Dale Reid / BPS, electrical, MOV charging
If you organize a 1-on-1 with this person, then fill in the timeslot and your name in the table below.
Timeslot | Organizer | Other notes |
---|---|---|
Mon. Feb. 22, 8:00pm-8:30pm | @Micah Black | Work time permitting @aashmika might join @Nita E to join |
|
|
|
He’s doing a part-time PHD: http://ampere.engr.uky.edu/people
Question 1: Is the situation in which a team tests the limits for there BPS system before hand, arrive at the site for scrutineering, then find themselves failing the BPS checks one that arises often? If so, can you think of any factors of the environment or situation that may cause that?
Teams will build custom circuitry - digital circuit
Microcontroller will shut down
flip a switch, power on a large load, get a large bus transient
Sucked all the energy out of the caps
holdup diodes to keep micro powered
DC-DC converters spread all over the car - thermal shutoff
When they were racing (2008) - Kentucky team - 2300mile, Austin to Calgary
120V-5V Vicor module, and a couple spares available
Had some low modules - used the vicors
Use the pack to charge the pack - even if it its manual
Notoriously had trouble with terminal breaking - pouch cells and road vibrations, etc.
Such low budget and resource-constrained
Question 2: When can we expect to have access to a charging meter before a race? (Receiving a meter before a race was briefly in exchange for a deposit was discussed in one of the meetings.)
Stay tuned…
Evan is building them. He has some already
Evan works really hard
Pilot and control signal must be wired separately
J1772 extension cords would have been a different solution
Let Dale know about the grounding of the metal chassis
IT industry (PCs, printers, etc.) use a real high standard of insulating - anywhere you can touch “doubly-insulated” or “reinforced isolation
Chargers are not built to that same standard - is NOT build for double insulation - chassis and AC power - chassis could become energized.
Build circuits to sell on the market - ALWAYS connect the chassis to GND.
Generator is not earthed
Instruction manuals likely state that
GFI protected outlets - How does it know?
Monitors current in both terminals
“Earth” is floating
Monitors current in both hot and neutral and a difference in them will trigger if
Prevent AC power from getting out into the chassis
Question 3: Your own projects on the solar team
“I was the electrical team”
”I was real stubborn”, just liked to build things - that’s just how its going to be. Not going to buy them, just going to build it.
MPPT trackers
“Textbook boost converter circuit”
Never successfully made it through a race. Somebody else’s subsystem would always apart.
2008 car
Just dreamed up a BMS and tested it, several iterations
Didn’t use any LTC chips, etc.
Used it 2008 until 2018
Some boards were still the original
A fan of capacitor circuits
Optical isolated FETs
Used a 1uF capacitor, 1 per module.
Across the capacitor was a set of FETs - 4 of them
Capacitor would charge to the battery voltage, then would get isolated from the pack
Opamp buffer circuit
Analog switches, piped into a microcontroller ADC
Berkeley team - the guy at Berkeley has done an incredible job - best power point tracking system
Has published some papers on it
Flying capacitor boost converter MPPT - google it.
Frequency multiplication effect - so inductor, etc. can be absolutely miniaturized - 1/4th of the size and 1/10th weight of anything you can buy
Crazy high boost ratio capability - all the way down to 3V - he might make it commercial someday
Might be this one: https://ecal.berkeley.edu/files/EECS498.pdf
Question 4: Do you often see teams failing insulation?
Not really - just run a
Question 5: Solar array to chassis?
“Our array is producing nothing” - “Oh look its shorted out”
Always check this
Question 6: Excel is awesome
VBA built in to it.
Factory had PLCs built in but needed a database system and number-crunching tabular data
Made ~50 different widgets
Controlled their whole assembly line through EXCEL and Access
Tables of data. Line would communicate with the system
Bins of parts that you needed to use would light up with LEDs.
Battery charge/discharge with python
Communicate with the SCPI built in. Program will pull data and throw it into Excel
Make it using a table - will grow dynamically - then chart data from a table
Dynamically charting the data.
Setup a timing loop - every second will go a pull the data - GPIB communication
If statements - for start/stop criteria for charging.
Industry uses battery cyclers - a lot of companies out there
multichannel devices
squidstat potentiostats - he met the guy that designed it https://www.admiralinstruments.com/product-page/squidstat-solo
Been using quite a bit of Keithley stuff
SMU, Loads, Power supplies
@Micah Black to send Dale info on the Python code and the Keithley Rpi app note
Toyota Prius had NiMH batteries - literally a process he developed. Cars that fail typically with one module bad
DAQ, Loads, Power Supplies, all connected. Excel VBA system will connect to it - come back next day and he’ll have all the data.
Macros - to ‘mine the data’
Charging time, etc. 3rd charge - what's the highest voltage, etc.
Manufacturer cells - some of the cells might be different
For a full cycle, he’s expecting
Brute Force hardcore do it with circuits kind of guy
When Tesla builds a module - (public info), say 30 paralleled cells and 1 of them decides to go bad. Fusible links connected - a cell shorts and the fuse blows.
Making sure everything is super perfectly matched - he’s not a big believer in this
The cells will all age differently.
Accommodate the difference with other ways
Background
Non-traditional student, worked for a number of years
Original team - he and a few people started a UK team 2002/2003
Lead Acid days - built a car
Lifelong learner - still taking classes now.
Worked in industrial maitenance
Lexmark, printer power supplies
Oil company now, working with NiMH - batteries would in Hybrids - coolant fluid, batter swapout
Always worked a full time job and went to school too.