Hi All,

I am wondering if there are any considerations to be taken into account if I were to mount 4 or 5 routerboards in the same outdoor enclosure (IP67 rated - metal or otherwise).

This is mostly to reduce space taken up on the mast by the regular enclosures and also to reduce patch cords going from one RB unit to the other.

Two of the RB's shall have 2xR5H's each (two backhaul units) and the other two shall probably have 3 x R52H's each (two 360 Deg AP's)
.
The third is mostly to be used for monitoring and providing external services to the same building.

I intend to utilise the power jack on all boards so as to be able to control power remotely.

Please do advise on this especially if you have done this already.

Mark.

Radio interference is something to keep in mind.

In some situations, heat is also a consideration.

Thank you for your feedback.

I thought of the interference between the miniPCI cards but then the RB600A can actually have eight on board with the daughter card and this by design.

What about the heat considerations? If the enclosure if more that 4 x the volume of a 'single' enclosure, would this not be enough?

Rgds,
Mark

600a with a daughter card can hold eight cards, but you don't want to.

I've got a 600 with 4 R52H's, 1 5GHz ptp in, 1 5GHz ptp out, 1 5GHz AP and 1 2.4GHz ap.

It is a backbone unit with backup AP's. Even with ptp1 on 5825 and ptp2 on 5765, ccq still drops signifigantly.

I have to run one ptp link on 5.3GHz band to keep ccq up.

One of my other towers has two 5GHz ptp links and doesn't have as much of a problem, but under a really good load, it is noticeable.

If you put multiple boards in a single case, I recommend using indoor cases inside an outdoor case to try to minimize self interference.

In response to a previous post regarding interference....

Multiple radios in an enclosure works. Use quality components and you'll be fine. Running 6 5ghz radios with 2 on the same frequency hasn't posed a problem yet (in 20 vertical feet.)

Good quality jumpers and feed line, antenna spacing, and antenna choice is critical. Using run of the mill pigtails is the first mistake most people make, followed by poor feed line choice.

Whats some good pigtails? I'd rather have my stuff in a single outdoor box, single board, but its become a problem recently.

600a with a daughter card can hold eight cards, but you don't want to.

I've got a 600 with 4 R52H's, 1 5GHz ptp in, 1 5GHz ptp out, 1 5GHz AP and 1 2.4GHz ap.

It is a backbone unit with backup AP's. Even with ptp1 on 5825 and ptp2 on 5765, ccq still drops signifigantly.

I have to run one ptp link on 5.3GHz band to keep ccq up.

One of my other towers has two 5GHz ptp links and doesn't have as much of a problem, but under a really good load, it is noticeable.

If you put multiple boards in a single case, I recommend using indoor cases inside an outdoor case to try to minimize self interference.

I have an RB433 with an R5H running at 5805 and an R52 at 5320 and the interference is terrible. I thought with that kind of separation I wouldn't have a problem but that is not the case.

Whats some good pigtails? I'd rather have my stuff in a single outdoor box, single board, but its become a problem recently.

We have ours manufactured per our specs. We don't use commonly available parts, as those tested have failed our tests.

We use an RG178 type cable, double braid over foil, that results in a -105dbc or better leakage figure per cable.

600a with a daughter card can hold eight cards, but you don't want to.

I've got a 600 with 4 R52H's, 1 5GHz ptp in, 1 5GHz ptp out, 1 5GHz AP and 1 2.4GHz ap.

It is a backbone unit with backup AP's. Even with ptp1 on 5825 and ptp2 on 5765, ccq still drops signifigantly.

I have to run one ptp link on 5.3GHz band to keep ccq up.

One of my other towers has two 5GHz ptp links and doesn't have as much of a problem, but under a really good load, it is noticeable.

If you put multiple boards in a single case, I recommend using indoor cases inside an outdoor case to try to minimize self interference.

I have an RB433 with an R5H running at 5805 and an R52 at 5320 and the interference is terrible. I thought with that kind of separation I wouldn't have a problem but that is not the case.

I will only say that I do not use any "R" cards due to issues experienced with all of them.

Running 6 5ghz radios with 2 on the same frequency hasn't posed a problem yet (in 20 vertical feet.)

You have two 5 GHz radios on the same frequency and in the same enclosure?!

Yes. We do it frequently.

Hi

Careful mate with multiple radios in one box.. I have had bad experienes with one particular brand.

Also watch out for poor vswr.

Poor vswr will force a coax that normally wouldnt normally radiate, to radiate from its shield. Some radio cards have poor screening around their RF stages and this unwanted radiation can get in to the RF circuits and cause mayhem.

Some antennas may appear to have a flat and useable vswr curve across the bands, but I see some specs of vswr at 2:1 or worse at band edges.
Add that to imperfect installations, ie metal objects or other resonant antennas in the same region and the vswr could be worse.

