Windows/Driver bugged on purpose so you must buy expensive routers.

so

500 u s

or

hen

ll

eption is universally weak here. The WISP signal runs through trees on thi s end for a couple of hundred feet, or maybe three. The Fresnel zone inclu des the water of the lake. I think the antenna is about a mile away and my antenna is about 20 to 30 feet off the water(5 to 10 feet less than the Fr esnel zone). The far end should be more like 50 to 100 feet up. I've neve r been able to find the antenna on Google or by sight. It seems to be in a gated community. There is also a small piece of land in the near end of t he Fresnel zone... maybe, I'm not sure how large the Fresnel zone is near t he ends.

ete brick that is used for sidewalks and walls. They run between 8 and 12 inches in size, maybe three inches thick.

ore so I can't give you the model number... or maybe I can if I dig out the data sheet I downloaded.

ntenna. I guess there's no magic to the unit, I just don't want to mess wi th what's not mine.

width.

sted *faster* than my network speed, then the bottleneck is somewhere else.

8.8.8.8 is googles DNS

e:

:

e so

t 500

h

ut

you

s

ves

t or

then

fall

n s .

eception is universally weak here. The WISP signal runs through trees on t his end for a couple of hundred feet, or maybe three. The Fresnel zone inc ludes the water of the lake. I think the antenna is about a mile away and my antenna is about 20 to 30 feet off the water(5 to 10 feet less than the Fresnel zone). The far end should be more like 50 to 100 feet up. I've ne ver been able to find the antenna on Google or by sight. It seems to be in a gated community. There is also a small piece of land in the near end of the Fresnel zone... maybe, I'm not sure how large the Fresnel zone is near the ends.

e

s

crete brick that is used for sidewalks and walls. They run between 8 and 1

2 inches in size, maybe three inches thick. g

ymore so I can't give you the model number... or maybe I can if I dig out t he data sheet I downloaded.

antenna. I guess there's no magic to the unit, I just don't want to mess with what's not mine.

ndwidth.

tested *faster* than my network speed, then the bottleneck is somewhere els e.

t m

I stand corrected.

Rick C.

A calendar might help. In the past, I installed wi-fi links to homes and small communities who could not obtain service from the major monopolies. Some have been operating for 10 years or more, during which time, the trees along the path have grown. I look at the signal levels occasionally to see how things are going and to try and anticipate any problems. As the signal levels drop, the speed at which the data connection operates decreases in an effort to maintain a low BER (bit error rate). You should be able to notice the changes in association or connection speed (not throughput) with the seasons. When the speed slows down enough for customers to notice, or when it can't go any slower, it's time add another 20 ft tower section. One link that is just barely above the tree line has been elevated twice in 15 years.

Cellular service sucks everywhere, no matter how strong or good the signal. It has more to do with the provider than the technology. Fortunately, Wi-Fi is not cellular (unless you get your WISP service from a cellular provider).

At 1 mile and 2.4Ghz, one Fresnel Zone is 23.5 ft at the middle of the path. If the path crosses the lake at 20-30 ft, you might see some cancellation of the signal if the lake rises or falls. The Fresnel Zone radius decreases towards the end points of the path, so if the lake is not in the center of the path, you might be ok.

If you want to do some direction finding with a direction antenna in the area, you could probably find the site using triangulation or trilateration. Driver around the gated community taking as many direction finding "fixes" as possible. Draw the lines on a map. Many will be reflections and produced useless lines, but most will converge on a single point. That's where to look.

The Fresnel Zone gets much smaller as you approach the end points. As long as the obstruction is NOT within the cucumber shaped zone, you're ok.

I really don't. I could have Googled for the answer, but it was easier to be obnoxious an ask. I eventually determined it to be a "paver" I've seen these, but never had any experience using them. In the mountains, everything is on the side of a hill. Pavers seem to be restricted to flat ground, which is in short supply in the area. I have seen driveways made from poured concrete, and then made to look like pavers, bricks, tiles with a steel tool that looks like a giant branding iron.

Just curious. I recall the discussion, but can't recall the model numbers.

