Has anyone successfully used a multiplexer in conjunction with gps sync to successfully use 2 APs on 1 sector.
Got a spot im out of wind load, and out of capacity. 1 more epmp ap 1 way would be great. I know its a 3 to 4 db hit to do this, but well worth it to me.
For 5Ghz?...do you have any model specs you could share?
No exp with multiplexers but would one of the KPP two antennas in one do the trick ? Looking at it I'm thinking it is almost twice as big a single sector so probably won't help your wind load problem.
https://www.kpperformance.com/two-5ghz-hv-antennas-in-one-radome
1 Like
We do this on the Motorola side of the house for 2-way radio systems. But..., we still need two separate antennas. One for transmit and one for receive. We transmit up to 6 channels into one antenna using a rf combiner system. We then receive the 6 receive frequencies of those channels using an rf multicoulper using a second antenna.
Unless there is some new technology on the markert I am unaware of, there is no way to do what you want without it being a full duplex system (separate transmit and receive frequencies, as in 2 way radio systems). The 11Ghz Ubiquiti radios are full duplex (2 frequencies) and use an rf duplexer to split the tx and rx. The ePMP are half duplex systems using one frequency. Just to combine the transmit you would need to add the components in pic below (each tuned for frequencies you plan to use) for each port of the antenna. Basically, you would repeat attached diagram twice, one for the V port and one for the H port.
This will -ONLY- combine transmit power. You will have little to no receive. A single stage rf isolator reduces any rf power coming from the antenna into an attached dummy load (the two cylinders underneath isolators in attached pic) by a factor of 20 to 30 db. This is to keep the transmitters from feeding into each other and causing horrible IMD's to be transmitted. They act as a one way valve, allow rf to the antenna, but not from the antenna. If you had a -60 rx signal hitting the sector, you would only have a -80 to -95db (-60 minus the filter insertion loss, connectros, and 20-30db of loss from the isolator) signal at radio coax input port.
You will also spend a lot of money just to be able to combine the transmit.
I think brubble1 has the best suggestion.
I also threw in some links for anyone who is interested in rf combining and receive multicoupling.
In theory this can be done, but it would require a very extensive isolator system to be installed behind the antenna. so not really practical. the epmps use a mimo system, so you would need two of everything up there. Yet more wind load!
better solution is is to use a KP dual antenna or place a pair of narrow beam antennas side by side.
even better would be to re-enforce the tower to handle the new wind loading.
@Douglas Generous wrote:In theory this can be done, but it would require a very extensive isolator system to be installed behind the antenna.
I am curious..., theoretically, how and with what equipment would you accomplish this?
I don't want to HiJack this thread... so Chris if you think this is HiJacking, tell me and I'll open a new thread, OK? :) But I have a multiplexer question so maybe it's related enough to be on topic?
Anyways - I have a bunch of ePMP2000 AP's on the shelf here and my plan is still to deploy them in places which don't need or justify the e3K. However, I'll have to buy sectors for them, and I was wondering about the idea of buying the Cambium 4x4 sector, and then using a multiplexer to connect both of the sector's V connections combined to the e2K's V, and both of the sector's H connectors combined to the e2K's H.
That way, I'm not buying older sectors, and those locations are e3K ready if they ever need that. How do you multiplexing experts think that would work out?
You could always try the UBNT multiplexer It's designed for the AF5x and I suspect the sync timing on ePMP isn't tight enough for the two radios to step all over each other, but it might work.
Mr. Turner, you and me had the same thought.
I just got 1 in to try out. Got 2 epmp2000s to use as well.
I’ll have to wait a bit though, its perfect weather for drilling season to start so for now we are focusing on putting fiber in. Got 16 miles to go and some new equipment that needs broken in!
Todd, hijack away.
You can use the 4 port sector and just leave 1 h and 1 v port capped for now. Drop in the new ap later. I haven’t tried, but you may be able to put 2 epmp 2ks on that 1 ap and have double the capacity if you have enough spectrum now and drop a 3k on there another day. I’m itching for some 3k action. Myself, but i need dfs unfortunately.
1 Like
@Chris_Bay wrote:
You can use the 4 port sector and just leave 1 h and 1 v port capped for now.
Oh, I thought that if I did that, that would essentially just give me a 45 degree antenna? I thought the 4x4 sectors were essentially two 45 degree 2x2 antennas mounted together. That's why I was wondering about using a combiner/multiplexer to combine both V's and combine both H's
@Chris_Bay wrote:
you may be able to put 2 epmp 2ks on that 1 ap and have double the capacity if you have enough spectrum now
Yes, I was thinking of that also. If those antennas are really two side by side 2x2 45 degree sectors in one housing, then I was thinking that 3 of those 90 degree 4x4's, would really be 6 - 45 degree 2x2's - so with 6 ePMP2000's doing ABCABC reuse, that might be awesome too.
@Chris_Bay wrote:
I'm itching for some 3k action. Myself, but i need dfs unfortunately.
SONY
(Soon, Only Not Yet)
On a 4x4 sector like this or those designed for Mimosa radios they are typically two arrays of equal horizontal beamwidth, probably stacked on top of each other. If in doubt, check the radiation patterns in the spec sheet.
@Jacob Turner wrote:You could always try the UBNT multiplexer It's designed for the AF5x and I suspect the sync timing on ePMP isn't tight enough for the two radios to step all over each other, but it might work.
I asked earlier if Chris had a model/specs in mind, because I thought this may be what Chris was talking about.
I do not think these will work with the Cambium radios. These seem to be mechanical in nature with RF triggering to engage/disengage RX/TX paths. I feel these will not work with Cambium radios because for these to function properly with the airFiber radios, in the airFiber radios GUI, you must check the NxN enable box. I suspect this feature adds some delay to actual data being sent so RF paths in multiplexer can switch before data is modulated onto the RF carrier. Cambium radios have no such box to check.
I personally would also avoid these because the so called data sheet, gives absolutely no data. No insertion loss (a tester claimed 3-4db of insertion loss), return loss, and most importantly no TX-TX isolation. This is very important. If the transmitters are not isolated well, you are going to have horrible IMD's that can weak havoc within your own network (3rd order IMD's). This could possibly put you in the hot seat with the FCC for an OBE (Out of Band Emissions) infraction. I do not know if these would or not and do not want to find out.
With this being said, I am still highly interested in results Chris will obtain.
*Edit*
So.... it was bugging me that switching speeds would be impossible, I decided to look into this deeper. This multiplexer is said to be built on the Wilkinson Power Splitter design. Impractical at the sub 1Ghz freqs I work with but not at 5Ghz. Due to design it also negates alot of concerns I had above. I am even more interested to see how these will work with Cambium. I probably should have taken a real look at this before knocking it. I just have a problem trusting products that give no real data.
@ninedd wrote:
@Chris_Bay wrote:
You can use the 4 port sector and just leave 1 h and 1 v port capped for now.Oh, I thought that if I did that, that would essentially just give me a 45 degree antenna? I thought the 4x4 sectors were essentially two 45 degree 2x2 antennas mounted together. That's why I was wondering about using a combiner/multiplexer to combine both V's and combine both H's.
The 3000 sector is 2 90 degree sectors in one shell. If not doing MU-MIMO the 2 sectors combine in phase at the SM for up to a 3db increase in RSSI. If you do not use one of the antennas/sectors (1 H and 1 V), it is best to install a small dummy load to help keep noise from being generated or re-radiated. It will also help to stablize the radiation pattern.