuracy will be

• Eric’s notes: https://uwmidsun.atlassian.net/l/cp/oPkT0eia

Solar Irradiance

• Has theoretical power of one array using pysolar

• Next steps:

  • Model all panels + effects of aero curvature

  • Obtain power for a given route

  • Add weather effects

  • Account for system efficiency over varying loads

Coulomb counting

• Found fuel guage IC that we might be able to use for main pack, a lot of FW complexity

  • MAX17261METD+T Analog Devices Inc./Maxim Integrated | Integrated Circuits (ICs) | DigiKey

• Strategy algo progress basically 0

  • On-vehicle calc limits model complexity → lower accuracy

  • Unsure of resource usage on micro

• Decision to be made next Tuesday on which method to pursue, to consider

  • FW effort of writing I2C driver for IC vs building in-house CC algo

  • Accuracy of OTS vs in-house

• Forest to build 36x1 pack this weekend to support testing of CC + AFE

Telemetry

• RF implementation is complex and involves dealing with FW drivers, regulations, and interference

  • Was initially proposed as a simpler alternative to dealing with networking protocols, but that doesn’t seem like it’s the case

• From strat, we only need accurate SOC every 5 mins for their model, possible to do over radio

  • What accuracy is needed?

• Baseline is no live data, will be a task but not on the critical path

  • Critical assumption made is that we will be able to get accurate SOC calculation entirely on the vehicle, which is possible especially with OTS ICs

Battery box board mounting

• Started mounts, mostly done with carrier

• To focus on AFEs since they are most complete

  • Backup is to use MSXIV AFEs, design mounts for those

• To get current sense as close as possible to carrier

• Figure out HV harnessing and where to place current sense

  • Should be on low side (connecting to neg pack terminal)

  • It is allowed to be before the main contactors as per regs section 8.6.A.4. Relays only need to disconnect HV wires exiting the pack.

Vehicle range sensitivities

Super rough range sensitivities I came up with. These represent the expected range loss per day due to an increase in one of these factors:

+ 1kg:

• - 0.5km/day

+ 1W non-drive:

• - 0.7km/day

+ 0.001 Cd:

• - 2.3km/dayTLDR: We should prioritize aero when considering trade offs

Aero Status

Areo ideal: 0.09 Cd

Aero current: 0.20+

• Skin friction drag largest source of non-ideality, followed by panel gaps

  • Lots of sanding work required to get that down, panel gaps can be taped up

• Sim also missing wheel wells

  • Maybe possible to cover bottom opening with cover attached to upright

• Path down to 0.13

Battery box wiring / ports

Connector options:

1. 2 external HV connectors (motor/dcdc) + 1 LV 4 pin

2. 1 external HV connector (splitter outside BB) + 1 LV 4 pin

3. 1 external HV (dcdc inside battery box) + 1 LV 8 pin (for DCDC out, BMS pwr from PD, CAN)

   1. Preferred to reduce umber of HV connectors (and simpler HV wiring vs option 2)

Boards Inside BB:

• BMS Carrier

  • AFE 1

    • AFE 2

      • AFE 3

  • Current sense

AFE Temp Sense

• 16 thermistors - 2 for board and 12 for 3x modules, 2 NC

BMS FW

Cell balancing

• Goal is to make sure the pack charges fully

  • Cells charged in series charge at slightly different rates due to manufacturing tolerances

  • Series must stop charging when first cell reaches full, even if other cells are not yet full

  • Cell balancing will discharge cells that are “too full” so remainder cells can charge fully and bring the whole pack up to max capacity

BMS Fans

• How to achieve fault detection without fan controller?

  • Monitor current through shunt

  • Read fan tach from CB

Lights

• Probably don’t need multiple wrap around lights since they only need to be as bright as reference, and there will only be 1 reference light

• Ref DRL discontinued, but replacement has similar specs, to confirm with asc if they are okay.

