slightly OT comment, but I post anyway: At work I've been playing with White Rabbit PTP hardware. Developed by CERN for control/data-collection of the LHC. It is all open hardware and software, available at: http://www.ohwr.org/ Very roughly this WR-PTP over optical fiber is maybe 100x better at synchronization/time-transfer than current copper-based PTP implementations.
The typical nodes in this network are PCI-E cards with a Spartan 6 FPGA + a front-end plug-in card that can be simple GPIO, an ADC/DAC, or for timing applications a time-to-digital converter. Most of these plug-in frontends are also open hardware. Commercial providers sell the PCI-E cards for around 500 euros and plug-in frontends starting at 400 euros. The heart of a PTP system is the switch. The WR-PTP switch is also open hardware, but more complex than a simple node in the network. Commercially available for around 3500 euros. Does TI or anyone else have a copper-PTP switch available as open hardware or otherwise cheaply? All of this results in a system where you can do data-collection and control with better than 1ns synchronization, in a network where nodes can be up to 10 km away from the switch. For feedback control one obviously is more interested in latency instead of synchronization. I haven't measured/tested this, but the WR-switch is supposedly developed with deterministic ethernet as a goal, so it should be very possible to try. This would be very interesting for a distributed cnc-control where any endoint device (VFD, servo-drive, jog-wheel/UI-terminal etc) can be plugged in to the optical fiber network anywhere and you can potentially have 10km of fiber between devices. 10km of fiber inevitably means 100us of two-way delay, since signals travel at around 2e8 m/s in the fiber. Not sure of the potential advantages for conventional cnc-machines/controllers? Anders On Sat, Mar 8, 2014 at 11:12 AM, Joachim Franek <[email protected]>wrote: > > On 07.03.2014 23:16, Frank Tkalcevic wrote: > > Cool. Do you have any plans for it? > > The price opens some possibilities for remote drives > (3ph pwm and quadrature counter are onboard). > > But a IEEE 1588 capable eth on the LCNC is necessary > to relax realtime demands on the LCNC board. > > I think the community have to decide about the > specs (protocoll etc.) and can only do the work > together. > > Joachim > > > ------------------------------------------------------------------------------ > Subversion Kills Productivity. Get off Subversion & Make the Move to > Perforce. > With Perforce, you get hassle-free workflows. Merge that actually works. > Faster operations. Version large binaries. Built-in WAN optimization and > the > freedom to use Git, Perforce or both. Make the move to Perforce. > > http://pubads.g.doubleclick.net/gampad/clk?id=122218951&iu=/4140/ostg.clktrk > _______________________________________________ > Emc-developers mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-developers > ------------------------------------------------------------------------------ Subversion Kills Productivity. Get off Subversion & Make the Move to Perforce. With Perforce, you get hassle-free workflows. Merge that actually works. Faster operations. Version large binaries. Built-in WAN optimization and the freedom to use Git, Perforce or both. Make the move to Perforce. http://pubads.g.doubleclick.net/gampad/clk?id=122218951&iu=/4140/ostg.clktrk _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
