# AP in cluster has poor downlink efficiency **URL:** \ **Category:** PMP\ **Created:** [September 8, 2004, 1:01pm UTC](https://community.cambiumnetworks.com/t/ap-in-cluster-has-poor-downlink-efficiency/24631 "2004-09-08T13:01:32Z")\ **Posts on this page:** 6\ **Page:** 1
**Author:** ![Bortnem1](https://d10gw3vjmzyp8.cloudfront.net/letter_avatar_proxy/v4/letter/b/a183cd/32.png) [@Bortnem1](https://community.cambiumnetworks.com/u/Bortnem1)\ **Post date:** [September 8, 2004, 1:01pm UTC](https://community.cambiumnetworks.com/t/ap-in-cluster-has-poor-downlink-efficiency/24631/1 "2004-09-08T13:01:32Z")
I have 4 AP’s in a cluster… all have 90%+ downlink efficiency, except for one, which consistenly has 50-55% downlink. All AP’s have the same settings. I have checked and re-checked everything, but just can’t figure out why this one is the exception. Any ideas? I’d like to try the easy solutions before I climb a tower. ☹ ---
**Author:** ![wpeople](https://d10gw3vjmzyp8.cloudfront.net/letter_avatar_proxy/v4/letter/w/d78d45/32.png) [@wpeople](https://community.cambiumnetworks.com/u/wpeople)\ **Post date:** [September 11, 2004, 3:59pm UTC](https://community.cambiumnetworks.com/t/ap-in-cluster-has-poor-downlink-efficiency/24631/2 "2004-09-11T15:59:17Z")
Not easy, but replace the AP with an SM and do a spectrum analysis. It should help. ---
**Author:** ![Bortnem1](https://d10gw3vjmzyp8.cloudfront.net/letter_avatar_proxy/v4/letter/b/a183cd/32.png) [@Bortnem1](https://community.cambiumnetworks.com/u/Bortnem1)\ **Post date:** [September 13, 2004, 1:14pm UTC](https://community.cambiumnetworks.com/t/ap-in-cluster-has-poor-downlink-efficiency/24631/3 "2004-09-13T13:14:05Z")
I was also thinking of taking a laptop up near the cluster and doing a spectrum analysis with NetStumbler. I’d prefer not taking down the AP if I can help it. ---
**Author:** ![ginovilla](https://d10gw3vjmzyp8.cloudfront.net/letter_avatar_proxy/v4/letter/g/c57346/32.png) [@ginovilla](https://community.cambiumnetworks.com/u/ginovilla)\ **Post date:** [September 13, 2004, 4:18pm UTC](https://community.cambiumnetworks.com/t/ap-in-cluster-has-poor-downlink-efficiency/24631/4 "2004-09-13T16:18:19Z")
Spectrum Analisys with NetStumbler?.. AFAIK, Net stumbler is not a Spectrum Analizer… its a 802.11b sniffer… So Im guessing your on 2.4? if your interference source is from a non 802.11b type of gear, netstumbler wont detect it Gino ---
**Author:** ![Bortnem1](https://d10gw3vjmzyp8.cloudfront.net/letter_avatar_proxy/v4/letter/b/a183cd/32.png) [@Bortnem1](https://community.cambiumnetworks.com/u/Bortnem1)\ **Post date:** [September 14, 2004, 12:57pm UTC](https://community.cambiumnetworks.com/t/ap-in-cluster-has-poor-downlink-efficiency/24631/5 "2004-09-14T12:57:43Z")
> | ginovilla wrote: | > | --- | > | Spectrum Analisys with NetStumbler?... > > AFAIK, Net stumbler is not a Spectrum Analizer... its a 802.11b sniffer... > > So Im guessing your on 2.4? if your interference source is from a non 802.11b type of gear, netstumbler wont detect it > > Gino | The way it was explained to me, NetStumbler is a SpecAn. Perhaps I got some bum info. I have never played with it, so thanks for the heads up. The reason I'm reluctant to use the SM as a SpecAn is that we've never really been able to make the SpecAn function work correctly, so I don't have that much confidence in it.... ---
**Author:** ![jduncanaz](https://d10gw3vjmzyp8.cloudfront.net/letter_avatar_proxy/v4/letter/j/82dd89/32.png) [@jduncanaz](https://community.cambiumnetworks.com/u/jduncanaz)\ **Post date:** [September 19, 2004, 1:06am UTC](https://community.cambiumnetworks.com/t/ap-in-cluster-has-poor-downlink-efficiency/24631/6 "2004-09-19T01:06:42Z")
Is the cluster using a CMM w/ GPS? I Have a new site going up w/ 2.4 and am waiting on a lmr 400 BNC connector to come in before turning on the GPS antenna sync pulse. Have 2 ap’s generating their own sync pulse mounted vertically on top of each other. One is producing high error rates the other is working great. This site is remote so I have not had a chance to do more spectrum sweeps. I suspect it is the fact that it is not timed or it maybe the fact they are mounted vertically to each other…this will change the next time I go up. Jason Duncan CableAmerica Corp. 480-461-6887 x222