In this test we did a 40 km link using NV2, we tested all he does better on top of Nstreme in terms of latency and processing continues over TCP using 100% CPU, no ping traffic ranges from 5ms to 24ms, we also realize he has remained online with no fall maximum of 1 day by setting the same power. As you leave the bridge even in common is swinging up to 5 days without anything, we Hw. Retries = 10, NV2 = 2 TDMA Period Size, Cell Radius 40Km, XR5 card, TX-POWER = 26dB.
The following additional attachments
Our testing was done using only one card XR5, Pluton Antenna 5.8 Ghz 29 Dbi produced in Brazil.
For not being 802.11N its Amazing!!!
I would suggest to upgrade this link to N card and try out as you could get even higher data rates even with one chain.
And then you would use also our new test release RouterOS v5.9 which improves the performance even more.
card rate at 26dbm? won’t that kill the card?
XR5 defaults to 28dBm.
TX SPECIFICATIONS
54Mbps-23 dBm
http://ubnt.com/downloads/xr5_datasheet.pdf
Card Rates mode will lower the power accordingly based on data rates.
@54Mbit most likely it will be 21dbm if the power is set to 26dbm
As friends and talk suggest we use this test to publicize the system funiona, as this link does not require more than 8Mb, used it as a test to get an idea, of course I want to use N cards and see how it comes that and link distance, more like Mum saw in Sao Paulo Brazil who speak more than 30Km NV2 did not run and was very poor, I decided to take the test using the worst case, a single card set optimally and do the tests, we use 26dB I think the test was a success when it comes to hardware used and the form used.
As mentioned we use version 4.17 because it ran Mpls, and still not test it on as versions 5.8 and 5.9
I see what you’re talking about now.
I don’t know. Might kill the card, might not. Those XR cards are pretty tough. The amp is still running at nearly a quarter capacity even tho it is being over driven at 54mb.
In my experience, overdriving the XR5 though the software kills the CCQ more than the card.
Also, no discussion of XR5 is complete without mentioning how Mikrotik screws around with the offsets so that I have no confidence that the card is even being run optimally anyway.
BTW, I have XR5 42km that gets similar speeds to OP using original Nstreme protocol. My link is PtP and I prefer nstreme for it’s more stable ping times.
I can understand your position more my test is not to sell anything or prove anything but pass as being traffic on Mikrotik protocols using this card, then questioning me the way you did is irrelevant, I mean that many squeaky-band links in doing so wrong, i just using the link to sample test of how he behaved, just using a card. And the CPU consumption.