Dual band Rootenna experience?

[valenti]

Just wondering if anyone has tried these:
http://pacwireless.com/products/R2T2458.shtml

It is a 2.4 / 5.8 GHz Rootenna (12 /15 dBi). I bought some thinking it could be the backhaul for a micro cell. I could start with 2.4 backhaul now, and then move to 5.8 later on.

But I haven't had much luck using the 5 GHz option. (using WAR2 + CM9) I was trying a 4 mile link, and I could get a connection using 2.4. But when I switched to 5GHz, nothing. Then I tried 24dBi Rootennas at each end and still nothing found on the site survey.

Just to make sure I wasn't doing something totally wrong with the 5Ghz connections, I tried connecting up with a laptap and an Orinoco Silver AB card, that worked fine. (maybe this thread is really about newbie mistakes with 5GHz?)

My intention is to use the dual band Roo for shorter connections, less than 3 miles. And it would be LOS, silo to grain leg, etc. I would have the second radio connected to an N connector on the backplate, hooked to an omni.

[luke541]

I have used about 6 or 8 of these for 5Ghz for distances up to about 6 miles and they work pretty well for me. Never tried them in 2.4 though.

What was your distance setting set at on the war2? Did you already try differant cm9's?

[valenti]

After trying a 24dB Roo at both ends, I gave up on 5GHz for now. 2.4 with 19dB Roo's works OK until some 24dB grids come in.

I've been trying so many things, I can't say for sure what my distance setting was. But that wouldn't affect the site survey results, would it? I thought that was only for association. I never saw anything on the survey, except "No sites found"

And I tried different Roo's with different WAR + CM9, but I don't think I ever swapped a CM9. Several combinations did let me connect in my basement with that Orinoco card.

[gunther_01]

Something sure seems amiss. 4 miles with LOS shouldn't be a problem. You should at least see something. It may seem simple but make sure you are using the correct ant. connector to go with your settings in Star. I have not used that particular Roo, but you should see something even if it's weak.

I dare say that maybe you burned out the RX side on your CM-9 while testing it close to your other AP.

You say OK with 2.4 and 19 grids. I would expect to see low 70's to mid high 60's with out even calculating it.

[rbolduc]

Something sure seems amiss. 4 miles with LOS shouldn't be a problem. You should at least see something. It may seem simple but make sure you are using the correct ant. connector to go with your settings in Star. I have not used that particular Roo, but you should see something even if it's weak.

I dare say that maybe you burned out the RX side on your CM-9 while testing it close to your other AP.

You say OK with 2.4 and 19 grids. I would expect to see low 70's to mid high 60's with out even calculating it.

I am using a 24 dbi roo and a 11 dbi sector at 5 miles and am getting a usable -80 signal, the 11dbi sector is feed with 50' lmr400 and a cm9..soon to be changed to a 26 dbi grid


Reed

[valenti]

Something sure seems amiss. 4 miles with LOS shouldn't be a problem. You should at least see something. It may seem simple but make sure you are using the correct ant. connector to go with your settings in Star. I have not used that particular Roo, but you should see something even if it's weak.

I dare say that maybe you burned out the RX side on your CM-9 while testing it close to your other AP.

You say OK with 2.4 and 19 grids. I would expect to see low 70's to mid high 60's with out even calculating it.

This particular link isn't LOS. I've even looked with binoculars, all I can see are trees. I set it up with 5GHz (Trango gear) in November and it worked for several months. The Trango AP got zapped about the same time leaves were coming out, I suspect it would have failed due to foliage anyway.

I do have the link working with 19dB Roos at both ends. I'm getting -78 both sides, noise is -93, at 2412 x4. I ran large pings overnight etc and the link seems solid, it passes about 8Mbs, which I think is the limit of what a War1 will do right now? (grids have been ordered to improve the signal)

But your comment about burning out the RX side worries me - I didn't even realize I could to that! Would the CM9 have different RX for different freqs? With the combo Roo, I could get a weak connection at 2.4 but see nothing at 5GHz. And it works at a hundred feet for 802.11A connections to a PC. So it seems like the RX is ok.

My thinking is that the leaves are eating more of the 5GHz signal. Plus, the Roo isn't exactly a precise antenna for aiming - at least in my hands!

[gunther_01]

5G with Non-LOS is going to be tough at 3-4 mile distance you have been refering too. In fact you can almost bet that it won't work. 2.4 is better for non-LOS and of course 900 would be your best.

In general what happens when a cm-9 goes "deaf" from a burn RX side, the RX signal drops about 20db. And I guess you can tell what would happen to an already low signal at that point. About the only way you can tell if the RX side went out is if you are graphing your link signals. Or are just aware of what your signals have been from the start of that links inception.

If you already have the PtP (this is a PtP? correct.) set-up you can swap out a new CM-9 to see also. As far as different signal for different freqs. A lot comes in to play there. Depending on how PAC makes the thing, you may need to realign (I would anyway). Changing channels in general will give you different readings depending on interference. But the change from 5g to 2.4 will be a large swing with Non-LOS, as you have found out. And since Trango is proprietary we can't really say that Star will work in the same setting. It may, but it may not, as well. It's not an apple to apples comparison.

My thought. Go with 900 and be done with the problems. If it's a backhaul, you want it to be stable even through foliage growth. 900 is going to help with that.

[c.davis]

This is a 'Correct me if I am wrong' type of post, so please do so.

