Re: [WISPA] Bad radio?
okay, now 3 days later the new south radio (originally north) is misbehaving in the same way (Tx/Rx signals are off by 10 where on the other sectors they're roughly balanced). The signals going back to the tower are 8 - 10 db stronger than the received signals. An R52 (or possibly R52H) is shooting back to the tower whereas an XR5 is shooting down to the CPE. If it were different radio specs, it'd be unbalanced the other way around. The SR5s that I have installed have been functioning just fine for years and are installed in the same fashion as the XR5s. Did UBNT make a bad batch of XR5s or am I doing something wrong? -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Mike Hammett [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 2:58 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? I changed the freq a bit and didn't see any significant change. I climbed and swapped the pigtails with North and changed all the settings so they were fully swapped. Everyone on the south sector (which did have the questionable radio) is now happy. Now the North sector is showing the same symptoms on my test CPE. It's a good thing everyone except two are on East and West until I get this sorted out. Time to contact Streakwave about a bad radio... -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: RickG [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 1:05 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? It would be interesting to switch the freqs around and see what happens. It's probably a bad radio or cable though. -RickG On Sun, Sep 28, 2008 at 1:06 PM, Mike Hammett [EMAIL PROTECTED] wrote: Does this seem like the radio isn't loud enough? I setup the test CPE on the TV tower at my house and pointed it at the tower. Radio Mobile reports the azimuth as 250 degrees, so well within the south sector's coverage and only at a distance of 230'. I am well below the vertical beamwidth of the sectors, explaining the relatively low signals, but ICS2 is horrible. 1 = North, 2 = South, 3 = West, 4 = East. It makes no sense that South is that much worse signal wise than the others, especially considering that it should be on the South sector anyway. 3 and 4 are SR5s while 1 and 2 are XR5s. I just replaced the towers with the XR5s. [EMAIL PROTECTED] Test Platform] /interface wireless scan wlan1 Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG NF SNR RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -72 -99 27 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -68 -99 31 00156D501709 AB R 00:15:6D:64:0B:59 ICS1 5ghz 5785 -77 -99 22 00156D640B59 AB R 00:15:6D:64:0B:55 ICS2 5ghz 5825 -85 -99 14 00156D640B55 Here is a listing of the signals when connected: [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5785MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS1 bssid: 00:15:6D:64:0B:59 radio-name: 00156D640B59 signal-strength: -77dBm tx-signal-strength: -74dBm noise-floor: -107dBm signal-to-noise: 30dB tx-ccq: 58% p-throughput: 5481 overall-tx-ccq: 58% authenticated-clients: 1 current-ack-timeout: 28 wds-link: no nstreme: no framing-mode: none routeros-version: 2.9.51 last-ip: 10.10.1.1 802.1x-port-enabled: yes compression: no current-tx-powers: 6Mbps:24(24),9Mbps:24(24),12Mbps:24(24),18Mbps:24(24),24Mbps:24(24),36Mbps:22(22),48Mbps:20(20),54Mbps:19(19) notify-external-fdb: no [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5825MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS2 bssid: 00:15:6D:64:0B:55 radio-name: 00156D640B55 signal-strength: -86dBm tx-signal-strength: -76dBm noise-floor: -107dBm signal-to-noise: 21dB tx-ccq: 59% p-throughput: 5535 overall-tx-ccq: 58% authenticated-clients: 1 current-ack-timeout: 167 wds-link: no nstreme: no framing-mode: none routeros-version: 2.9.51 802.1x-port-enabled: yes
Re: [WISPA] Bad radio?
