Let's talk about 2+0 configuration here and clarify a few concept and restrictions you may face when you planning a 2+0 link.
People will like to use 2+0 configuration to achieve high capacity. In addition, it also provide some level of redundancy. if one of the link fail, all traffic will be routed through the other link.
Normally we talk about three configurations:
You can achieve a 2+0 configuration with:
- 2 x PTP820S
- 1 x PTP820C with 2nd core activated
- 1 x PTP820G dual modem with 2nd modem activated
If the 2+0 link has seperate antenna, then there is no restrictions.
But when you plan a 2+0 configuration share a common antenna, both channel must be covered by the same sub-band radio.
Let's take an example - 2+0 11 GHz with 40 MHz channel share a common antenna using PTP820S
Following configuration is supported as they are covered by same sub-band (Ch1w6)
Following configuration is not supported as they can NOT covered by same sub-band (Ch1w6 vs. Ch7w13)
Commonly in a 2+0 configuration, both channel are using the same channel size. In the cases you need different channel size, it is ok as long as both channel also covered by same sub-band. For example, you could have one channel using 40 MHz channel and the other use 30 MHz channel, itis supported.
The only exception is 80MHz. in a 2+0 configuratoin, if you have one channel using 80MHz, the other channel must be 80 Mhz as well.
I'm wondering what is the reason for the requirement to use the same sub-band radios when using a single antenna? We recently installed a new 2+0 link with PTP820S radios. We needed to use a single antenna due to space on the tower. However, the only two available channels were in separate sub-bands. These two channels are 80 MHz, on opposite polarities in the 11 GHz band. We installed the radios and antennas, never realizing that this was not recommended. Our vendor never mentioned that this would be a problem (we sent them the frequencies prior to ordering the equipment), and I know that they worked closely with Cambium before they ever sent us a BOM. I just happened to stumble upon this post tonight, and realized our configuration is not in compliance. We have traffic on it, but not much at this point. On one side, we have a 2 foot dish with an OMT combining the two radios on opposite polarities. On the other side, we have an 8 foot dual-pol dish with the radios on remote mounts and flexible waveguides connecting the radios to the dish. At this point, the link is operating on both channels at full signal, and modulation does not appear to be suffering.
So are the implications here? What sort of problems is this going to create?
The rule we put together is guidance to provide guaranteed perfomrance in 2+0. The reason is when both channel fall to same sub-band, the filter at depliexer provide further help on radio MSE (Mean Square Error Rate).
In the cases when you don't have the radio covers by same radio, the 2+0 may still work, but we can't guaranteed in each case it will work.
It is good to know your links works, can you send me a screen short for MSE stats?
Here are the screenshots from all four radios comprising this 2+0. Please note that the time/date on the radios is not synchronized with any NTP server, and the dates are currently all different. However, these screenshots were all taken within about 10 minutes of each other. Please let me know your thoughts based on this information.
I'm new to the PTP820s, and am configuring a link for 2+0 ACAP using the OMT and 2 radios on each side. Do the radios get setup as though they are 2 seperate 1+0 links? I assume that that need to setup LAG as well?
What would you like to do?
Have you read this post?
You can use external switches to do the traffic combining, you can use internal LAG, or you can use MC-ABC.
Each of these options has tradeoffs:
MC-ABC combines the traffic at Layer 1 and is the most efficient way to use the channels. (Depending upon the configuration and the traffic, LAG may allow one link to be fully utilized while the other link is under utilized. LAG doesn't necessarily guarantee even traffic distribution.)
There's no additional cost with internal LAG. External LAG requires two drop cables per end, one for each link.
I hope this helps!