Fwd: CAN HW compability

Job Race <[email protected]> Thu, 7 May 2015 10:50:28 +0300
Newsgroups gmane.comp.hardware.bus.can
Message-ID <CAGB76Pgc1N_TqU9aZWBk=T2xFHrScg=rDQHCdZmgU-293MCD4g@mail.gmail.com>
---------- Forwarded message ----------
From: Job Race <[email protected]>
Date: 2015-05-06 17:03 GMT+03:00
Subject: CAN HW compability
To: ‫[email protected]‬


Hi There

I am having an issue when connected to peugeot C4 picaso 2015

My objective is to filter some of the messages coming from the original
sensor unit so i cut the CAN wire directly from the original sensor unit
and connected a PCB with two CAN buses, CAN1 is connected to the original
sensor side and CAN2 to the car side

The PCB has two MCP2551 CAN transciver connected to a dspic30f5011
MICROCHIP microcontroller

It seems to work fine but when i look via the osciloscope at the side of
the sensors (CAN1) the signal does not look very good i.e. bit of ringing
on rize and slow fall times as in the picture below (picture 1)

https://drive.google.com/file/d/0B9RlBLf0YeD6NXo4ZEwtbWdPUW8/view?usp=sharing

Picture 1: CAN1 (sensor side)

When returning to original state (without the PCB & CAN returned) the
signal of course looks much better as in picture 2:

​
https://drive.google.com/file/d/0B9RlBLf0YeD6WU02VklkVWUtZFU/view?usp=sharing
​
Picture 2: CAN original

First i thought it needs terminating resistors as i basicaly disconnected
the unit from the reset of the bus but that is not the case as:
1. the original bus does not have terminating resistors
2. The bus rate is 125kbps
3. When i try to put any resistor between CANH & CANL it completly disables
the bus

My other thought is that the natural 47kohm resistance of the MCP2551
(between CANH & CANL) might cause the issue

What do you think?

Thank you &
Best Regards
Job