First, read this: http://www.enterprisenetworkingplanet.com/netsysm/article.php/3765946 It might sound strange but sometimes RSSI can be effected by interference. It's important to run the calcs on what you're running. I have a nice spread sheet if you need a calculator. I really really miss the YDI online one. Nice and simple. Not full of junk no one uses. Oh well. To offer more realistic help here we need to know a lot more info What TX power are the radios set for? How much coax? Lightning arrestors? Amps (db)? What antenna gain? If sectors, what coverage and how are they pointing? (example, customers are 500' lower than the antenna and 1 to 20 miles away. Antenna is downtilted 25*) How long are the cat 5 runs? Any other radio systems on that tower or near by (less than one mile)? If you do an ap scan from each AP (usually have to put them in client mode but some will do so via ap mode) how many other systems do you see and what levels do you see them at? Do you have the ability to run a general spectrum scan? If so, what does it show? (NOTE: If you run this test make sure to run a second test with all of the other AP's that you control turned off.) I turn the power wayy down on almost all of my AP's these days. Most are only putting out 15 to 20 dB. I use slightly larger customer antennas to make up for the AP TX power losses. This has REALLY helped the speeds and stability on my overall system. Believe it or not, I have customers at nearly 15 miles pulling 2 to 3 megs both ways via an 8 dB omni fed by a radio turned down to 17dB. I now have several customers over 15 miles that pull from 13dB sectors fed by 17dB ap's. Speeds are HIGHER than they were when I was running amped systems at 4 watts. Happy hunting! marlon - Original Message - From: Mike Hammett [EMAIL PROTECTED] To: WISPA General List wireless@wispa.org Sent: Wednesday, October 01, 2008 6:48 AM Subject: Re: [WISPA] Bad radio? okay, now 3 days later the new south radio (originally north) is misbehaving in the same way (Tx/Rx signals are off by 10 where on the other sectors they're roughly balanced). The signals going back to the tower are 8 - 10 db stronger than the received signals. An R52 (or possibly R52H) is shooting back to the tower whereas an XR5 is shooting down to the CPE. If it were different radio specs, it'd be unbalanced the other way around. The SR5s that I have installed have been functioning just fine for years and are installed in the same fashion as the XR5s. Did UBNT make a bad batch of XR5s or am I doing something wrong? -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Mike Hammett [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 2:58 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? I changed the freq a bit and didn't see any significant change. I climbed and swapped the pigtails with North and changed all the settings so they were fully swapped. Everyone on the south sector (which did have the questionable radio) is now happy. Now the North sector is showing the same symptoms on my test CPE. It's a good thing everyone except two are on East and West until I get this sorted out. Time to contact Streakwave about a bad radio... -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: RickG [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 1:05 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? It would be interesting to switch the freqs around and see what happens. It's probably a bad radio or cable though. -RickG On Sun, Sep 28, 2008 at 1:06 PM, Mike Hammett [EMAIL PROTECTED] wrote: Does this seem like the radio isn't loud enough? I setup the test CPE on the TV tower at my house and pointed it at the tower. Radio Mobile reports the azimuth as 250 degrees, so well within the south sector's coverage and only at a distance of 230'. I am well below the vertical beamwidth of the sectors, explaining the relatively low signals, but ICS2 is horrible. 1 = North, 2 = South, 3 = West, 4 = East. It makes no sense that South is that much worse signal wise than the others, especially considering that it should be on the South sector anyway. 3 and 4 are SR5s while 1 and 2 are XR5s. I just replaced the towers with the XR5s. [EMAIL PROTECTED] Test Platform] /interface wireless scan wlan1 Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG NF SNR RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -72 -99 27 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -68 -99 31 00156D501709 AB R 00:15
Re: [WISPA] Bad radio?