PD validation update + lessons learned @jenna k

• Power rails were up at first, encounter smoke

• Observed current spike on display of pwr supply, board started smoking

  • Intermittent issue, happened 3rd time immediatly but not first 2

  • Zener diode was shorted between pwr/gnd

  • Something else was smoking but it still works

• 3v3 from CB was at 1.8V

  • 12V*0.042A = 0.5W

  • Replaced CB connector which fixed the issue

• 5v outputs 5v5, should be 5v2 based on FB

• In the future to set low current limit when powering boards

Update from ASC @Jonathan Xie

• Basically okay, small brained some things but we good?

• Stated only issue was can’t power steering and drive

LV Power Distribution

• Quiescent draw of AFEs from main pack

Current sense architecture @Soumav Maiti @Mitchell Ostler

• Pass relay power through for cutoff

MCI/transition to drive state @Forest Zhou @Mitchell Ostler

Press button to drive, what happens?

• Enable HSD to MCI

• MCI triggers precharge start

• Precharge closes motor contactor once complete

• MCI reads precharge complete

• Drive mode indicator comes on in CC

Pack Design

• Went with module design since we did it for msxiv

• Device on number of modules first → drives number of series/parallel cells

• 2x4+1 config

Solar Testing

• Solar panel connected to MPPT

  • 7x4 stack, 1 mppt, connected to dc load

  • Panel lying flat with protective film

  • dc load set to 25V, 60W as measured at load with no clouds

  • Dropped to 45W with clouds

• Compared with protective film vs without 100W vs 125W without

  • 56 cells (200W max → 63%)

  • Voc -> 19v

  • Isc → recorded

  • Both MPPTs connected series

• Alastair Correya (Unlicensed) to update confluence tmrw

• Tested with batteries as well, similar power to eload

• Spoke with Elmar, set output pot so that they are proportional to stack sizes

Battery Box Board Mounting

Participants:

Lights

• Strongly consider use reference lights if they can solve for brightless while being within aerobody

• Decision on custom vs ref lights pending response from regs

BMS

Questions

• do we pull over

  • What do we do to swap

• Are all the things we have BPS since they control BPS

• Justify why all the components are required for safe startup

AUX

• To ask if we can keep driving if aux dies

  • Assuming relay is redundantly powered, are any of the other loads drive critical

  • We think not, we can keep driving if aux dies pending regs

• UVcutoff isn’t pwoered on startup, add output to a load switch on PD

  • Otherwise you cycle if aux batt is dead

• CC should be on during startup since its the central fw element

  • Also in case of BMS fault during statup we want to have indicator on CC

• Ask regis if we can power PS CC PD UVC + other things (what is included in BMS)

• Aux batt needs to go in a battery enclosures

  • Ask about process of swapping aux

Power Up Logic

• Fill in later

Active SCP

• In firmware from current sense to carrier ~10-15ms

Participants: Former user (Deleted) Mohamed Goha Jonathan Xie Owen Li Shem Kim

Dashboard

• Steering stalk is super huge and blocks the whole thing

  • Alternate idea to have console on one side and display on the other, with ventilation behind the wheel

    

• Ben came up with a new design and it seems to fit tho? Mohamed Goha can you check

  • Need to decide estop location

• Can we just use an iPad?

  • Yes, but we need to make an app and have the rear camera feed streamed to it

  • More effort than its worth

Power Distribution Location

• Two primary options: under the seat and in the front

  • Under the seat is the most “central”, optimal for harnessing but hardest to access. It is the furthest point from the outside and you need to climb in for access

  • Want to prioritize serviceability, will have PD on top of chassis in the front

Startup and Power Switches

• External estop switch will be in the rear b the lights/cam, will be normally closed

• Internal estop will be flight switch like device accessible to the driver, can go to the left of the main display

• Proposed startup sequence, changes required

• 

  PHASE 1 (initial start-up, before HV pack is engaged):