If you were having problems with lack of signal, more gain is more better. However, you're right to leave it to the WISP. He might have some reason to use one particular antenna over another.

Agreed, but I've seen home networks with problems (usually bad wiring) that are slower than the ISP link. This is becoming a real problem when the customer orders Comcast cable data and gets 150Mbits/sec download speed, but their home network is 100Mbits/sec maximum. A few of the wireless links I've installed will do about 280Mbits/sec (using iPerf3) but the older computers, assorted boxes, and home wiring can't keep up. I had that problem on Monday, where the customer had installed a power line networking device that would only deliver

20Mbits/sec and was wondering why his download performance test results stunk. I suggest you not assume that your home network and PC's are faster than the wireless link and just test them to be sure.

Windoze use tracert Linux use traceroute It will be the 2nd machine listed and probably include your WISP's name as the domain. The 1st machine is probably your router. For example: C:\>tracert

Tracing route to [63.249.92.87] over a maximum of 30 hops: 1 If I can't see the problem in my use I don't bother them with it.

Run traceroute (or tracert) to 8.8.8.8 and see how much of the 30-40 msec is caused by intermediate routers and exchanges. From here, it takes 10 hops to get to 8.8.8.8 with most of the delay near the Google end.

On the public wi-fi, the local hospital (Dominican) blocks any type of streaming via UDP packets to guests. Patients are issued a login and password, which gets them enough streaming bandwidth to make a phone call, but little more. This creates big problems for web pages with obnoxious streaming video advertisements, that will not finish loading until the video is working.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

You do realise that you could have saved circa one hundred dollars by using a couple cheap Chinese woks in place of those rather pricey dishes? :-)

HINT: do a search for "wok antenna".

Here's a good starting point:

--
Johnny B Good

I was thinking more of the cellular service. It is so poor that I can't ma ke a call sometimes or the phone doesn't ring and then I get a voicemail. That happens rather randomly. If the trees and water had much to do with i t I would expect to see less of this in the winter and the worst would be i n thunderstorms.

The WISP guy says he puts the antennas low to get "under" the trees. Yeah, at least that's what my neighbor said. I think he just doesn't want to bo ther with mounting them high.

Similar frequencies, 900 MHz for the WISP. Cellular ranges higher, but sta rts at what, 800 MHz about?

900 MHz'ish. I looked it up once and use 950 as an approximation.

Doesn't matter if you can't see it on the map. I have a distance according to the Ubiquiti box and I know where I am. I can draw the circle and look in the community with Google. A tower isn't that easy to see actually. I 've seen some in cities that google adds perspective to and it actually fri ghtens me when the tower top swoops past as I move the map, lol. I have th is fear of water/heights that kicks in hard sometimes. Even when looking a t maps oddly enough. Here in the boonies Google doesn't do that.

Yeah, but like, how big is a cucumber in cubits??? The land is maybe 400-5

00 feet away and as high as I am, plus the trees, etc. It's got to be in t he Fresnel zone some.

I don't think they are used for driveways much, sidewalks and small walls f or sure.

Ok how? I think I did use a wifi computer to computer link once to transfe r files. I'm pretty sure it was much faster than the 5-7 Mbps I get from t he web. But it was nowhere near hardwire Ethernet speeds. They really doe s approach 1 Gbps.

1 1 ms 1 ms 1 ms 192.168.1.1 2 1 ms 4 ms 2 ms 192.168.125.1 3 25 ms 39 ms 35 ms 10.19.37.1 4 59 ms 48 ms 31 ms 10.20.10.1 5 70 ms 43 ms 256 ms ge-4-1-4-3481-sur01.alief.tx.houston.comcas t.net [50.203.240.233] 6 47 ms 31 ms 29 ms 96.108.140.73 7 167 ms 42 ms 37 ms ae-72-ar02.charlvilleco.va.richmond.comcast .net [96.108.140.77] 8 53 ms 47 ms 221 ms be-21508-cr02.ashburn.va.ibone.comcast.net [68.86.91.53] 9 71 ms 41 ms 69 ms be-10130-pe04.ashburn.va.ibone.comcast.net [68.86.82.214] 10 100 ms 78 ms 80 ms as36040-2-c.ashburn.va.ibone.comcast.net [7 5.149.229.86] 11 238 ms 96 ms 115 ms 108.170.246.65 12 41 ms 35 ms 39 ms 216.239.40.51 13 60 ms 51 ms 46 ms google-public-dns-a.google.com [8.8.8.8]

I'm not able to make much sense out of this. The numbers are all over the place. My WISP IPs seem to be 10.19.37.1 and 10.20.10.1, then he connects to Comcast.

