Ethernet over Power (480V)

johansen · Dec 18, 2020

An ato transformer with a capacitor across one of the coils will couple the signal directly into the output, without attenuating the signal. Obviously, one capacitor, not two.

motion guru · Dec 18, 2020

digger doug said:
175' might not be festooning, rather it might be bus bars.

Can you run an IR link ? Put stationary unit at one end of rail, shooting along the rail.
pickup on bridge.

Do you need to just get to bridge, or all the way to the trolley ?

Might be a simple mirror on the bridge to turn the IR 90 degrees.

Just to the bridge - and the power is already on bus bars . . . with the favored solution using those same bus bars for Ethernet.

DDoug · Dec 18, 2020

motion guru said:
Just to the bridge - and the power is already on bus bars . . . with the favored solution using those same bus bars for Ethernet.

I understand the original request sir.

You'd like to use that power system for Ethernet.

I can read.

I recall seeing about 35 years ago (and handling) an IR t/r unit, cast aluminum (made for outdoors) that you aimed like a sight. Distance was supposed to be pretty long.

But whatever..... :crazy:

motion guru · Dec 18, 2020

No offense intended Digger - just want to preserve a working concept if we can that is cheap and proven. I know zero about IR (assume Infrared laser?) transceivers . . . the first google search showed something that looked a bit spendy.

I wonder how much this would cost? FSO IR Laser Wireless Links

My experience with using lasers on long crane runways to measure position is that that even the slightest skew in the bridge requires that you have a massive target on the stationary part of the crane works. If you mount the laser on the stationary part of the crane . . . then you have to figure out how to get the signal to the moving part of the crane. With an IR solution, we would need to figure out how to keep the stationary and static units aimed at each other even with a little skewing of the bridge on the runways. I'll check it out.

D Bronson · Dec 18, 2020

MG:

Take a look at QEI. They supply scada communications equipment for substation control. https://www.qeiinc.com/

Also, your preferred vendor Siemens has similar equipment: Powerline Carrier – PowerLink IP | Smart communication | Siemens Global.

We have no experience with either's product offerings - so, sorry, no help there.

Regards,

DB

motion guru · Dec 18, 2020

Thanks DB . . . however those products are rated for Medium Voltage power transmission lines. These operate at a minimum of roughly 10x 460V.

I found 2 each 50VA 480:120 transformers in our test inventory and am going to give them a shot on Monday.

Milland · Dec 18, 2020

motion guru said:
I found 2 each 50VA 480:120 transformers in our test inventory and am going to give them a shot on Monday.

So if it works, does it mean my question in #3 wasn't totally dumb?

thermite · Dec 19, 2020

Milland said:
Thanks for that.

Figured it was a dumb Q, just didn't know why...

It's done all the time anyway. Small-value caps just carry the freq AROUND the XFRMR. Just buy it, tested and proven. Do NOT "DIY" it!

I'd run flexible CAT5-S meself. Or fibre.

The "patch cord" stuff, even when "screened", is meant to tolerate lots of bending.

I HAVE run point-to-point Ethernet laser-shots, too. Had a busy road to get across. The gear is cheap.

Not really a big fan of "WiFi".

The fewer potential intruders, "inherently", the lower the risk and lesser the cost of mess-with to guard against malice or accident.

Residence here is wired Old Skewl Gig-E over copper to one or more points in every room, for example. That's faster than the ISP's uplink (25/25) so no barrier to progress.

Visiting family with their menagerie of hand-helds get a "throttled" and heavily firewalled / proxy-filtered WiFi. Only.

The few who still run Microsoft go down the road to Mickey-D once or twice a day!

Never ONCE fails they have several cooties! tcpdump -i finds them. Then I firewall-off their MAC address!

You can always tell a WinSERF, if only by tracking the rude traffic of their ever-present parasites!

You just can't tell them MUCH!

A(ny) shop floor will have similar risks of exposure to WinCrobes and newer chrome/Android/Linsucks/Mac malware.

Best to have a plan for that "up front".

Not after it has already harmed your operation by AT LEAST sucking all the bandwidth off the local network.

Strostkovy · Dec 19, 2020

At one point there was a standard for wireless communication intended for dorms and apartments and such. An IR transceiver was mounted on the ceiling, and and another was on an ethernet dongle that connected to the computer. It was led based and only required a line of sight, and no aiming at all.

I personally think this system would be more reliable, but maybe for the data rates a bunch of lost packets don't matter much.

IIRC more modulation frequencies are available than for wifi, but I'm not totally positive about that. Speed is certainly lower, but not awful. Transmitters and receivers are also very easy to physically shield from sources of interference with a small shroud.

thermite · Dec 19, 2020

Strostkovy said:
At one point there was a standard for wireless communication intended for dorms and apartments and such. An IR transceiver was mounted on the ceiling, and and another was on an ethernet dongle that connected to the computer. It was led based and only required a line of sight, and no aiming at all.

I personally think this system would be more reliable, but maybe for the data rates a bunch of lost packets don't matter much.

