Wireless Service Providers → Xagyl vs. XR9 Rootenna vs. Arc

So after severe disappointment with UBNT 900mhz performance and for some applications the Moto 900mhz just not cutting anymore I go back to my roots and consider the 'tik solution again.

The above questions say it all:

Would you use the Xagyl 900mhz card or the XR9?

And for the antenna?

Pros and cons appreciated.

I have tested the UBNT 900 and not been impressed. Ended up selling my test equipment to a customer who wanted to connect a small handful of building spread across their property where 2.4 wouldn't quite cut it.

Anyway, I use MT & XR9 in a Rootenna. Works great. I've tried a couple of the Xagyl cards and can't say that I had much if any improvement over the raw XR9. What I haven't tried is switching a whole tower over the the Xagyl cards to see how they perform compared to just XR9's.

Would you say the performance is comparable between the two cards?

Yeah, pretty much. The noise floor readings are always so good on the Xagyl card that I simply don't believe they are accurate. The last tower I tried one in it seemed to help some people slightly and hurt others slightly. Overall, I would say the performance was about the same. In the end, the XR9 is cheaper for me and I have a few hundred of them in service and understand how they work, so I didn't see much of a compelling reason to use the Xagyl card. I have heard other people say the Xagyl card did improve their AP's and clients drastically, I just haven't seen it with my own eyes yet.

Ill chime in having all this hardware in the field.

XR9 vx xagyl.- Mostly the same, but in noisy areas, the xagyl card has won. (I put it in APs, but for CPEs the difference is marginal unless you want the extra dbs)

We use 13db laird yagis, outperformed the rootenna/arc panels for us but susceptible to all the things yagis are known for of course.

Performance between UBNT and mirkotik 900mhz, the ubnt wins for raw throughput all things being equal, I recently did 17 megs in 8mhz to a customer, where 10-12 megs (nstreme) was max with mikrotik in 10mhz. The extra chain of course giving the extra throughput.

We dont deply new mikrotik 900mhz APs, but we still put it up for customers in legacy deployments.

Got burned by the SR9/XR9 disaster a few years ago. I've had good luck with the UB900 stuff. Mostly have switched away from Moto 900 to UB 900. Good performance and stable connections - what kind of problems did you see?

Signal variation moving from connectorized Canopy to UBNT. Same antennas. Very low throughput in noisy areas where Canopy just worked, etc.

Thats why I quit with the UBNT stuff originally. I got burned by the SR9's.

I have a major WISP competitor here that swamps the 900 spectrum (maybe as a way of preventing competition ?) with Canopy 900. In some places the noise floor went from -95, 8 years ago to -66. (We were Synced to them on Moto with GPS). So my 900 shots were -62 or better to stay connected anyways, otherwise we had previously had to move them to a LOS repeater.... I didn't see a lot of change moving the Moto900 to UB900. In some cases we needed to install a tripod and mast, but most were a straight swap.

Same here. Just replace "where Canopy just worked" with "where XR9's just worked".

I swapped out one small tower (AP and 3 clients) from XR9's to 900M's. I was going from a 12db rootenna to an 8 or 9db (if I remember correctly) internal antenna, yet most of my signal drops were in the range of 10 db or so, rather than the 3 or 4 I was expecting. I had issues with noise too. An xr9 link that was perfectly stable at -78 absolutely sucked with the 900M's and seemed to need to be about -70 or better to be as stable. I also saw absolutely zero performance benefit. Still maxed out at around the same speed as my XR9 systems no matter what I tried. Add to that the fact that I have too many towers to do dual pol in many locations (due to self interference between towers) and I really didn't end up with a single advantage to the 900M's other than price.

That may be my saving grace I guess - most of my 900's are like above -65 or better.

And I'm on the opposite side of the equation where 75% of my 900's are -65 or worse, with most being somewhere in the -70's.

Is that all 2x2?

How is 5.5.2 working ? I'm always paranoid

5.5.2 is the best performing most stable firmware to date for me and others also say the same.

Above the first image everyone is 2x2 but the 2nd is mostly 1x1

We are getting very good results from the new 16dbi+rocket 900 as a CPE but slightly on the pricey side.

Thanks.
Watch out for icing on the Yagis. Not good.

Wet snow is no hell either