No, this was an online game. In two browsers it worked fine and one it wou ld not run properly. Maybe this had to do with DNS as it only crapped out completely when I reloaded the page. Does the browser remember the IP addr ess of a page rather than use the URL for every access?

Rick C.

Kinda busy right now. More later.

Yech. The gateway machine is either 192.168.125.1 or 10.19.37.1. The Ubiquiti radio is probably setup as a wireless bridge, which would not show it's IP address in traceroute. Your WISP is using non-routeable IP addresses hung onto probably a single comcaset routeable IP (line 5). I can't tell why the jump from 192.168.125.1 to 10.19.37.1 with such a large increase in latency. I would need to know more about what kind of equipment the WISP is using. Notice the large variations in latency on line 5 and 7. There's packet loss (and retransmissions) in there somewhere.

I'm having the same problem. One extra change in non-routeable IP's and large variations between minimum and maximum latency. Why the big increase between hops 2 and 3 has me mystified. Wireless is fast, typically returning ping times of less than 2 msec for a 1 mile link. So, the extra latency is something else, probably inside your ISP's network.

Yes, that's likely. One of those IP's might a router used specifically for bandwidth management (rate limiting), traffic monitoring, and billing by usage. Dunno.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

I already have too much fried food in my diet. Therefore a salad bowl is better than a grease platter.

Do you realize how awful a parabolic dish or wok make when used with a USB wi-fi dongle? One problem is that most of the transmit power coming from the USB dongle does not hit the dish and go in the desired direction. In a real dish antenna, the feed is carefully matched to the dish to minimize spillover to have most of the RF hit the dish. With the USB dongle, only a small part of the tx power hits the dish. This comes into play with the "feed efficiency" part of the calculations. Typical is 50% efficiency.

Here's my calculations for the maximum gain for a DBS dish and a USB dongle feed at 2.4GHz: Same for using a DBS dish with a cell phone at cell phone frequencies:

In receive, there's no spillover problem because all the RF that hits the dish, also makes it to the feed. However, that causes a different problem, where the antenna gain is different in transmit and receive. How much different depends heavily on the feed design, which I don't want to go into now because a late dinner awaits my appearance.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

cast.net [50.203.240.233] >

ast.net [96.108.140.77]

192.168.125.1 is the Ubiquiti radio. I use that IP to access the control p anel.

I suspect the latency is a bit random. I see this in traceroutes all the t ime. Sometimes there are a few nodes that time out, but the end point is r eached.

cast.net [50.203.240.233]

ast.net [96.108.140.77]

et [68.86.91.53]

et [68.86.82.214]

[75.149.229.86]

he place.

10.19 and 10.20 seem to be my WISP. Then it goes to comcast in texas and t hen back to richmond (near me). Don't know why, but that's what they are d oing. Maybe some server required is in Texas. Then it goes up to Ashburn where the big network connection is for a large area.

mcast.

This guy doesn't really limit usage and billing is monthly or annually. I lucked out that he doesn't cap like a lot of them do. I use 10's of GB a m onth... no, my computer says 110 GB in the last 30 days. Seems like a lot.

Rick C.

Methinks you looked this 'gift horse' of mine far too closely in its mouth. :-)

I do understand what you're talking about but we shouldn't let practical limitations get too much in the way of assembling a cheap dish antenna out of a USB wifi dongle and a Chinese wok.

