[Etherlab-users] What Hardware

Andreas Stewering-Bone andreas.stewering-bone at gmx.de
Mon Feb 15 17:53:38 CET 2021


Dear Karl,

You have to sort your IOs in following categories:

- High Speed IOs, ex. force, position sensors
- Low Speed IOs, ex. temperature sensors
- Digital Interfaces, ex. RS232, Here you have to define data width and reaction time

Based on this categories you can decide
- only one master with one cycletime
- or one master with two cylcetimes

I do not think you will need a multi master system. This ist only used in very stiff and fast systems or you have a big data width to transport.

Depending on your application code
- complex Signal calculation and control
- or simple reacting on Signals

A small Atom based board can fit your needs. But this depends on your application needs.

On a modern Linux Kernel you will get good latencies with the generic driver at 1kHz. On the other side you do not have have to select a special board according to your selected Chipsets.

Best regards

Andreas



Am 15. Februar 2021 16:26:19 MEZ schrieb Karl Zeilhofer <karl at zeilhofer.co.at>:
>Hello users of EtherLab!
>
>We are planing a retrofit a control system in the industrial field with
>EtherCAT components.
>
>* 500 digital I/Os
>* 4 Analog Inputs
>* 15 Analog Outputs
>* SSI Input
>* Some 9.6kbaud RS232 ports
>
>Cycle Time should be in the range of 1 to 10ms.
>
>What hardware would you recommend?
>
>I have an offer for a Beckhoff IPC for around 2.000€ for an Core i7
>with
>4 LAN ports.
>Another offer is about 800€ for a similar desktop system.
>
>How critical is the decision for or against the supported Ethernet chip
>sets (e1000, e1000e or r8169)? Would the generic driver be sufficient?
>
>Kind regards,
>Karl Zeilhofer

-- 
Diese Nachricht wurde von meinem Android-Gerät mit K-9 Mail gesendet.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.etherlab.org/pipermail/etherlab-users/attachments/20210215/e9b1f8cd/attachment.htm>


More information about the Etherlab-users mailing list