Can anyone tell me what an XR5 puts out for signal if it has blown it's amp? I'm starting to believe something is causing the XR5s to blow their amps, reducing them to pre-amp power. What that is remains to be seen. I'm hoping it's the radio itself, pigtail, or the coax vs. the $350 MTI I waited 3 weeks for. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Mike Hammett [EMAIL PROTECTED] Sent: Wednesday, October 01, 2008 8:48 AM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? okay, now 3 days later the new south radio (originally north) is misbehaving in the same way (Tx/Rx signals are off by 10 where on the other sectors they're roughly balanced). The signals going back to the tower are 8 - 10 db stronger than the received signals. An R52 (or possibly R52H) is shooting back to the tower whereas an XR5 is shooting down to the CPE. If it were different radio specs, it'd be unbalanced the other way around. The SR5s that I have installed have been functioning just fine for years and are installed in the same fashion as the XR5s. Did UBNT make a bad batch of XR5s or am I doing something wrong? -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Mike Hammett [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 2:58 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? I changed the freq a bit and didn't see any significant change. I climbed and swapped the pigtails with North and changed all the settings so they were fully swapped. Everyone on the south sector (which did have the questionable radio) is now happy. Now the North sector is showing the same symptoms on my test CPE. It's a good thing everyone except two are on East and West until I get this sorted out. Time to contact Streakwave about a bad radio... -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: RickG [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 1:05 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? It would be interesting to switch the freqs around and see what happens. It's probably a bad radio or cable though. -RickG On Sun, Sep 28, 2008 at 1:06 PM, Mike Hammett [EMAIL PROTECTED] wrote: Does this seem like the radio isn't loud enough? I setup the test CPE on the TV tower at my house and pointed it at the tower. Radio Mobile reports the azimuth as 250 degrees, so well within the south sector's coverage and only at a distance of 230'. I am well below the vertical beamwidth of the sectors, explaining the relatively low signals, but ICS2 is horrible. 1 = North, 2 = South, 3 = West, 4 = East. It makes no sense that South is that much worse signal wise than the others, especially considering that it should be on the South sector anyway. 3 and 4 are SR5s while 1 and 2 are XR5s. I just replaced the towers with the XR5s. [EMAIL PROTECTED] Test Platform] /interface wireless scan wlan1 Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG NF SNR RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -72 -99 27 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -68 -99 31 00156D501709 AB R 00:15:6D:64:0B:59 ICS1 5ghz 5785 -77 -99 22 00156D640B59 AB R 00:15:6D:64:0B:55 ICS2 5ghz 5825 -85 -99 14 00156D640B55 Here is a listing of the signals when connected: [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5785MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS1 bssid: 00:15:6D:64:0B:59 radio-name: 00156D640B59 signal-strength: -77dBm tx-signal-strength: -74dBm noise-floor: -107dBm signal-to-noise: 30dB tx-ccq: 58% p-throughput: 5481 overall-tx-ccq: 58% authenticated-clients: 1 current-ack-timeout: 28 wds-link: no nstreme: no framing-mode: none routeros-version: 2.9.51 last-ip: 10.10.1.1 802.1x-port-enabled: yes compression: no current-tx-powers: 6Mbps:24(24),9Mbps:24(24),12Mbps:24(24),18Mbps:24(24),24Mbps:24(24),36Mbps:22(22),48Mbps:20(20),54Mbps:19(19) notify-external-fdb: no [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz
Re: [WISPA] Bad radio?
Not sure if this was asked, but is your board powering these cards properly? Mark Nash UnwiredWest 78 Centennial Loop Suite E Eugene, OR 97401 541-998- 541-998-5599 fax http://www.unwiredwest.com - Original Message - From: Mike Hammett [EMAIL PROTECTED] To: WISPA General List wireless@wispa.org Sent: Wednesday, October 01, 2008 8:34 AM Subject: Re: [WISPA] Bad radio? Can anyone tell me what an XR5 puts out for signal if it has blown it's amp? I'm starting to believe something is causing the XR5s to blow their amps, reducing them to pre-amp power. What that is remains to be seen. I'm hoping it's the radio itself, pigtail, or the coax vs. the $350 MTI I waited 3 weeks for. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Mike Hammett [EMAIL PROTECTED] Sent: Wednesday, October 01, 2008 8:48 AM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? okay, now 3 days later the new south radio (originally north) is misbehaving in the same way (Tx/Rx signals are off by 10 where on the other sectors they're roughly balanced). The signals going back to the tower are 8 - 10 db stronger than the received signals. An R52 (or possibly R52H) is shooting back to the tower whereas an XR5 is shooting down to the CPE. If it were different radio specs, it'd be unbalanced the other way around. The SR5s that I have installed have been functioning just fine for years and are installed in the same fashion as the XR5s. Did UBNT make a bad batch of XR5s or am I doing something wrong? -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Mike Hammett [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 2:58 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? I changed the freq a bit and didn't see any significant change. I climbed and swapped the pigtails with North and changed all the settings so they were fully swapped. Everyone on the south sector (which did have the questionable radio) is now happy. Now the North sector is showing the same symptoms on my test CPE. It's a good thing everyone except two are on East and West until I get this sorted out. Time to contact Streakwave about a bad radio... -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: RickG [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 1:05 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? It would be interesting to switch the freqs around and see what happens. It's probably a bad radio or cable though. -RickG On Sun, Sep 28, 2008 at 1:06 PM, Mike Hammett [EMAIL PROTECTED] wrote: Does this seem like the radio isn't loud enough? I setup the test CPE on the TV tower at my house and pointed it at the tower. Radio Mobile reports the azimuth as 250 degrees, so well within the south sector's coverage and only at a distance of 230'. I am well below the vertical beamwidth of the sectors, explaining the relatively low signals, but ICS2 is horrible. 1 = North, 2 = South, 3 = West, 4 = East. It makes no sense that South is that much worse signal wise than the others, especially considering that it should be on the South sector anyway. 3 and 4 are SR5s while 1 and 2 are XR5s. I just replaced the towers with the XR5s. [EMAIL PROTECTED] Test Platform] /interface wireless scan wlan1 Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG NF SNR RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -72 -99 27 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -68 -99 31 00156D501709 AB R 00:15:6D:64:0B:59 ICS1 5ghz 5785 -77 -99 22 00156D640B59 AB R 00:15:6D:64:0B:55 ICS2 5ghz 5825 -85 -99 14 00156D640B55 Here is a listing of the signals when connected: [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5785MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS1 bssid: 00:15:6D:64:0B:59 radio-name: 00156D640B59 signal-strength: -77dBm tx-signal-strength: -74dBm noise-floor: -107dBm signal-to-noise: 30dB tx-ccq: 58% p-throughput: 5481 overall-tx-ccq: 58% authenticated-clients: 1 current-ack-timeout: 28 wds-link
Re: [WISPA] Bad radio?
*shrugs* It's a Mikrotik 4 slot mPCI - PCI adapter modified (by Mikrotik) to power higher powered cards. It previously contained all SR5s. I'm looking to convert to 4x RB411AHs instead of 1x PC. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Mark Nash [EMAIL PROTECTED] Sent: Wednesday, October 01, 2008 10:39 AM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? Not sure if this was asked, but is your board powering these cards properly? Mark Nash UnwiredWest 78 Centennial Loop Suite E Eugene, OR 97401 541-998- 541-998-5599 fax http://www.unwiredwest.com - Original Message - From: Mike Hammett [EMAIL PROTECTED] To: WISPA General List wireless@wispa.org Sent: Wednesday, October 01, 2008 8:34 AM Subject: Re: [WISPA] Bad radio? Can anyone tell me what an XR5 puts out for signal if it has blown it's amp? I'm starting to believe something is causing the XR5s to blow their amps, reducing them to pre-amp power. What that is remains to be seen. I'm hoping it's the radio itself, pigtail, or the coax vs. the $350 MTI I waited 3 weeks for. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Mike Hammett [EMAIL PROTECTED] Sent: Wednesday, October 01, 2008 8:48 AM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? okay, now 3 days later the new south radio (originally north) is misbehaving in the same way (Tx/Rx signals are off by 10 where on the other sectors they're roughly balanced). The signals going back to the tower are 8 - 10 db stronger than the received signals. An R52 (or possibly R52H) is shooting back to the tower whereas an XR5 is shooting down to the CPE. If it were different radio specs, it'd be unbalanced the other way around. The SR5s that I have installed have been functioning just fine for years and are installed in the same fashion as the XR5s. Did UBNT make a bad batch of XR5s or am I doing