I equate multi-band to multi-purpose... kind of like using a leatherman multi tool... sure, it does the job in an awkward and less than efficient kind of way but it never really does the job in a timely and/or desirable fashion.

When you use a multi-tool like a leatherman (or multiband antenna), you are making a sacrifice for a 'do all' compared to the right tool for the job and will therefore pay some sort of inconvenience/penalty.

In the case of a multi-band antenna, you obviously have more than one grill (one for each frequency), sure this is great in a pinch but really is what you have is layered grills which will, of course, absorb some signal, just like the handle of the trusty old leatherman has tools in the way of the one that you are using. I know that these are two totally separate frequencies but the fact remains that one grill is on top of the other and will cause some sort of interference or blockage at the source which never translates well as distance increases.

If you are using the 5.8 part of the multi-band antenna, the 2.4GHz radiator is still there and will still cause some issue(s), big or small, it is still there absorbing, deflecting or otherwise causing problems and really, not being the best tool for the job.

For example, if you are using 5.8 for a back haul and using 2.4 for ptmp and using the multi-band antennas in the same area/segment of your network... would not the 5.8 grill in your multi-band antennas, although not used or 'plugged in', absorb signal since they have the right pattern and therefore cause problems?

Like I said, correct me if I'm wrong, but I do think that there is some relevance to this.

[valenti]

C.Davis, I think you are right on target. My limited experience has lead me to put too much faith in the RF calculators. While this link isn't a good first choice, some of my other ones will truly be LOS and <3 miles. I think there will be an excess of fade margin there to use the combo Roos. I just need to overcome my frugality and laziness - it seemed like a great idea to be able to switch freqencies on the fly, without climbing a silo.

Gunther_01, yes it is a ptp link. I was trying to reserve 900 for customer links, but that might be a good solution. I'm also thinking about raising the antennas about 20' at both ends, that might get LOS.

I have been graphing RSSI on my Trango radios, and was disappointed to find out that I couldn't do that with StarV3. (right?) My understanding is that feature might show up soon in this release that fixes snmp/wpa. (I've been hitting refresh on the StarOS home page about 10 times a day waiting for the announcement...)

[gunther_01]

I too think Mr. Davis is on to something there. And of course makes total sence.

No offense here, but calculators don't take in to account for trees. So those calculations don't mean much with this link. Trial and error is all you have, and of course you have been trying, but that's how it goes sometimes.

StarV3 as of right now does not have a working SNMP platform in it (without memory leak) I too am waiting for it, and have for a while now. But it seems it will be whenever. I, personally, am not holding my breath anymore for SNMP. I won't deploy any clients without it. And frankly, I hate having my backhauls deployed without it either. I am relying on my client radios to tell me if my backhauls are working, ping graphs to give me an idea of signal qualities, and my other brand P2P link to help with bandwidth graphing. Some have been able to use Starutil in the past to over come this (I think) But myself I use standards for my monitoring. I have a hard enough time trying to get a webpage built, without learning how to script something up just to monitor a signal strength.

From some of the other threads, it doesn't seem that SNMP will have signals and such until the main driver is working and stable. We will be waiting a while it seems.

[lonnie]

The other thing to keep in my with multiband antennas is that they deliver noise to the radio that is way out of band from the frequency you are using. Noise is noise, so any extra noise you deliver, be it in band or way out of band, affects the receiver. Wireless is ALL about signal to noise ratio, there let's check the pros and cons for the dual band antenna.

Pros
Only need to stock a single antenna
Allows flexibility for quick band changes

Cons
probably cost more than a single band antenna
lower gain than a single band antenna of the same size and price
more noise than a single band antenna

I do not recommend them, but it is up to the end user to decide on their own.

[Stratolinks]

I did use a dual band parabolic dish antenna at one time. It was at a site where we pay rent per antenna so I wanted to feed it initially with a 2.4 connection and move to 5.8 later. I plugged the 2.4 cable via a tuned 2.4GHz surge suppressor into the "A" port on the CM9 and the 5.8 cable via a 4-6GHz tuned lightning suppressor into the "B" port. When I was ready to change, I went to the site connected via the other radio and then applied the settings and activated changes. I works like a charm.

The key thing is to have a parabolic antenna that has dual feeds. The dish is just a reflector so it is up to the element at the focal point to tune in the frequency it wants. Yes the dual band antennas are slightly leff efficient, but they do sometimes have an application that you have to determine if they are suitable or not.

[tog]

Comparing dual-band parabolic dishes to dual-band panel antennas is like comparing apples and oranges though.

[valenti]

Lonnie,

Would this noise be indicated in the client's noise value displayed on the main screen? (or is it at a deeper level, so hard to quantify)

I'll probably swap the combo (12dbi) for a 15 dBi Wartenna soon. Hoping to pick up 3dBm. But I can note the noise value too. (24dBi grids are due in Monday, that is the longer term solution)

I'm also wondering what switching these to horizontal polarity would do. I usually pick up 1 or 2 dB when I do that on a Trango 900.

[lonnie]

The Receive noise would show on whichever unit has the dual band antenna. Normally the out of band noise is attenuated by the antenna but with a dual band it is all delivered to the radio.

If things around you are vertical then a move to horizontal will be a good thing.

The main reason people use vertical is because they started out with omnis and just kept going. You can use polarization to reduce noise by at least 10 dB and as much as 20 dB in extreme situations.