Hi all, we put up a 3.1 mile link and had great results, 1097mbps DL and 1091 UL, -65/-65 signal, DS 8 DL DS 7 UL, SNR 31/31. No traffic was flowing through the link at time of testing. Now have traffic going through the link and the modulation is fluctuating alot…
If you perform bandwidth testing across it (like via two Mikrotik’s), are the speeds consistent? What do each sides radio’s MCS %'s look like?
What channel size are you running? Assuming it is not an interference issue. We have found the internal DS reporting seems a bit flakey when running a 160 Mhz channel. Like Eric suggest, we test across the link Mikrotik to Mikrotik and the results have been consistent, regardless of what the DS is reporting and in fact seems to “correct” itself when we run a router to router test. Seems to be more accurate on lower bandwidth setting as we do not experience that anomaly on 80 and 40 Mhz PTP links… Just our observation and to date no indication throughput is actually being affected. We consistently see the real throughput we are expecting.
As evidence of the “flakiness” this is the internal test on a 160 Mhz link. Notice it shows the uplink as SS0 yet an uplink speek of 2576 Mbps. The downlink side is accurate. The slave side shows accurate DS levels. When I change the link to 80 Mhz, it reports as expected. It does not appear to affect the throughput as I consistency get 980 meg with a Mikrotik to Mikrotik test, limited by the 1 gig ethernet connection. Not that I wouldn’t love to be get 2.5 gig on my uplink, LOL.
PS upgrading to 5.9.1 resolved the reporting error.
I was surprised to see that they made a premium 2gbit radio, with a 1gbit Ethernet. I know it has SPF+ options, but 2.5g was right there at their fingertips. 
thank you for your replies, here are some screen shots. I have it set to 160mhz channel…
Here is throughput chart, the starting on the left is cambium link test done on the slave, then cambium link test done on the master, and the last 2 are tik to tik bandwidth test.
Your MTik test is also set to test simultaneously bidirectional. Whereas the Cambium test, tests download and upload separately.
If you do the MTik test again, but specify a download test, followed by an upload test, that’ll give you a more similar look at what you’re capable of.
Thank you for your reply and I did as you instructed and the MTik to MTik was more reflective of the Cambium link test.
Yes, below is the same test you ran on a 160 Mhz channel, Master to Slave and Slave to Master and then Mikrotik to Mikrotik. Mikrotik maxes out at 980 meg as it is just a gigabit ethernet connection to the radios.
Nice! What distance is that link? What did you use for antennas?
Short distance, about 1.3 miles. Using 4600C with RFE ultrawideband dishes. We use it to bring in a 2nd geographically separated redundant fiber feed into our primary NOC/network. A Tachyon 60 Ghz is our primary and this is the backup Both have been very solid but the 60 will fade in heavy rain. We have quite a few 6 Ghz PTP links using both 4600C with RFE dishes and 4625’s. Our shortest link is about 1/3 of a mile. We replaced 60 Ghz with a pair of 4625’s on a 40 Mhz channel getting 420+ meg both ways. Plenty of bandwidth and we do not have to worry about fade. They have been working great. Highly recommend following Cambiums advice to set the slave side to only scan the actual frequency. It will reduce the reconnect time from minutes to about 35 seconds to reconnect.