I agree with a former post regarding higher quality pigtails.

And am astonished too when you say that you have 2 radio cards in same enclosure on same frequency.

Yes. We do it frequently.

And you aren't bothered by the 50%, or worse, performance hit but the huge amount of interference?

Yes. We do it frequently.

And you aren't bothered by the 50%, or worse, performance hit but the huge amount of interference?

There are no issues. Proper system design is the difference between having a solution that performs well and one that does not. I have attached a shot of a unit with 2 radios running on 5300. The customers are limited to 10m u/d.

It is very frustrating to have folks continually state that you can't do something when you can. RF is a science. There are far too many people who cobble crap together and it 'works', yet they don't know they are causing others harm, or that their 'working' system is only marginally functioning at best.

Of course we don't run multiple radios on the same freq if we don't have too, but there are cases where you must, and it is doable. Reference my previous comments regarding using good components.

ne0031, what kind of antennas are these? something highly directional?

Whats some good pigtails? I'd rather have my stuff in a single outdoor box, single board, but its become a problem recently.

We have ours manufactured per our specs. We don't use commonly available parts, as those tested have failed our tests.

We use an RG178 type cable, double braid over foil, that results in a -105dbc or better leakage figure per cable.

Care to share your supplier? or do you make them yourselves?

ne0031, what kind of antennas are these? something highly directional?

The link quoted uses MT-485002 panels. Even though they are 9deg panels, these are mounted about 5 feet from each other approx 75deg off axis.

5 feet is the minimum separation we will use when freqs are in close proximity to each other.

Whats some good pigtails? I'd rather have my stuff in a single outdoor box, single board, but its become a problem recently.

We have ours manufactured per our specs. We don't use commonly available parts, as those tested have failed our tests.

We use an RG178 type cable, double braid over foil, that results in a -105dbc or better leakage figure per cable.

Care to share your supplier? or do you make them yourselves?

We make them ourselves for this application. Otherwise we use Laird, which are the next best we have located. They run -100dbc normally.

Thanks for the info. Just checking, are the Laird pigtails any different than PacWireless?

I did notice your signal levels were ridiculously good. That also has a part to play in the self interference issue.

My testing here on the ground, -50s, no problems being on adjacent channels. Get'er up in the air, get a -77 instead of -52... whole different situation.

What ne0031 said...
I agree...

We have lots of RB600's in the field - fully loaded as well as RB800's FULLY LOADED..!!! - ie: with daughterboards and all slots occupied.

Use good components and you'll be OK.

Tanker//

I forgot to mention:-

- we use RG-316 DOUBLE Braided pigtails to drop cross-over to minimums. A tad pricey at $8.50 each - but well worth it.

Got a source for those?

I just finished up a ptp AP for to feed a three new towers.

With three R5Hn, I had poor ccq with the card in the middle, regardless of the channels.

I swapped the middle card for an R52H to run as a 2GHz ptp instead, high 90 ccq on all three radios, even on adjacent channels.

I'd like to try out those pigtails, but I think in this case, the actual card is just too close.

btw, using Laird MMCX to NF bulkhead now, definitely less noise than the ufl.

Thanks for the info. Just checking, are the Laird pigtails any different than PacWireless?

I did notice your signal levels were ridiculously good. That also has a part to play in the self interference issue.

My testing here on the ground, -50s, no problems being on adjacent channels. Get'er up in the air, get a -77 instead of -52... whole different situation.

Without knowing which ones, I can't say. We get varying product for the same part number, which is one of the reasons that high density deployments transition to in house components.

I can only assume that 'getting it in the air' means you have significant feed line? Otherwise you shouldn't see any difference in signal levels.

Got a source for those?

I just finished up a ptp AP for to feed a three new towers.

With three R5Hn, I had poor ccq with the card in the middle, regardless of the channels.

I swapped the middle card for an R52H to run as a 2GHz ptp instead, high 90 ccq on all three radios, even on adjacent channels.

I'd like to try out those pigtails, but I think in this case, the actual card is just too close.

btw, using Laird MMCX to NF bulkhead now, definitely less noise than the ufl.

I won't bash the "R" cards. I will say that we don't use them - for a variety of reasons.

Without knowing which ones, I can't say. We get varying product for the same part number, which is one of the reasons that high density deployments transition to in house components.

I can only assume that 'getting it in the air' means you have significant feed line? Otherwise you shouldn't see any difference in signal levels.

Thanks for the info on the varying quality of Laird pigtails.

Max cable for me is 6ft. I was just commenting on difference between testing equipment and using it on location, free space loss.

-50s signal is pretty hard to come by on long shots, -70s isn't as friendly towards multiple radios on a single enclosure, at least with the parts I have available.