| Author |
Message |
DarkRothaus Guest
Free account, no CAN development support
|
25-06-2010, 16:54 Subject: Two "buses" connecting |
Quote |
|
Hi,
Perhaps the question is naive and can be dismissed with a simple 'No'...
I could just try it, but I haven't done so yet, and I'm not entirely clear about the potential side effects.
Vehicle:
Mercedes-Benz C-Class 202 facelift (12/97)
Upgraded with an E-Class 210 instrument cluster and multifunction steering wheel.
The vehicle has a CAN bus with a speed of 500 kbit, which I'm not particularly interested in for the time being.
The CAN bus, with its features like mirrors, interior lighting, reverse gear detection, window lifters, etc., operates at 83.3 kBit and can be accessed at the passenger-side rocker panel.
I had already connected my CANUSB device, and then I could 'see' and control things like the mirror on my PC.
The main reason I bought the adapter was to control my car PC via the multifunction steering wheel and to display data in the instrument cluster. To communicate with the steering wheel and the instrument cluster, I need to connect the CAN USB adapter to the CAN bus line that runs from the instrument cluster to the (non-existent in my case) radio. Everything works perfectly. This bus also operates at 83.3 kbit.
I'm not entirely clear why the two buses are separated. Sometimes, information is present on both buses, such as the speed signal, but it's then identified with different IDs. The only possible explanation might be... Reduce cable length or minimize error susceptibility?
Since, as far as I understand, all participants have equal rights in the CAN bus, theoretically, there shouldn't be anything preventing connecting the two buses to a single CAN distributor? (Assuming there are truly no duplicate IDs).
Do I think I'm being too naive, or are there other obstacles besides bitrate and ID? Perhaps the 83 kbit bandwidth is only sufficient for a limited number of transmitters/receivers?
Why do all of this? It would be interesting to process information like reverse gear data from the car's computer to, for example, implement a curb detection system. A second CAN USB interface is likely also a solution, but it would be quite expensive.
Thank you and best regards,
Patrick.
|
|
| Back to top |
|
 |
p.ostrich Guest
Free account, no CAN development support
|
29-06-2010, 9:56 Subject: Re: Connecting two "buses" |
Quote |
|
Hi, I'm sure a lot of the same data will be received, but there will also be a lot of data that is unnecessary for each individual bus.
At 83 Baud, it might be a good idea to avoid unnecessary data. It would certainly be possible to combine them, but, as you mentioned, IDs must not be duplicated.
There are several reasons why this is separated, but for your project, it's important to first determine how much data, of what type, and in what quantity is coming on each of the two buses.
|
|
| Back to top |
|
 |
DarkRothaus Guest
Free account, no CAN development support
|
01-07-2010, 11:35 Subject: Re: Connecting two "buses" |
Quote |
|
Okay, thank you.
According to another source, the separation was likely due to the relatively low data rate of 'only' 83 baud.
Is there a tool that allows me to read out the occupancy rates of buses?
Or does the 'simple' calculation of Messages * 8 bits per time unit suffice?
I heard that the utilization rate shouldn't be much higher than 50%.
'Nothing is likely to break' during the attempt, so I think I'll try a practical experiment.
|
|
| Back to top |
|
 |
|