  1. In order for Main Power Switch relay to be engaged, it needs to receive power and signal from the AUX batt

  2. aux batt needs to first flow through the N.C. rear switch in the back of the car (external E-stop), as well as the front switch (N.O), which is located next to the driver. Closing this switch is 'phase 1' of the power-up sequence

  3. AUX power flows through UV cutoff, which currently has active protection to cut off the power if its voltage is too low

  4. AUX flows into Power Select, which through pure hardware, selects the power source used to power the LV system. The priority is JUMP > DCDC > AUX

  5. during phase 1 (initial start-up), only AUX is available, so Power Distribution will output power from AUX (purpled line on diagram)

  6. BPS will be powered off this purple line, and will check the status of the main pack. If the main pack is okay, then BPS will close the 'BPS Switch' (N.O. relay, powered by purple line)

  7. If all the above conditions are met, then purple line will power and enable the Main Power Switch relay, activating the HV system

  PHASE 2 (HV pack is engaged, but car is not fully on yet)

  1. Main pack power flows into DCDC, which creates DCDC LV power

  2. Power Select now identifies DCDC is available, and allows PD to distribute DCDC to power the rest of the LV system

  3. Note that the purple line now becomes DCDC, so BPS and BPS Switch are now powered by DCDC. BPS MUST be powered off DCDC after start-up, but BPS Switch doesn't have to, so maybe we can leave it powered by AUX after start-up?)

  4. Fans, horn, telemetry, will always be powered by AUX (unless aux dies, in which... maybe telemetry & horn need to be taken over by DCDC. See Phase 3 #3)

  5. Centre Console is now powered by DCDC, meaning that its buttons and indicators become active.

  6. Pressing the 'POWER' button on

  PHASE 3 (car is fully on)

  1. Motor controllers, MCI, and all other HV and LV systems are enabled.

  2. In the event of a BPS emergency:

  3. BPS will open the BPS Switch, thus cutting the power and signal to the Main Power Switch, thus disabling the HV and DCDC systems

  4. Power Select will now select AUX as the LV power source such that the driver can safely steer to a stop, and the driver + onlookers are notified of the emergency (this require BPS lights, CC, steering, pedal, etc to all still be active and powered off AUX)

  5. Once the car is in a safe place, the driver can disconnect the Front Switch, or someone from outside the car can hit the Back Switch, to fully disable the car

  6. In the event of AUX failing: (which is likely to never happen, because aux batt should last an entire racing segment)

  7. BPS Switch (if we choose to power it off AUX in Phase 2/3) will switch to DCDC as its power source

  8. Horn may have to be powered off DCDC (is it critical to still have horn?)

  9. Telemetry may have to be powered off DCDC (or can the driver just walkie-talkie to the trailing vehicle?)

  10. In the event of both failing:

  11. Car completely dies

  12. To turn off the car:

  13. Press the POWER Button on CC (with brakes pressed?)

  14. Flick the Front Switch to OFF

  Micah Black 
  This looks pretty good. A few thoughts:

  • 5.a. Yes, brakes should need to be pressed, as in all vehicles.

  • You should expand on point 4. Think through what happens if BPS then aux fails, or the other way around. How will you tell what happened, where are the logs, are the failure states latched, etc.

  • 3.b. if aux fails, and you're still driving and want to pass someone at FSGP, you will need a horn (unless regs have changed significantly).

  • 3. Make sure you have an indicator of this somewhere in the car.

  • Think through what happens when someone pushes the e-stop. What happens when it gets re-enabled? What if this re-enabling happens really quickly (i.e. someone bumped the e-stop but it didn't latch - will the car still have power but think it shut down?) Edge cases like this are important to consider.

  • Precharge should be part of your diagram. What happens if something fails during precharge? e.g. car gets half-precharged, then aux does not have enough power left to close the main contactors?

