3
\$\begingroup\$

I've run into the concept of IEEE 1588 synchronization over Ethernet several times, and am curious what its most common uses are and the details of why it is required. For example, I read that cell phone towers require time synchronization on order of microseconds, but I am not sure exactly why that is the case.

Judging by the amount of support for it, clearly, IEEE 1588 is an important and widely utilized standard. I can think of some relatively obscure cases where I would want microsecond time sync between devices miles apart, but I am not sure what exactly industry does with it.

If anyone has had any experience with it, please share an example or two, or generally talk about what you think are the most important/common applications of it.

PS. Someone please re-tag this appropriately, I don't have enough rep to add tags, and apparently this has been an obscure topic around here.

\$\endgroup\$
2
  • 2
    \$\begingroup\$ Long distance time sync is generally done with GPS. Cell phone towers generally have GPS receivers for time sync. A network using IEEE 1588 generally derives its time from GPS as well with an IEEE 1588 grandmaster clock with a built-in GPS receiver. IEEE 1588 is used when you want sub-microsecond time synchronization over short distances (e.g. within a datacenter). \$\endgroup\$ Commented Jul 5, 2014 at 21:26
  • \$\begingroup\$ Thanks, everyone, for your help. All the answers were helpful. I'm accepting the answer that I had least previous knowledge about. \$\endgroup\$
    – Dmitri
    Commented Jul 7, 2014 at 15:44

4 Answers 4

2
\$\begingroup\$

Electrical power substations are perhaps one of the biggest applications. Phase measurements of voltage and current need to be taken more or less at the same instant (microsecond precision) over spatially varied locations within the substation.

Since substations are poor GPS environments, precise time is distributed via IEE1588 - which can also go over fibre if necessary too. A "grandmaster" clock for the network is usually a GPS-synchronised clock in a location with good sky visibility - perhaps on the edge of the facility.

\$\endgroup\$
2
\$\begingroup\$

Synchronizing time stamping of data acquisition over an Ethernet network is an example.

The individual nodes will each have a clock, but there will be an unknown offset and a slight drift between clocks. If you don't care within milliseconds about synchronization then it's not necessary but many applications are not that sloppy and need microseconds or better sync to achieve control or compensation.

Of course there are proprietary or ad hoc methods that achieve the same ends, but standards help ensure interoperability and may allow COTS boxes to be combined with bespoke ones.

\$\endgroup\$
2
\$\begingroup\$

Regarding cellular towers: CDMA require synchronization for the code generation and recovery, also, it helps reception on cell edges, as the signals from adjacent towers do not interfere much with each other. TDMA needs tight synchronization for time slot assignment and guard intervals. In both cases, a loose time base will result in lower usable bandwidth. Cell towers often use GPS receivers with OCXO for their timing source.

\$\endgroup\$
1
\$\begingroup\$

Motion control servo motors are typically controlled via Ethernet these days. Think CNC machines and robotics. They need microsecond time synchronization between all the axis of movement. 1588 is one of several protocols used to achieve that.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.