This is a story about reverse-engineering the analog circuitry of the Tesla Model 3. Taking measurements on the "12V" line and seeing voltage dips as low as 2V. Watching this voltage-dip propagate over to the 1.65V reference voltage used by the TI and STM32 chips's ADC.
The ADC taking the (assumed 1.65V reference) at only 0.3V, and then how the voltage applied at the pedal mechanism being misinterpreted as 100% depressed (when the pedal remained completely flat), as well as hardware tests on "Pin44" (see page 19) showing how this changing voltage-reference could cause unintended acceleration.
---------
I know Musk / Tesla stuff is always a hot topic. But this is incredibly detailed electrical engineering, reverse-engineering, and other such factors at play here. I feel like Hacker News is still an ideal place to discuss, even if I know its going to be flamebait.
Please read over the .pdf and skim over the technical allegations on how the Tesla Model 3 reads/interprets voltages from the pedal, and how the ADC its associated voltage-reference could mess up here... at least for the EEs on this site who can follow the discussion.
---------
EDIT: The overall argument is as follows:
Root cause: high torque required by the power-steering uses more than 300 Amps of current. Tesla Model 3 12V battery can only supply 100 Amps, and the DC-DC converter to the main batteries only supplies 200 Amps. Low-speed turns requiring more torque could cause the 12V line to drop down as low as 2V.
The software calibrates off of a 1.65V reference point. If the software happens to calibrate during this voltage-drip, it will think that the 0.3V (reality) voltage represents 1.65V.
The accelerator pedal outputs 4V when fully pressed and 0.65V when the foot is off. 0.65V vs the 0.3V reference (thinking its 1.65V) will come back as "thought to be 3.5V" nominal, or nearly a fully pressed pedal (!!!). This will be locked in until the next calibration event. And that next calibration event will not happen for minutes.
10V swing for 0.5 seconds (500 milliseconds) across 300 Amps is... a 1500 Farad capacitor. (not uF. Big F, full size Farad. Lol)
Methinks that won't work, lol. The numbers discussed are pretty insane, to be expected of a car-sized load / amperage.
-----------
What they need is a well designed power-network (Pi-network? Capacitor-Inductor-Capacitor-inductor-Capacitor) to isolate the microcontrollers + analog pedal reader from the rest of the car.
PDF was spitballing ideas for a software-only detection of this brownout condition. But I don't like that, any uC / analog sensor circuit that has 10V swings on a 12V nominal circuit is just going to go haywire in unpredictable ways. I'm frankly surprised that the uC is still sending CANbus messages on the pedal state despite the 12V supply voltage swinging down to 2V. I guess TI and STM32 chips / ADCs have enough stray capacitance / local capacitors to survive even when the sensor network/analog circuitry is browning out in Teslas.
> Taking measurements on the "12V" line and seeing voltage dips as low as 2V. Watching this voltage-dip propagate over to the 1.65V reference voltage used by the TI and STM32 chips's ADC.
Did we read the same paper? He takes no measurements, everything is based on theoretical assumptions, schematics and the three youtube videos he mentions in the beginning of the paper.
From the paper:
> The enclosed paper proposes a simple test that can be done to prove or disprove the explanation provided. This test has not been done by the author because of the cost involved with acquiring the needed Tesla Model 3 inverter PWB.
> The details of Tesla's Model 3 inverter design as revealed by Irish engineer Damien Maguire have been presented. These details were used to construct *a hypothetical model* of all hardware and software operations performed on the two accelerator position sensor (APP) sensor signals inside the inverter as they pass from the APP sensor to the electric motor controller.
He doesn't have access to the parts in question nor the car in question.
Also of interest is the authors own comments on his previous papers about sudden unintended acceleration:
4 comments
[ 3.1 ms ] story [ 22.7 ms ] threadhttps://www.autoevolution.com/pdf/news_attachements/breaking...
This is a story about reverse-engineering the analog circuitry of the Tesla Model 3. Taking measurements on the "12V" line and seeing voltage dips as low as 2V. Watching this voltage-dip propagate over to the 1.65V reference voltage used by the TI and STM32 chips's ADC.
The ADC taking the (assumed 1.65V reference) at only 0.3V, and then how the voltage applied at the pedal mechanism being misinterpreted as 100% depressed (when the pedal remained completely flat), as well as hardware tests on "Pin44" (see page 19) showing how this changing voltage-reference could cause unintended acceleration.
---------
I know Musk / Tesla stuff is always a hot topic. But this is incredibly detailed electrical engineering, reverse-engineering, and other such factors at play here. I feel like Hacker News is still an ideal place to discuss, even if I know its going to be flamebait.
Please read over the .pdf and skim over the technical allegations on how the Tesla Model 3 reads/interprets voltages from the pedal, and how the ADC its associated voltage-reference could mess up here... at least for the EEs on this site who can follow the discussion.
---------
EDIT: The overall argument is as follows:
Root cause: high torque required by the power-steering uses more than 300 Amps of current. Tesla Model 3 12V battery can only supply 100 Amps, and the DC-DC converter to the main batteries only supplies 200 Amps. Low-speed turns requiring more torque could cause the 12V line to drop down as low as 2V.
The software calibrates off of a 1.65V reference point. If the software happens to calibrate during this voltage-drip, it will think that the 0.3V (reality) voltage represents 1.65V.
The accelerator pedal outputs 4V when fully pressed and 0.65V when the foot is off. 0.65V vs the 0.3V reference (thinking its 1.65V) will come back as "thought to be 3.5V" nominal, or nearly a fully pressed pedal (!!!). This will be locked in until the next calibration event. And that next calibration event will not happen for minutes.
This could very well lead to the behaviors discussed in videos like: https://www.autoevolution.com/news/what-made-a-tesla-model-y...
Methinks that won't work, lol. The numbers discussed are pretty insane, to be expected of a car-sized load / amperage.
-----------
What they need is a well designed power-network (Pi-network? Capacitor-Inductor-Capacitor-inductor-Capacitor) to isolate the microcontrollers + analog pedal reader from the rest of the car.
PDF was spitballing ideas for a software-only detection of this brownout condition. But I don't like that, any uC / analog sensor circuit that has 10V swings on a 12V nominal circuit is just going to go haywire in unpredictable ways. I'm frankly surprised that the uC is still sending CANbus messages on the pedal state despite the 12V supply voltage swinging down to 2V. I guess TI and STM32 chips / ADCs have enough stray capacitance / local capacitors to survive even when the sensor network/analog circuitry is browning out in Teslas.
Did we read the same paper? He takes no measurements, everything is based on theoretical assumptions, schematics and the three youtube videos he mentions in the beginning of the paper.
From the paper:
> The enclosed paper proposes a simple test that can be done to prove or disprove the explanation provided. This test has not been done by the author because of the cost involved with acquiring the needed Tesla Model 3 inverter PWB.
> The details of Tesla's Model 3 inverter design as revealed by Irish engineer Damien Maguire have been presented. These details were used to construct *a hypothetical model* of all hardware and software operations performed on the two accelerator position sensor (APP) sensor signals inside the inverter as they pass from the APP sensor to the electric motor controller.
He doesn't have access to the parts in question nor the car in question.
Also of interest is the authors own comments on his previous papers about sudden unintended acceleration:
https://www.autosafety.org/wp-content/uploads/2022/01/Note-t...