  Forest Zhou  
  @Jonathan Xie The rear estop switch should not be controlling aux power to power select. That needs to directly de-energize the main relays. Otherwise, if you push the estop the car will just go into 3a and keep on running. Also the diagram says msxiv lol (edited)

Aux Batt

• SOC indicator is not needed tailing car can communicate to driver pack voltage

  • Can also make BPS indicator RGB, for both batteries

  • Or switch main pack voltage display to aux voltage momentarily

• Maintain active protection, since its pretty simple

Main BMS

• Can we use an integrated BMS IC that does coulomb counting Former user (Deleted)

• Can we do SOC on car so we can display percentage Mitchell.Ostler (Deactivated)

Single Motor

• Too late now

Comms

• Different from telemetry, telemetry is only for sending vehicle diagnostic data and not for communication with the driver

  • Telemetry is unidirectional, will simply spill can bus data to trailing car and nothing else

• Communication with driver is to be established via handheld radio?

Power

• NEED TO REVISIT FOR SUMMER VALUES Calculation of Solar Insolation | PVEducation

• Power input estimate

  • Theoretical 3.5W/cell * 256 cells = 896W

  • With 4 peak sun hours per day, 896W * 5h = 4.5kWh/day

  • 5kWh pack over 10 days = 0.5kWh/day

    • Battery average current 0.5kWh / 7h / 130V = 0.55A or 0.07A/cell

    • Battery max current 10kW / 110V = 90A or 11.25A/cell

  • Total = 5kWh/day

• Power budget estimate

  • Solar inefficiency 4.5kWh * 0.95 = 4.275kWh

  • LV power = 30W * 8h = 0.24kWh

  • Motor controller power 4kWh / 7h = 570W

  • Driver power 570W * 0.95 = 540W

• Speed Estimate

  • Theoretical Aero Cd 0.09, realistic 0.15

  • Frontal area 1.36m2

  • Crr 0.0109

  • Mass 300kg Mohamed Goha is this with the driver?

  • Cruise speed at 540W = 11.2m/s or 40kph

Owen’s Numbers:
HV Battery Pack Specs:

• 36S8P 

• 151.2V fully charged (4.2V per cell)

• 130.68V nominal

• 58.2 Max discharge continuous current

• 176A pulse current draw (10sec)

• 5241.6Wh

Motor Controller Specs: 

• 160V max continuous bus voltage

• 122A max continuous bus current

Weight Limit: 

• 20 kg lithium ion cells

Solar input: 

• 800W max (guesstimate 300W? 400W?)

Race length: 

• ~8 days (72h of racing)

• 13h solar input (12.10.A 9 hour race day, 12.18.B.1 impound time)

• 13*6 + 11 + 11 = 100h (solar input, not including first race day morning and last race day evening)

Average speed: 

• 56 - 64 kph (for competitive vehicle FSGP)

• 48 - 56 kph (for competitive vehicle ASC)

What is our energy budget?

• 300*100 = 30,000 Wh (solar input)

• 5241.6 Wh (full capacity battery)

• (35,241.6 Wh)/ (72h) = 489.47 W

• 400*100 = 40,000 Wh (better solar input)

• 5241.6 Wh (full capacity battery)

• (45,241.6 Wh)/ (48h) = 942.53 W (48h because based on route lengths between checkpoints we don’t need to be driving 9 hours each day)

Calculate power required for cruising speed based on Cd, Crr, mass, etc:

• Avg Cruising Speed (489.47 W) =  9.5 m/s = 34.2 km/h

• Avg Cruising Speed (942.53 W) =  14.5 m/s = 52.2 km/h

Attendees: Mitchell.Ostler (Deactivated) Jonathan Xie

Objectives: UV cutoff merging with PD, power up sequence

UV Cutoff:

• Good to merge with PD

• maybe don’t merge with PS, cuz 1 MCU for all 3 might be too congested for firmware. TBD

• delete fuel gauge if battery/owen is ok w it

• fuel gauges

• conduct purely hardware validation on existing UV cutoff board soon (before new one is sent out)

• conduct firmware validation on existing UV cutoff (can read BATTERY_EN and send load switch enable signals). To be done after hardware validation

Power-Up Sequence

Critical Questions (in general):

• does UV cutoff power itself on start?