Thanks for the w1ghz antenna book links btw. I downloaded a couple of chapters but don't have the time right now to read any more. I'm pondering whether a recently purchased (from Aldi) Workzone 1800/2000 watt inverter genset is worth hanging onto which rather depends on whether the three cermet trimpots poking out of the inverter module's potting 'lake' would allow me to properly tune its performance (it bogs down dreadfully when hit by a 900W load before resuming pretty much the same rpms as at no load eco-idle with a 10v drop on the 235 open circuit output voltage. There seems to something seriously out of kilter with these gensets - the first two were even odder, if you can believe it! :-(

It's the reverse situation attributed to what used to be referred to as an "Alligator Station" (all mouth and no ears). :-)

Presumably, your dinner ingredients are mostly in the salad bowl. Bon

--
Johnny B Good

A bargain. I was thinking more like $539.

Oddly, in other newsgroups and mailing lists, people thank me for tearing apart their systems and calculating what level of performance they might expect. But, not in this S.E.D. I started mentioning RF a few days ago and you are now the 3rd person in as many days to suggest that I should shut up and possibly find something else to discuss. Perhaps politics or something off topic would be less offensive than RF. This is new to me and I don't understand it.

Nitpicking? Well maybe, except I call it analysis.

When I see a point to point 2.4GHz wireless link with 5ft diameter

30dBi gain dish antennas on each end, I often find that the large dishes are accompanied by excessive transmit power. Pushing 2.4GHz RF through 1500 ft of dense foliage is not going to be easy, so high power and high gain will probably be needed. For a 30dBi antenna, the maximum transmit power on 2.4GHz is 22dBm (158 milliwatts) which is near the minimum that than most commodity wireless bridges will produce (usually +20 to +28dbm). In other words, for this link to be legal, the power would probably need to be reduced to almost minimum.

TX Pwr Ant Gain EIRP (dBm) (dBi) (dBm) 30 6 36 29 9 38 28 12 40 27 15 42 26 18 44 25 21 46 24 24 48 23 27 50 22 30 52

If you had provided the equipment list and details which I requested, it would have been easy to calculate whether the system was within FCC Part 15 limits.

However, I still can't tell from "Yep, you guessed it" whether this is a 2.4GHz or 5GHz system. On 5GHz, 30dBi gain dishes are more common and more practical because they're smaller. The rules are more complicated on 5GHz and depend on which sub-band is used. The 5Ghz rules changed in 2014, which is probably after you installed your system: For example, you can now run 53dBm (200 watts) EIRP in the U-NII 1 band.

However, due to a lack numbers and details, my comments are mostly conjecture speculation. I was hoping that you could fill in the blanks with numbers. Since you apparently fail to appreciate my nitpicking and analysis, I'll not waste my time trying to be helpful on RF topics in the future.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Once upon a time, I had a technician working for me that was in the CAP (civil air patrol). During working hours, he would use the most sophisticated RF test equipment available to work on the company's radio products. After hours, he would put his own CAP radio on the bench, ignore all the technology that he had just used, and tune the transmitter for maximum with a light bulb, and tune the receiver by ear. Despite my attempts at interrogating my tech, no explanation was ever produced.

Suggesting that I throw together an antenna out of available junk is much like the technician tuning his radio with a light bulb. Yes, it does work. However, I would have no clue how well it works, whether the signal is clean, if it's on frequency, whether the deviation is set correctly, if the pre-emphasis curve is correct so the audio doesn't sound like garbage, etc.

Same with the wok. Yes, the signal increases over the tiny chip antennas used in most USB dongles. But, is it the best that can be done with the available junk? Would a reflector behind the USB dongle improve the transmit gain? Will it reduce spurious side lobes? Is the f/D ratio optimum? Is the feed beamwidth matched to f/D ratio for maximum efficiency? Is the VSWR across the operating bandwidth close to 1:1 and wide enough to cover the entire 2.4GHz band? Yes, you can throw something together that will meet some of these requirements, but I wouldn't call it a decent antenna. (Please don't remind me that my salad bowl antenna has all the same problems as your wok antenna).

If you don't have the test equipment, metalcraft skills, and general understanding of antennas to perform all these tests, at least make the attempt to understand why you might have problems trying to use the antenna. For example, I've seen as much as 10dB difference in gain between transmit and receive on a dish with a badly designed feed.

