Single master I2C bus and CSP

[lo...@lesmasson.com]

Hello,

First of all I would like to congratulate you on the fantastic work you did with CSP, it looks like a very solid protocol in and of itself ! 

I am investigating CSP as a possible solution for the cubesats developed in my lab, and one question came up on my part.

One of the strengths of CSP, as I understand it, is to alleviate the risk of single point failure by removing the situation of a single master node. 

However, I would like to know how packet forwarding to other nodes of the network is handled in the case of a single master I2C bus ? Would a multi master bus be obligatory ?

Thank you in advance for the help !
Cheers,

L

[Jeppe Ledet-Pedersen]

On 2014-02-04 14:42, lo...@lesmasson.com wrote:
> Hello,

Hi Louis,

> First of all I would like to congratulate you on the fantastic work
> you did with CSP, it looks like a very solid protocol in and of itself
> !

Thanks :)

> I am investigating CSP as a possible solution for the cubesats
> developed in my lab, and one question came up on my part.
>
> One of the strengths of CSP, as I understand it, is to alleviate the
> risk of single point failure by removing the situation of a single
> master node.
>
> However, I would like to know how packet forwarding to other nodes of
> the network is handled in the case of a single master I2C bus ? Would
> a multi master bus be obligatory ?

Yes, CSP devices are meant to run as I2C masters. You can still add
regular slave devices to your bus (sensors, simple subsystems, etc.),
but they will not be able to run CSP.

Hope this helps.

-Jeppe

[lo...@lesmasson.com]

Hi Jeppe,

Thanks for the swift answer, that is all I needed to know. It confirms what I suspected. :) Thanks again for the help !

Best,
L

[Johan Christiansen]

There is nothing that stops you from having multiple masters on an i2c bus and still have slave only devices too. 

An alternative would be a single master which polls all csp devices at regular intervals and acts like an active bridge between them. But that would of course be against the philosophy. 

--
You received this message because you are subscribed to the Google Groups "cubesat-space-protocol" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cubesat-space-pro...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.