• does BPS need to be on before HV is turned on? Does it need to on before CC/PD are enabled? Or can it be on at the same time as LV system (CC/PD/etc)

• how is power selected when main switch is pressed (HV relay closed)

• what systems are powered by aux (on startup, and when do they ‘get off' aux)

• how does BPS cut off the battery (BMS relay?? what does it do and how, at each state)

• what do each of the driver controls (physical button vs digital button) do?

Next steps:

• Email ASC to clarify regs (when can BMS be powered by aux?)

• Reach out to past elec leads (Micah, Aashmika, Nita) to find out what they did

• Meet with Forest, Mitch, Owen, Battery Box, etc to discuss this (Wed July 12th)

Participants: Former user (Deleted) Mohamed Goha Jonathan Xie Yanshen Zhou Jasmine Thind Jenna Kong Owen Li Kostubh Agarwal Nicole Choe Shem Kim

Objective: sync discussion on: lights, power startup sequence, brakes, steering

Lights:

• Current enclosure is designed based on MSXVI

  

• LED board is in the YZ plane (if x the front-back axis of the car)

• Might have optics to focus the light more

• ENture surface is thermoformed over so light gaps can be as large as needed

• Todo get reference lights if we don’t already have

• How to get lighting requirements for side repeater? 5 degrees from body side

  • try not to have light stick out

  • What direction is reference light pointing? How is the reference light set up so we can replicate? Jasmine to email

• Taillight must be visible 10 degrees from perpendicular, so might have to drap around the side of the car

  • Requires 2 boards if we need that

• Confirmed will combine rear brake and turn

• To figure out mounting for center brake light and BPS strobe

• No combined turn/brake

  • One LED driver for both side DRLs

  • One driver per turn signal side

  • One driver for all 3 brake

  • One driver for BPS

  • Total 5

• Combined turn/brake

  • One LED driver for both side DRLs

  • One driver per turn signal side

  • one driver per brake light

  • one for BPS

  • Total 7

• Mostly todo: clarify reference

Whats the turn on process for the car?

• Person closes main control switch

  • Aux batt is connected to main relays (2 of them)

  • Aux batt directly connects to HV contactor switch, which you close

  • How do we power BMS from aux battery and then switch the power source?

  • What are we checking? Voltage and temperature and current?

  • When we close the contactors, what gets connected?

    • DCDC

  • Power BMS first to run self check

• How does BMS disconnect the pack if its connected to the

• Can we email the people to ask if BMS can be powered for a few seconds after

  • Also ask if power select/power distribution can be powered with BMS at the start

  • Also ask if it can power UV cutoff

• Talk to Mitch

• UV cutoff is active protection as defined by 8.3.A.5, which is no protection when there is no power. Do we need to add passive protection?

• Should we use primary cell for aux?

  • Probably not, since we can recharge aux batt any time so energy density is not that big of a concern

• Owen to get back to jonathan about cutoff voltage

• Aux battery will last at least one day, @Forest to see if thats is still true

  • 5Ah @ 12V, 60Wh, total driving time is 9 hours aux battery power consumption has to be less than 6W

  • Loads: 3.5W contactors, 1W for driver fan, Telemetry 2.5W, horn 0W which is 7W

  • To check regs if we can swap aux batteries whenever

• There is CAD for the battery box, Mo is working on main assembly

• Do we have thermal load for battery box

  • Allows us to reduce the LV power, BMS fans is the single largest LV load

  • Fans are part of the air duct that goes to the wheel

• To send battery box with battery if that wasnt sent already

• have searate battery box elec fw sync to talk about BMS architecture and general elec stuff in box

• mo working on assembled CAD model to come next week (to get git to work properly as well)

Solar Panel covering

• Nothing gets put on them

• The roof will have indents for where the panels will sit, everything will the flush

  • Some mech stuff mech people have figured it out

• Mo has solar panels in CAD

  • To get cutout locations

  • To make the specific 3d layout

How does regen braking work?