IIRC more modulation frequencies are available than for wifi, but I'm not totally positive about that. Speed is certainly lower, but not awful. Transmitters and receivers are also very easy to physically shield from sources of interference with a small shroud.

EM spectrum is continuous. Absolute top of what we call "RF spectrum" is the bottom of the light band share. Radiated heat.

IR's transport layer is not slower. Latency and error-correction for S/N ratio ain't as easy.

Too many natural heat sources around, plus a ton of other man-made players.

Strostkovy · Dec 19, 2020

thermite said:
EM spectrum is continuous. Absolute top of what we call "RF spectrum" is the bottom of the light band share. Radiated heat.

IR's transport layer is not slower. Latency and error-correction for S/N ratio ain't as easy.

Too many natural heat sources around, plus a ton of other man-made players.

IR LEDs are actually near-IR, and are better described as just redder than visible. Things have to be pretty hot to emit in that spectrum, but there are plenty of things that get that hot in industry.

The reason the protocol I was talking about is slower is because it is modulated on a carrier frequency to distinguish it from other IR channels and background light.

Background light is comparable to electrical noise in radio communication, (or over power lines) but easier to shield against (because it's light) and less susceptible (hot things don't have sharp temperature transisents).

The exception is arc welding, which is rough on both but not insurmountable.

Arclight · Dec 19, 2020

motion guru said:
We started down this path when we embedded 3 motion control enabled servo drives inside a 15 foot long 16 inch diameter spinning tube that was fed via a slip ring. The drives were mounted to a carriage that travels inside the tube from one end to the other. Having communications go through a single flex rated power cable inside the tube and a slip ring to go from stationary tethered point outside the tube dramatically saved cost and complexity. This worked well @ 230VAC using a pair of consumer grade units from NetGear.

Now we are looking to solve this problem with auto positioning cranes where adding flex rated Ethernet over 175 feet of bridge travel is more complex and more expensive than if we can get this system to work @ 480V.

I work in IT, and I would trust any of the new 5.8Ghz enterprise WiFi devices over one of the power line solutions. If they already have managed access points from Cisco, Meraki, Aruba, or even Ubiquiti, they can spin up another "virtual AP" on the existing bandwidth and equipment, but connecting to a private VLAN for the motion control system.

You can also get ethernet bridges that run on spectrum your customer doesn't use, such as 900Mhz, 10Ghz, etc.

motion guru · Jan 14, 2021

We have successfully tested our system with Ethernet/IP between a PLC and HMI over 120, 230, and 480 power . . . then when doing an IP search to figure out if this was something that was anywhere else in the public domain, we stumbled across this - News: News | United States of America

Looks exactly like what we need already UL approved and ready to go!

garyhlucas · Jan 14, 2021

What is the cost of this system? Looks very good and those guys have been around for a really long time.

Ohio Mike · Jan 14, 2021

Very interesting, I'd be curious how this costs out.

motion guru · Jan 19, 2021

Well . . . the maker of this product will not sell it to us for use on an existing craneway power track system - we must purchase all new which is a non-starter for this application where we are adding automation to an existing set of cranes. Back to the drawing board.

Mud · Jan 19, 2021

What is their stated reason for that?

JST · Jan 19, 2021

Thermite's point about coupling capacitors is valid.

"Y" rated capacitors for 277 V should be available, although I have never had to use them. They will already be UL, so it would be fairly easy to couple to the line.

If you devise a unit to couple the signal, you might need to run it through ETL or similar however, for things "used in a workplace" but not carrying a listing for the usage.

Given that, the (already UL) transformer for connecting to a unit that also already has its UL mark is going to be way less work, and offers the potential to be ready to go as soon as proven to work.

sfriedberg · Jan 20, 2021

It has been more than 15 years, probably more than 20 years, since I last looked at datasheets for Ethernet isolation/coupling transformers. This was back when an Enet transceiver was a chip, rather than a integrated IP component of a system-on-a-chip, and the input circuitry had a level isolation transformer so you didn't have to worry about differing ground potentials for Enet gear plugged in at buildings with different service entrances and grounding electrodes. ISTR the bog-standard isolation transformers were good for several kilovolts, which surprised me (and made it memorable) as they were teeny-tiny devices.

If today's specs are comparable (at least for comparable levels of integration), there may be enough margin to operate Enet over 480V even if the maker isn't marketing them that way. Does nothing for UL certification, of course, which is probably required by motion's customers.

countryguy · Jan 22, 2021

Off today and wandering the threads. sucked in too but only as 'interested w/ no clue'. But fun topic here.

I went looking for point to point over high voltage rather than ethernet as a term. Found PLC term as discussed here as Power Line Communications in a SciDi article:
Power Line Communication - an overview | ScienceDirect Topics

a PLC grid comms device-
ET9 PLC equipment - Power Line Carrier Communications - Iskra

and a GE device:
Communications: e-terrapowercom

Ethernet over Power (480V)

Stainless

Diamond

Diamond

Diamond

Aluminum

Diamond

Diamond

Diamond

Stainless

Diamond

Stainless

Aluminum

Diamond

Stainless

Titanium

Diamond

Diamond

Diamond

Diamond

Hot Rolled

Similar threads