It's a very good book on parabolic dish antennas. Skim this section on high efficiency feeds and notice the effect of feed design on overall dish efficiency. Also, search for the word "spillover".

Ummm... that's a rather drastic change in topic.

This one?

2 winters ago, I did battle with a similar Genrac inverter generator. I forgot the model number, but something like this: The output voltage was stable at half of max load, but engine was running up and down ever few seconds. I eventually replaced the carburetor and fuel filter which solved the problem. I later rebuilt the carb and found the jets were clogged with tar. Check YouTube for repair videos on other brands of similar inverter generators (Genrac, Honda, Kipor). Also, tuneup instructions for similar inverter generators.

I suspect that I first coined the term and was first to use it in alt.internet.wireless. That's what happens when someone installs an RF power amplifier on their wi-fi router. Everyone can hear the router for miles, but nobody can connect to it.

The salad bowl antenna has a big hole punched into the bottom for the feed pipe and is therefore unsuitable for holding a salad. Dinner turned into a can of chicken soup, a small romaine salad, and a raisin bagel with peanut butter and blackberry jam.

Argh. 1AM. Time to kill the computer.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Oops. You're right. The term alligator was in use during the 1970's and perhaps the 1960's in ham radio. I built my first repeater in

1965: Well, if I didn't invent the term, perhaps I was the first to introduce it in the wi-fi mailing lists and newsgroups? Nope, only 2002. Ok, I yield. I didn't invent the term.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

few days ago and you are now the 3rd person in as many days to suggest that I should shut up and possibly find something else to discuss. "

Learn Krav Maga. Once they hit the floor 20 or 30 times they will be quiet.

You can do that ? Well OK on the intended output but if the case leaks or whatever, all bets are off.

Lasse Langwadt Christensen wrote in news: snipped-for-privacy@googlegroups.com:

It works, however, if it is not a crossover cable the full Gb capacity is not present. It is the different twist rate that allows the system to detect it and switch internally.

This is of course a reference to full capacity cat 6 wiring. A cat 5e does not do full Gb/s capacity on a Gb link.

Jeff Liebermann wrote in news: snipped-for-privacy@4ax.com:

I never said it did. YOU need to learn how to read better.

Jeff Liebermann wrote in news: snipped-for-privacy@4ax.com:

CAT 6 was "made to perform" at Gigabit speeds.

CAT 5e was meant to keep up and OPERATE on Gigabit hardware. But I can gaurantee you that a proper CAT 6 standard ethernet link cable performs faster than a CAT 5e. The connector leaks alone causes the interface to back down.

You do not know what you are talking about.

Sure folks have been using CAT 5e in place of CAT 6 BECAUSE IT STILL FUNCTIONS. CAT 6 cable is more expensive BECAUSE it matches the standard SPECIFICATION, and it proves that it does in operation! It does not cost more because of some magic claim.

So you put some CAT 5e in some offices.., WOW.

I built satellite baseband gateways with double redundancy feeding to and from a pair of 8 meter dishes. 17 US cities and 11 in Austrailia.

I think I know about link optimization.

Even as it relates to such a mundane thing as a standard ethernet link cable. You see... it IS the minutiae that matters.

Jeff Liebermann wrote in news: snipped-for-privacy@4ax.com:

Again, dipshit... I never said that.

In: Message-ID: or:

You wrote: "It works, however, if it is not a crossover cable the full Gb capacity is not present. It is the different twist rate that allows the system to detect it and switch internally."

Ummm... please tell me how the gigabit ethernet NIC can measure the cable twist rate?

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

At what length ? Full 100 m or something less ?

After all, the 100BaseT and 1000BaseT are not electrically that much different. Of course there is two extra pairs, but the symbol rate is the same but 1000BaseT uses 5 level pulses. Thus a slightly better Signal/Noise ratio (SNR) is require at the receiving end of a 100 m long connection. The wire diameter and dielectric losses will determine the losses and hence reduction of SNR. Bad twist may also degrade the crosstalk and hence cause more errors.

Anyway, for cables less than 100 m long, the 1000BaseT will go through Cat 5, 5e or 6 cables at full speed.