Just completed an install of a Force180 to a ePMP 1000 AP. This is the 11th SM to attach to this tower. Close range (.09 miles) and perfect line of sight with no obstructions. I can't get the SM to break an MCS of 7. RSSI is -65 (not awesome but not terrible) and SNR is 31. MCS is 7/5 while every other SM on the AP is 15/12 or 15/13 (we limit at 13 upload). The new SM is the bottom one:
AP is running on a totally clean DFS channel. No interference on the AP or any SM attached to it:
The only thing that has me a little concerned is while installing this Force 180, it slid down the shingle roof on the face of the radio for about 10 feet before coming to a rest in the gutter (didn't fall). It scratched the face of the radio a little. Certainly didn't think it looked bad enough to cause a reception issue.
The RSSI and SNR should be giving it a 14 or 15 MCS, not 7. Speeds on the radio are off as well - they match what I would expect with a 7 MCS on a 40 MHz channel (~ 50 Mbps). We need to give this customer 75 Mbps or better. Any thoughts before I head out and replace the Force 180??
Thank you for provided screenshot. From statistics we can see radio operates on high MCSs in Uplink, however in DL mostly it operates on MCS7. One of the most possible reasons could be interference in Downlink. So it would be nice to get spectrum captures from SM. Also make sure Force180 mounted horizontally to avoid polarization issues.
From a pure RF stance, it seems that one of the receive chains is dead. You have almost 94% of your receive data at MCS7 which is the single chain version of MCS15 (dual chain). You also have practically no data at dual chain MCS levels. With that being said, I would not think inerference would be the case or you would not have such a high single chain MCS.
I would go to eAlign and check your Ch0 and Ch1 levels. You may find one of those chains are not registering a receive level, which would indicate a receive problem in that chain. This could be from a faulty pin switching diode, faulty coupling capacitor, or even from the slight fall the unit took (possibly broke a solder connection).
There could be other reasons, but being an RF man, and with the data presented, a dead receive chain is my first thought.
The two chains don't line up very well. I've looked at all the other SMs on this AP and the two chains are within a dBm of each other on all the rest - except this one. It is dead level (we actually use a level when we mount them) and it has a clear shot.
I can't run a spectrum analysis once it's on a customers house so late at night, I am going to change frequencies and see what impact this has. I don't think it will have much since I am in the middle of nowhere and using DFS - but I'm willing to try before I go swap hardware.
I still would not rule out a receive issue. I have seen some crazy things in the world of RF. If you are LOS and aimed correctly into the AP like you say you are, you should not have a 10dbm difference in the RF chains as you do with this SM. Especially since all other SM's Ch0 and Ch1 are within 1dbm. That 10dbm difference could be from a faulty component that is not only knocking your receive level down by 10db in that chain, but could also be distorting the waveform so much that no useful information can be decoded from it, causing SM to revert to single chain MCS 7.
Once again, this is all based on years of RF experience working with radio communication systems. Could be other reasons for your issue.
Like you, I would try any and all other means to find problem besides getting on a roof to trade units if I could.
I got a chance to change frequencies a couple times. No change in the dispartity between chain numbers. Got a little MCS improvement up in UN-I-III but that is probalby due to the higher power output. So, last step is to do another site visit and replace the radio. I know it is installed perfecly horizontal and aimed well so I'll see if new hardware fixes this...
if a new radio doesn't solve it, try moving it up or down a couple of feet. really close in and at a very sharp angel we've noticed some odd behavior with 5ghz stuff, not just cambium. the sectors don't express the most consistent pattern outside of the main antenna lobe, a small movement can correct that sometimes.
