Hi everyone,
One of our team members has a node that's about 11 miles from our other nodes (2.4 GHz band using 5 MHz width). He sees the node at 100% LQ, but 0% NLQ. Is there anything that he should do to improve that? Or is it just the fact that he's too far out? We have another member who's going to put up a second node in between his and the main ones as a relay, but that won't be until next week. The nodes are 263' on a tower and they're TP-Link CPE210's. His node is 50' up on a tower, and it's a Ubiquit Rocket M2 with Yagi antennas pointed at the other nodes (one vertical and one horizontal).
Thanks and have a great day. :)
Patrick.
One of our team members has a node that's about 11 miles from our other nodes (2.4 GHz band using 5 MHz width). He sees the node at 100% LQ, but 0% NLQ. Is there anything that he should do to improve that? Or is it just the fact that he's too far out? We have another member who's going to put up a second node in between his and the main ones as a relay, but that won't be until next week. The nodes are 263' on a tower and they're TP-Link CPE210's. His node is 50' up on a tower, and it's a Ubiquit Rocket M2 with Yagi antennas pointed at the other nodes (one vertical and one horizontal).
Thanks and have a great day. :)
Patrick.
https://sites.google.com/site/orangecountymeshorganization/lqnlqandetx#:...(LQ)%20is%20the,node%20displaying%20the%20LQ%20value.&text=The%20percentage%20of%20packets%20that%20it%20receives%20are,NLQ%20(Neighbor%20LInk%20Quality).
Hi, Patrick:
The SNR at each end would be meaningful to me.
The LQ/NLQ does not tell much of the situation.
11 miles is not too far as some links are up to 40 miles on a quiet channel.
You did not mention the channel, but I assume '-2'.
You mention 3 nodes, (2) CPE210 and (1) Rocket M2, but I am awash about
who is doing what, with which, and to whom!
Is there a map?
What is the path profile, 'twixt and 'tween those 3 nodes?
0% LQ or NLQ indicates a serious problem.
As a reference, here are the SNRs on a 13.5 mile link, ends ~ 120-130' elevation, sharing a quiet part 15 channel:
NC8Q-MVHS-DARA
Location: 39.654 -84.115
Channel 140
Bandwidth 20 Mhz
Signal/Noise/Ratio -60 / -95 / 35 dB
NC8Q-DARA-MVHS
Location: 39.847597 -84.099131
Channel 140
Bandwidth 20 Mhz
Signal/Noise/Ratio -62 / -95 / 33 dB
LQ/NLQ 100% 100%
TxMbps 100+ both ways
A pair of MIMO nodes at 5 MHz bandwidth has maximum 32.5 TxMbps on a quiet channel.
At 10 MHz, 65, and more FEC to reduce retries.
Chuck
I don't have a map of the nodes per se, but I can give you a link to the area that we're located in (the nodes aren't Internet-connected as of yet). The terrain between them is hilly and the single node is at a higher elevation than the three nodes on the tower.
On the tower, we have three nodes (two of which are working). Node 1 is on Channel -2 at 5 MHz and node 2 is on Channel -1 at 5 MHz. The node that is trying to connect is on Channel -1 at 5 MHz and is pointed at Node 2 on the tower. I don't have the SNR information as of right now, but I will get it and add that to the post. I'm adding three pictures that show the map of the area with expected coverage (in red) from HeyWhatsThat, the same map without coverage but with contours, and their prediction for the coverage. It's hard to tell, but his house is in a spotty red area.
Here's a link to the map at HeyWhatsThat http://www.heywhatsthat.com/?view=XW262LEJ
Thank you for your help, and when I get the SNR information, I'll update the post.
Patrick.
Hi, Patrick:
I noticed that 'nodes' was plural.
Are 'both nodes' the same direction?
I hope you don't have one yagi pointed one way and the other yagi pointed another way.
:-|
At 11 miles, you will likely need the gain of a dish at one or both ends.
I ran some test on our 13.5 mile link and determined that with
28 dBm and 40 dBi (gain from TX and RX antennas) there would be a SNR of 20 dB.
Add up each end of your link:
(node-A's dBm+dBi+node-B's dBi)=?
Node-B's dBm+dBi+node-A's dBi)=?
I am guessing that you may need a value of 65 or more.
At any bandwidth, 1/2 of your signal on channel '-1' is in the bandwidth of channel '1' at 20 MHz bandwidth.
:-|
Chuck
Hi, Patrick:
18 dB of SNR is sufficient and could be the reason Scott is getting 100% LQ.
"So his is essentially pointing at all three of them."
But they are not all pointing back. ;-)
I am making assumptions, so let me know what errors I introduce.
The Rocket M2 is 28 dBm and the CPE210 is 27 dBm so I assume that the reverse path should be 19 dB SNR.
