Re: Q: Replace ethernet cables with POF backbone on home Powerline?

x13 wrote:
> Hi all.
> 
> I regularly transfer many large files to/from my media device (like a mini NAS) 
> two rooms away. Though this box has both Ethernet and Wi-Fi connectivity, neither 
> provide a satisfactory throughput. So I caved and installed a pair of Devolo Magic
> LAN 2 Powerline adapters. Sure there's a lot of bandwidth loss, but even that is 
> faster than using a "networked" connection via my router.
> 
> The setup is very simple:
> 
> Mediabox -> CAT-7 (S/FTP) Ethernet cable -> Powerline
> 
> (two rooms further)
> 
> Powerline -> CAT-7 (S/FTP) Ethernet cable -> Laptop
> 
> For some time now, I've been thinking (dreaming) it would be great if I could replace
> my CAT-7 cables with glass fibre... and then came POF (Plastic Optical Fibre). Using 
> POF, there would be 0% signal loss, plus it's not subject to induction which is an added bonus.
> 
> So how would I go about doing this, if it's at all feasible?
> Besides the obvious POF cable (which I wouldn't mind running though the house as 
> it's so small), what else would I need? A POF switch? Something else?
> 
> It's possible that the signal frequencies used by Powerline may not be compatible 
> with POF (I hope not !). I have no idea. I'd appreciate the input of an electronics 
> guru on this.
> 
> Thanks.
> x13

Could you make this more complicated ?

What if the laptop is such a loser, it doesn't
deserve all this frippery in the first place ?

As long as you have *some* connectivity, that's a real win.

*******

This article says plastic optical fiber has a loss
of 180dB/km. That's not exactly a wonder-pony. Could
you do a set of EO modules to work with that ? Of course.
Plastic fiber is used for TOSLink at 6Mbit/sec, using
EOs that "cost a dollar" on the ends. All components
are optimized for that single application. The EO
probably doesn't even have AGC to compensate for
various lengths of fiber.

https://www.fiberoptics4sale.com/blogs/archive-posts/95051846-plastic-optical-fiber

    "SIPOF, today has a best bandwidth of 12.5 MHz-km
     and an attenuation of 180 dB/km"

This article helps interpret bandwidth-distance product,
which is typically dispersion limited (for the lousy
number stated above).

https://www.rp-photonics.com/bandwidth_distance_product.html

At 100 meters then, it could pass 125MHz. At
20 meters, it could pass 625MHz. The encoding
method then converts that into some baud rate.
At 20 meters, the loss would be 4dB. I expect the
connector loss, or launch losses, will add to that.
Module design with AGC could give maybe a 15dB budget
with ease. But, you'll be paying money for this too.
The solution no longer "costs $1", because the
modulation method and transmitter type have to
change to hit really high rates. The EO guys at
work, it used to be a bad joke, that like fusion
power, they'd promise us "cheap modules" and
we'd say in return "yes, only 50 years to go...".
Certain components can be cheap, like lasers
stolen out of DVD players for $1, but the modules
just never seem to get cheaper.

    Paul