Hello,
For those who have installed/used the 450i 900 platform coming from 900 FSK what type of range / signal performance have you seen? We tried our first install attempt for a customer on a 180 degree FSK 900 mhz sector to the new 450i with 60 degree cambium sector antenna. Saw ~ 14 dB lower signal (17 yagi on 900 FSK vs the 12 dual slant on 900i sm) 1x Mimo-A at ~ 3.7 mile range.
What type of signal level performance have people seen vs existing 900 FSK deployment?
Tim
That's interesting.
We haven't done a straight comparison between 900 100s and 450s yet, but I'm seeing much better speeds through the 450s, even with the typically low signal strengths we get in our area. The 450 seems to handle weaker signals much much better than the 100s do.
We had somebody else in the forum post a month or two back about being able to get working connections with a 900 450 at a site that wouldn't even register on a 900 100.
What kinds of link capacities are you getting for capacities with the weaker signal? Given that we're dealing with OFDM vs FSK, I'm not sure the signal strength on it's own is all that significant.
So far we only put up one customer but ran some tests at various distances from tower. There is some thought on the existing 900 fsk causing some issue with the new 900 due to how close they are but they are on different frequencies with a pretty large guard range between. I’ll update the thread with results from another swap or two in next few days.
Tim
We didn't have any 900 FSK on the tower we put the 450 900 AP on, so that could very well be it.
I did find the noise spectrum for both the AP and SM was very different with the 45 degree slanted antenna than I saw in the horizontal. There were tiny little spikes of noise that I had to change frequencies to get around. How do the spectrums look for both the AP and SMs?
@Timothy Alexander wrote:
So far we only put up one customer but ran some tests at various distances from tower. There is some thought on the existing 900 fsk causing some issue with the new 900 due to how close they are but they are on different frequencies with a pretty large guard range between. I'll update the thread with results from another swap or two in next few days.
Tim
Tim - Just want to make sure that you have seen the frame calculator tool and migration white paper for optimizing co-location of the 900 MHz PMP 100 with the PMP 450i.
These tools can be found here.
Other customers have reported better success after making sure the parameters are configured per the recommendation in these tools... let us know if this helps, or if you've already used these.
@Ted Stewart wrote:
We had somebody else in the forum post a month or two back about being able to get working connections with a 900 450 at a site that wouldn't even register on a 900 100.
I think that was possibly me, but I can't find the post... anyway, we did a direct apples to apples real world, field test comparison between PMP450i and PMP100. This was difficult to accomplish because the PMP100 radios can output up-to, I want to say 28dBm, and many people use the 17dBi yagi's with them. When we turned down the power on the PMP100 system to match the same EIRP as the PMP450i system, the PMP450i system would beat it in all of our testing in terms of the ability to connect and also the speeds at which connections could be made. In some use case scenarios, like high interference environments, even when we cheated and turned up the power on the PMP100 to the max and used 17dBi yagi's on both sides, the PMP450i would provide for a more stable connection and more consistent throughput.
Just want to note here that the EIRP maximum for FCC is 36 dBm. Therefore, when using a 17 dBi Yagi, the radio Tx Power should not exceed 19 dBm.
We have had several customers report that the 450i 900 MHz radios connect even when the PMP 100's struggle, and have proven that in our own outdoor testing. Feel free to watch the webinars here where I discuss this (I have charts and data and everything!)
We were able to switch to 5 ms frame timing and get a few more customers on this AP and are seeing better performance.
The one thing I do see so far is most customers showing 1x A on uplink/downlink if not currently being used (activly transmitting data) where as the 5.7 450 sectors we have show modulation even when not in use. Is this expected behavior?
@Eric: Thanks for pointing this out. The combination of a lower dB rating Yagi with AP @ 22dB per pole adds up to the signal difference.
Thanks,
Tim
I meant to ask about frame size, but it slipped my mind. That's definitely going to make a difference.
I would love to see higher gain antennae for the 900 450. There's a significant difference in length between our PMP100 yagis and the PMP450 yagis.
@Ted Stewart wrote:
I would love to see higher gain antennae for the 900 450. There's a significant difference in length between our PMP100 yagis and the PMP450 yagis.
I've found a couple examples and suggestions for 3rd party antennas in this post.
I don't really want a flat panel. I'd prefer to keep the wind load down.
Ubitquiti makes supposedly 16 dB gain dual yagis for their own products. The mounting and antenna connectors look pretty much the same to me. So that piques my curiosity.
We purchased 1 x 450i-900AP, and used a 13.5 external flat panel from KP Performance. We have 1 x SM 450i and same 13.5 external flat panel.
The sector is placed several kilometres away from a 300 foot tower with screaming 360 degree 900 FSK
Did a test 2 miles away from the sector pure line of sight, and keep getting SYNCING, REGISTERING, back to SYNCING, never connects.
Did a another test 4.9 miles away from the sector nLOS and got SYNCING, REGISTERING, REGISTERED, performed a ping to the sector and got up to 1000 ms response, 2Mbps through speedtest.net
For a product that should do extremely well and up 100 Mbps this is a very bad and expensive experience with the product.
Note: I tried the full 20 MHz to get my maximum performance, then 10 MHz, then 5 MHz
Anyone have a high throughput performance AP configuration file I could upload to my sector and test?
There is a lot of information left out of your post to offer any adive at this point. But what does your specrum analisis show for the noise floor? Are you running GPS and is the FSK on GPS with matched settings using the white paper?
ncsnetwork, you using a sector antenna or an SM panel antenna? I'm not finding a 13.5 dB on kp performance for a sector antenna.
I asked about this earlier, but didn't see a reply.
Are you using a 2.5 ms or 5 ms frame length? The PMP100 900 MHz uses 5 ms, so if you don't match it you're going to get a lot of self-interference.
Running 14.2.1 firmware on a 450i 900mhz AP and 10mhz channels, 2.5ms frames and i was seeing some link tests at 3-4 miles NLOS at 40mbps down and 10mbps up. This product is amazing! Even when the signals aren't good its still getting 10mbps download same setup. I don't have any FSK on this tower to compare too but i am going to test another tower later this year that has FSK and an omni for direct compaison.
If you using that on your ap thats part of the problem. Below is a link to proper sector antenna
I had a weight and space restriction and I only need 30 degree of beamwidth.
We switched the setting to 5ms, and enabled MODE1 for coexisting with old Canopy FSK900 system.
Link is stable, performance not great, 2 kms with near line of sight, 10 down and 5 up, but latency of 500 ms
Because the number of slots on the 450s is so much higher than on the 100s, the timing is quite a bit different, so using the same settings as the PMP100s will still cause interference. You can either use the 900 colocation tool or the FSK and OFDM frame calculators in the PMP100 to correct that.
I've done a lot of work playing with timing to reduce interference and improve throughput, so send me a message if you'd like a hand with that.