(cpe210 27dBm+9dBi+Yagi+1 == RocketM2 28dBm+Yagi+CPE201's 9dBi antenna)
Is it known that the MCC-North node is working?
- Is there any other node linking with MCC-North with LQ/NLQ>0% ?
Is there any other node linking with Scott?
- Has anyone done a 'drive-by' ?
- Did it link 'on the bench' ?
Suggestion:
1. Move the other CPE210s off of channel '-2' and put MCC-North and Scott on channel '-2'.
- Test at both 5 and 10 MHz bandwidth.
- Assuming all factors being equal, Scott should hear the CPE210 at 15 dB SNR at 10 MHz.
2. If you are going to use channel '-1', it may be prudent to do a WiFi Scan at 20 MHz at each end to note the overlapping-channel QRM levels.
- If MCC-North has a high noise floor from channel '1' part 15 devices, perhaps this node must be on channel '-2'.
My HeyWhatsThat path profile assumption is not as 'rosy' as yours. :-|
I see a significant 'bump' around 130th St.
I hope this helps,
Chuck
2. We're assuming that the .230 node is the North node because 1) no one bothered to properly mark them and 2) one of our hams did a mobile test with an omnidirectional antenna, and the .230 node was seen North of town and throughout the town. He's the ham that lives between the Wilton node and the ones on the tower, and the .230 node is the strongest signal of the ones he received.
3. I've been down to the college and have confirmed that two of the nodes are working. I switched the SE node (we know that because it was the strongest signal into Illinois of the three nodes) to Channel -2.
4. I have a node in my apartment that sees the .230 node occasionally at varying LQ's with an NLQ at 0%. But, my apartment is SW of the tower and surrounded by taller buildings. The nodes around me don't see it, or if they do, it's intermittent. We're working on getting more nodes up to test.
5. More or less, yes they did link 'on the bench'. Before we moved the nodes up on the tower, they were all in my apartment. And, I flashed all of the nodes that Scott has at his house (he only has one on his tower, but about three or four in boxes) in my apartment. Everything saw everything, although they were 75' apart at the most.
A question for you. I don't know the exact dimensions of the tower at the top, but it is a 270' tall one, so I'm assuming the legs are still a few feet apart. Would all three nodes be able to work on -2 together, or would they create too much interference with each other? The reason I switched the other node to -2 and not this one is that the ham who's already connected didn't want to have to change his equipment. LOL Plus, I didn't want to break something that was working. I made the channel swaps before Scott put his node up. I'm going out to the tower today, so I may switch it up and see what happens. If it works on Channel -2 at 10 MHz, then I guess the other ham will have to make accommodations. :)
Thank you and have a great day.
Patrick.
As a P.S. I'm hoping that node three is just plugged in wrong. Or that the cable is bad on the PoE adapter end. I like heights, but I'm not going to climb 263' without fall arrest equipment to fix the other end.
Hi, Patrick:
It would be a miracle.
For isolation, if not channel separation, then vertical separation works best.
"About halfway between the node in question and the node on the tower is another ham."
The plot profile indicates that, if between 4 to 8 miles out, this location may not have a path to Scott.
If there is a path how about linking there to Scott on 5 GHz?
If someone does climb the tower to repair the dead node, how about swapping it for a CPE510 or equivalent?
Are there 3 ethernet cables?
If it turns out that the cable has failed, maybe carry a Ubiquiti Power Switch up along with the CPE510.
"I have a node in my apartment that sees the .230 node occasionally at varying LQ's with an NLQ at 0%."
So, there are 2 nodes displaying LQ * NLQ 0%! Hmmmmm.
"the ham who's already connected didn't want to have to change his equipment. "
Move that link to channel '1' through '8' at 5, 10, 20 MHz bandwidth and he can keep his same equipment. :-|
Chuck
A mea culpa. I misspoke about Larry. He didn't want to change things until he had finished his signal testing. The changes have been made and everything to the North is working fine.
As for my apartment, I'm surrounded by 8 to 11-floor buildings. The fact that I can see anything is a miracle. But, as of right now, I see both nodes on the tower, although the .60 one drops out. The .230 varies between 32 and 73 LQ and 2 to 16 NLQ, which is better than I expected.
The third node isn't functioning yet. I think it's a bad firmware issue though. When I tested the cable, I got the same results that I got on the two that are functioning. And when I plug the cable into my computer, I get a 169.254.x.x address (which I know is that DHCP didn't assign one). I wouldn't get that from the PoE adapter, and I don't think it would show a network. Monday I'll be going back to try and reset the node. Worst case, it's bricked. Best case it works. Either way, we're not out anything.
Thank you again for all of your help.
Patrick.