Monkey Brains

Mediocre eyesight, but you have great imagination to think that a thousand wi-fis aren't colliding furiously. Your new link is at an unused part of the spectrum, won't turn unreliable this year.

Every IOT gismo that uses 'b' or 'g' to save money is taking chunks out of the usability of WiFi. So is your microwave oven and cordless phone.

Oh, wireless is certainly very useful. I just don't think it is useful for /everything/.

Nor earthquakes.

Recently.

Nice, but no hand rails? However, I would guess(tm) that it was installed to protect the roof, not to make it easier climb. When I was working on radio stuff on various roofs, there were usually rubber mats installed between the roof entry hatch and the HVAC equipment or antenna farm to protect the roof. Most of what I found had crumbled and in need of replacement. I asked why they didn't build a wood or steel walkway and was told that the semi-sharp edges tend to dig into the roofing material can cut holes and slits. I just gave myself a quick rooftop view of the various Santa Cruz industrial parks with Google Maps. Such walkways are not very common in my area. I only found one.

Is the roof EPDM rubber, TPO vinyl or modified bitumen torch down? Methinks you should double check with your roofing contractor before partying on the roof. I'm not a roofing expert and therefore hesitate to give advice, but what I've heard suggests that adding a ton or more of human weight on most roofs is a bad idea. The roofs I've that have a party area, have a deck (wood or plastic) built over the actual roof. For smaller areas a "widows walk" platform. Both types have perimeter hand rails which your roof lacks.

This looks like a good roofing tutorial without the usual sales pitch: "Top 3 Flat Roof Systems Explained"

So, where are you going to put the solar panels?

We had something similar here in 1989.

We lost power for a couple of days, so we had an all-block ice cream party before it melted.

I don't know of any antenna alignment problems. It would take Gs of horizontal acceleration to move an antenna, and a fraction of a G will tear a building apart.

Brick chimneys (like ours) commonly crumbled, so an antenna strapped to one of those would fall. The big killer in San Francisco was bricks.

My wi-fi's at work and at home always work. Collisions get resolved in milliseconds. It all works, and uses a fraction of the potentially available spectrum.

Our wi-fi at home probably collides with ten others. That's the advantage of short-range, high-attenuation RF nets; collisions are limited.

Broadcast TV and FM are anachronisms, and getting rid of them would open vast spectrums.

That microwave dish on our roof costs something like $60. High-volume electronics is incredibly cheap. So a town could be covered with thousands of gigabit microcells at a tiny fraction of the cost of running wires and fibers to every house.

The future will be different.

Not for power, water, or gas. But why not one wireless network for all data to your house? I'd love to get rid of phone boxes, dsl modems, wifi routers, hubs, poking CAT5 and coax through walls, sprinkler controllers, garage door openers, light timers, separate security systems, scattered smoke and CO detectors, cable TV contracts and wiring.

On the commercial side, even low voltage/data wiring is run in conduit here. That's fabulously expensive and inflexible and ugly.

I'm always impressed by how many people refuse to consider progress, and fight against speculation on what their own profession is capable of. Most people spend their energy shooting down ideas instead of playing with them.

My earthquake experience was mixed. In 1989, we had the Loma Prieta earthquake: Most DBS dishes maintained alignment if the building wasn't damaged by the quake. Some buildings shifted on their foundations, which certainly caused alignment problems. Anything mounted on a chimney went out of alignment. In my area, every chimney had some kind of damage (except one that was build around a large steel pipe). On the commercial side, I lost 2 out of 7 point to point microwave links to alignment problems. I lost 2 more when the tower respective owners decided to adjust the tower so that it was properly plumb (vertical) and everyone had to realign their dishes. The heavy 1 meter fiberglass satellite dishes (mostly paging downlink and broadcast audio) were also mixed. All continued to work, but at reduced signal levels that were insufficient to trigger LoS alarms. We realigned them eventually.

As Rob mentioned, the 60GHz Mikrotik dishes are self aligning and should survive a shake that leaves the building intact. (I didn't know that it had that feature). It also makes realignment much easier.

The common 20" DBS dish is quite light. However, the J mount is not very secure. It doesn't take much to ruin the alignment as I've found when helping friends and neighbors with their installations (in the trees). The usual problem is that lag bolts used to secure the dish to a wooden part of the house. The bolts loosen over time.

