OBD-II and ELM327 (8), MITM Failure Analysis 2

Introduction

In school, the textbook illustrates theorems in sequence. It seems there's some path that we can get the answer. In reality, I can write some failure analysis with confident, but I would say it's wrong in this article. I think that's engineer's daily life, struggling in debugging, until reaching the answer

New Weapon

I bought STM32 Nucleo F767ZI in Mouser, which is high-end MCU fabricated in 90nm
  • Cortex-M7 @ 216Mhz,512KB SRAM,2MB Flash
    • Extremely fast processor, very large PM/DM
    • Stupid code still runs quite fast
  • 3x CAN,USB 2.0HS,Ethernet
    • With 3x CAN, I can do packet forwarding within MCU
    • I can play USB/Ethernet if desired


FedEx is quite efficient. Ordering on Friday and the EVB dispatched from USA and arrived to Taiwan on the next Tuesday. With some efforts, I completed CAN forwarding within MCU. In the scope picture below, the latency reduces from 400us to 20us (finally...)




HUD was NOT killed by latency

In the picture below, with extremely fast packet forwarding, the left red-rectangular is speed query from HUD. The right rectangular is Pi-CAR responses to the speed query. The latency between two rectangular is 4ms = 4000us, and HUD is still alive !!


I tried to enlarge the latency, Pi-CAR waits 0.1sec=100ms=10000us then responses to HUD. The HUD is still alive !!


Opps

So I bought new boards from USA, did some coding with STM32, and I was totally in wrong direction. Anyway, I still like the new EVB. Linux is hard to manipulate, while MCU is easier to use. I can use familiar tool (J-Link debugger) to analyze. It's extremely fast to forward packets, I would keep using it


Eventually, I put the equipment to the car, and the HUD is not working yet

Comments

Popular posts from this blog

OBD-II and ELM327 (1)

OBD-II and ELM327 (3), try to decode CAN package recorded by ELM327

OBD-II and ELM327 (2), the underlying interface of ELM327