• one pedal driving?

• Jonathan to figure out how to deletect brake press and what it should connect to

• we need button to turn poff regen for brake testing

• No throttle is full braking

• Full pedal is full power

• to adjust gradient in testing

• Battery max charge current is 0.7c, 27A max charge current

  • 2700W

  • Set regen level to be lower based on batter percentage

  • Set cutoff for high SOC

  • regen will have to be limited

• Test regen efficiency on dyno when we get that working

• Jonathan to figure out how much braking is 2700W

• Wheel weels have not been designed

  • Don’t know how the openings will work

• Do we want a button to control fan speed? Probably

  • To check if there are extra circuits on steering stalk

• Forest to order some comically long USB cables

  • to see if long able off digikey

  • Cam to dash is 4-5m depends on how much we route it around

Participants: Former user (Deleted) Mohamed Goha Jonathan Xie Yanshen Zhou

Objective: lights, power startup sequence, brakes, steering

Auxiliary Battery:

• Currently planning on using 10S NiMH batteries (10 in series, electrically)

  • NiMH Cell: 1.0 V discharged, 1.2 V nominal, 1.45 V max

  • 46 x 22 x 22 mm per cell, so a 2x5 mechanical arrangement would be 110 x 92 x 22 mm, or 110 x 46 x 44 mm total size

• Alternatively, regs allow primary cells (non-rechargeable batteries) to be used, so could potentially look into that as an option

Lights (on canopy):

• Need to have 1 blinking red light to indicate a BPS emergency (by regs)

• Need 1 white light as a headlights

• Can maybe put the two next to each other, but have to discuss with Jasmine Thind

Telemetry (antennas)

• Antennas can’t be placed under solar panels (RF blocking)

• will most likely be included

Rear Camera

• Rear-view camera must be placed in the tail of the car, and can’t be in the top/roof enclosure

  • needs too large of a thermoform + distortion in order to fit on roof

• But, the camera in the tail can be routed to the telemetry Rasp Pi (in the rear enclosure). to process the data and send to dashboard via HDMI

Fans:

• By regulation, the driver’s face needs to receive airflow which is controlled by the car

  • so we NEED at least 1 fan for the driver

  • considering 2 fans, one on each side of the driver

Mounting Location Discussions

Battery Box Enclosure:

• The battery pack will need to be impounded and lifted directly up and out of the vehicle

  • Any boards located above the main pack in the battery box enclosure will need to have their cables disconnected from the car, and impounded along with the pack

Rear Enclosure

• There will be a relatively large space in the ‘driver’ section of the car (behind front enclosure, in front of battery box)

  • This depends on where the ballasts (weights to compensate for a light driver) are chosen to be placed

• Options for where this enclosure will be are:

  • 1) behind the driver’s seat

  • 2) under the driver’s legs

Tail Enclosure

• There is a very small (1in tall) but wide section along the ‘tail’ of the car, to house LEDs and rear camera

• LEDs will be mounted mechanically in series, in a thin strip along the width of the car

  • One LED on each side, so they can be turn signals (if powered individually), or brake lights (if both on together)

• Waterproofing:

  • LEDs: naturally waterproof?

  • Rear-camera: will require an O-ring, gasket, etc

Other Discussions

Power Budget

• Need to begin measuring the power consumption of all the boards to make sure we’re not consuming too much power!

  • DCDC is rated at 240W max! (20A at ~12V)

Connectors

• Need to figure out what type of connectors we will use throughout the car.

• Consider:

  • Max current & voltage

  • Type of mechanical connection (latching/non-latching, ease of disconnecting, etc)

  • Space constraints

Tentative Board Mounting Locations

Action Items

• Former user (Deleted) to sync-up hardware with strategy
• Former user (Deleted) to sync-up hardware with firmware
• Schedule regular hardware-mechanical integration meetings (Former user (Deleted) Mohamed Goha)
• Confirm mounting locations
• Follow up with members on dashboard design
•