If you need some entertainment, attach an accelerometer to a wood mallet and beat on the dish at its center of mass. Watch the peak acceleration on an oscilloscope: Ground motions during the earthquake were unusually large at near-source distances for an earthquake of its size, registering up to 2.2 g (vertical) and 1.7 g (horizontal) near the epicentre and up to 0.8 g (vertical) and 0.7 g (horizontal) in the city centre. Aim for 1.7g acceleration and see what happens to the dish. Loma Prieta 1989 was 0.64g horizontal.

onsdag den 22. august 2018 kl. 17.22.05 UTC+2 skrev John Larkin:

the whole building might move

I suppose that might mis-align a dish.

A few years ago, I read that the biggest earthquake risk to homes in my area (mostly single-story wood-frame, built in 1960 or so) is chimney failures. The codes at that time did not require reinforcement. An unreinforced chimney falling over and crashing through your roof would definitely spoil your whole day. Even if it doesn't hit somebody, it could easily be the difference between a house which takes superficial damage, and a house that's red-tagged as "uninhabitable, do not enter" for months.

The chimney at our house had taken some minor cracking damage during Loma Prieta (several years before we bought the place) but was evaluated as sound during the pre-purchase inspection.

About two years back, I got a Round Tuit and did something about it... had a local mason disassemble the chimney, brick by brick, down to the "shoulder", and then rebuild it to meet current code standard... rebar up each corner, steel straps going through the wall and tying it to the house beams, a larger-diameter flue, cement "fill" in each corner around the flue and rebar, and a new cap and spark arrestor. Visually you really can't see any difference from outside, but internally it's a very different beast.

Good, cheap investment.

And, I decided not to ever fasten an antenna of any sort to it... turns out that the repeated bending stress from an antenna being blown around by the wind is one of the things which tends to weaken brick antennas over time. I installed a separate mast at ground level (goes through a hole in the eaves, bolted to the roof beams, and sunk into concrete at ground level) for the TV antenna.

The "resolution time" is an irrelevant metric.

In cities over here (and in high density US cities) it isn't unknown for your wifi to work until someone else's new installation clobbers it.

And in the ISM bands you have to accept such interference; that's the cost of them being unregulated.

That entirely depends on the neighbourhood and density. Over here (where 100 miles is a long way and 100 years isn't a long time) wifis frequently become "slower", and it can be difficult to attribute a specific cause let alone a cure it.

One common use of wifi over here is video streaming, where each terminal can soak up noticable bandwidth over a long period.

One use of the 60GHz band that I expect to see is as a wireless replacement for the cable going to monitors. That will soak up bandwidth, hopefully only over a very short radius!

A couple of decades ago someone noted that "in the future everything that is transmitted by wires will become wireless, and everything broadcast will be transmitted over wires".

After pondering that for a couple of months, I started to suspect that would come true.

That's why it is useful to get back to the physics and thermodynamics; they change very slowly.

I've helped my friend install a dozen microwave links, mostly in the

We would sweep them horizontally and vertically several times to find the best signal, measured to 4 digits with a meter on the AGC test point. Tweaking microvolts with a big crescent wrench is fun.

Eggs in basket? Security? Physics of the propagation channel?

Wouldn't we all, but that all has a different non-financial cost.

It must include the radio in the dish's focus because I see coax instead of waveguide. I used to install similar things but I didn't know how far it has advanced. Or radios were huge and mounted on the opposite side of the mast with a short waveguide.

But cell tower antennas don't seem to have changed as much.

What you see is not coax, it is UTP. The entire radio and router is integrated in the feed portion of the dish, and a 1 Gbps ethernet link is going inside, where a power-over- ethernet injector is used to send power up. You connect the UTP to your network and get the IP packets in and out.

The router can do many things besides simple routing, these devices are quite versatile. Of course they can do bridge mode as well, that is the default configuration.

I'm not yet 100% sure that what John has is actually the new 60 GHz model, the sticker he showed is applied by the installer and there is a 5 GHz model which has been on the market longer and is cheaper. It has the same functionality except no self-alignment and of course it is slower. Those are the ones available for $70 a piece. (the $300 is for a pair of 60 GHz units)

Is the self-alignment mechanical?

The systems I helped install were around 20k per pair.

Lower cost could allow sufficient redundancy. If the dishes are a few hundred dollars then you could have several directional dishes on the roof of each large building to link them in a wider mesh, each aimed at a different elevation and direction. Small buildings in between could use omnidirectional antennas to form a smaller local mesh that links to several of the larger buildings. The small buildings could also have multiple antennas with intermediate angles each.

Of course it would suffer to an extent when any part goes down, but that happens when a router fails somewhere.

And of course I know I'm missing something.

My friend maintained the system that linked the WTC to the NY metro area airports. It never went down until the collapse.