something wrong? -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Mike Hammett [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 2:58 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? I changed the freq a bit and didn't see any significant change. I climbed and swapped the pigtails with North and changed all the settings so they were fully swapped. Everyone on the south sector (which did have the questionable radio) is now happy. Now the North sector is showing the same symptoms on my test CPE. It's a good thing everyone except two are on East and West until I get this sorted out. Time to contact Streakwave about a bad radio... -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: RickG [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 1:05 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? It would be interesting to switch the freqs around and see what happens. It's probably a bad radio or cable though. -RickG On Sun, Sep 28, 2008 at 1:06 PM, Mike Hammett [EMAIL PROTECTED] wrote: Does this seem like the radio isn't loud enough? I setup the test CPE on the TV tower at my house and pointed it at the tower. Radio Mobile reports the azimuth as 250 degrees, so well within the south sector's coverage and only at a distance of 230'. I am well below the vertical beamwidth of the sectors, explaining the relatively low signals, but ICS2 is horrible. 1 = North, 2 = South, 3 = West, 4 = East. It makes no sense that South is that much worse signal wise than the others, especially considering that it should be on the South sector anyway. 3 and 4 are SR5s while 1 and 2 are XR5s. I just replaced the towers with the XR5s. [EMAIL PROTECTED] Test Platform] /interface wireless scan wlan1 Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG NF SNR RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -72 -99 27 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -68 -99 31 00156D501709 AB R 00:15:6D:64:0B:59 ICS1 5ghz 5785 -77 -99 22 00156D640B59 AB R 00:15:6D:64:0B:55 ICS2 5ghz 5825 -85 -99 14 00156D640B55 Here is a listing of the signals when connected: [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band
Re: [WISPA] Bad radio?
The TX is currently set to default in MT 2.9.51, whatever dB that turns out to be. I was going to tune it once the system stabilized. Maybe 20' of LMR-400. No lightning arrestors. The other sectors on the tower sat there for years without them and no damage, so when I upgraded and had the wrong pigtails to use the existing arrestors, I removed them. No amps, just the XR5s... I'd never use an amp. Roughly 16 dBi gain for all 4 sectors pointed North, South, East, and West. They all have roughly 5 degrees of downtilt. No PoE, the MT system is a PC with a 4 slot mPCI adapter. The only other cat 5 is going to an Orthogon Gemini (and it's PoE). This is pointed east and is on the opposite side of the grain leg from the troubled sector. Nothing else within 1 mile that uses upper 5 GHz. Well, there's CPE in unknown bands, but they didn't affect the previous radio\antenna combo. All customers use 19 or 24 dBi RooTennas, depending on distance from tower. I believe everyone is between -60 and -75. For some reason, some of these sectors are much louder than they were previously, though I don't have any documentation as to who saw what before. What do band-pass filters cost and where can I get them? Maybe I ought to invest in some of those. North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 North = 1 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -49 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -37 00156D501709 AB R 00:15:6D:64:0B:59 ICS2 5ghz 5805 -43 00156D640B59 AB RN 00:0C:42:05:51:B7 Walter 5ghz 5260 -71 000C420551B7 South = 2 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -51 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -38 00156D501709 AB R 00:15:6D:64:0B:55 ICS1 5ghz 5785 -29 00156D640B55 East = 4 === Will run a scan during non-peak times. West = 3 === Will run a scan during non-peak times. Mikrotik reports a debatable noise floor reading, which is supposed to represent all non-802.11 systems. It isn't worse than -99 on any sector. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Marlon K. Schafer [EMAIL PROTECTED] Sent: Wednesday, October 01, 2008 9:59 AM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? First, read this: http://www.enterprisenetworkingplanet.com/netsysm/article.php/3765946 It might sound strange but sometimes RSSI can be effected by interference. It's important to run the calcs on what you're running. I have a nice spread sheet if you need a calculator. I really really miss the YDI online one. Nice and simple. Not full of junk no one uses. Oh well. To offer more realistic help here we need to know a lot more info What TX power are the radios set for? How much coax? Lightning arrestors? Amps (db)? What antenna gain? If sectors, what coverage and how are they pointing? (example, customers are 500' lower than the antenna and 1 to 20 miles away. Antenna is downtilted 25*) How long are the cat 5 runs? Any other radio systems on that tower or near by (less than one mile)? If you do an ap scan from each AP (usually have to put them in client mode but some will do so via ap mode) how many other systems do you see and what levels do you see them at? Do you have the ability to run a general spectrum scan? If so, what does it show? (NOTE: If you run this test make sure to run a second test with all of the other AP's that you control turned off.) I turn the power wayy down on almost all of my AP's these days. Most are only putting out 15 to 20 dB. I use slightly larger customer antennas to make up for the AP TX power losses. This has REALLY helped the speeds and stability on my overall system. Believe it or not, I have customers at nearly 15 miles pulling 2 to 3 megs both ways via an 8 dB omni fed by a radio turned down to 17dB. I now have several customers over 15 miles that pull from 13dB sectors fed by 17dB ap's. Speeds are HIGHER than they were when I was running amped systems at 4 watts. Happy hunting! marlon - Original Message - From: Mike Hammett [EMAIL PROTECTED] To: WISPA General List wireless@wispa.org Sent: Wednesday, October 01, 2008 6:48 AM Subject: Re: [WISPA] Bad radio? okay, now 3 days later the new south radio (originally north) is misbehaving in the same way (Tx/Rx signals are off by 10 where on the other sectors they're roughly balanced). The signals going back to the tower are 8 - 10 db stronger than the received signals. An R52 (or possibly R52H) is shooting back
Re: [WISPA] Bad radio?
This really looks like you are causing yourself all kinds of interference. Using the channels you listed: North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 I'm sure they are stepping on each other. The signal doesn't just drop completely off on the edges. I assume 20mhz wide channels (because you didn't specify), meaning North and South edges are right next to each other, but in the same case, on the same card, I'm sure you are stepping on yourself. Same with the East and West ones. So, on the North you are actually using 5775 to 5795. While on the South you are 5795 to 5815. I can tell you right now that even having 10mhz between channel edges isn't going to be enough... you will need 20-40mhz of spacing between the edges of the band to keep from interfering with yourself. Travis Microserv Mike Hammett wrote: The TX is currently set to default in MT 2.9.51, whatever dB that turns out to be. I was going to tune it once the system stabilized. Maybe 20' of LMR-400. No lightning arrestors. The other sectors on the tower sat there for years without them and no damage, so when I upgraded and had the wrong pigtails to use the existing arrestors, I removed them. No amps, just the XR5s... I'd never use an amp. Roughly 16 dBi gain for all 4 sectors pointed North, South, East, and West. They all have roughly 5 degrees of downtilt. No PoE, the MT system is a PC with a 4 slot mPCI adapter. The only other cat 5 is going to an Orthogon Gemini (and it's PoE). This is pointed east and is on the opposite side of the grain leg from the troubled sector. Nothing else within 1 mile that uses upper 5 GHz. Well, there's CPE in unknown bands, but they didn't affect the previous radio\antenna combo. All customers use 19 or 24 dBi RooTennas, depending on distance from tower. I believe everyone is between -60 and -75. For some reason, some of these sectors are much louder than they were previously, though I don't have any documentation as to who saw what before. What do band-pass filters cost and where can I get them? Maybe I ought to invest in some of those. North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 North = 1 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -49 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -37 00156D501709 AB R 00:15:6D:64:0B:59 ICS2 5ghz 5805 -43 00156D640B59 AB RN 00:0C:42:05:51:B7 Walter 5ghz 5260 -71 000C420551B7 South = 2 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -51 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -38 00156D501709 AB R 00:15:6D:64:0B:55 ICS1 5ghz 5785 -29 00156D640B55 East = 4 === Will run a scan during non-peak times. West = 3 === Will run a scan during non-peak times. Mikrotik reports a debatable noise floor reading, which is supposed to represent all non-802.11 systems. It isn't worse than -99 on any sector. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: "Marlon K. Schafer" [EMAIL PROTECTED] Sent: Wednesday, October 01, 2008 9:59 AM To: "WISPA General List" wireless@wispa.org Subject: Re: [WISPA] Bad radio? First, read this: http://www.enterprisenetworkingplanet.com/netsysm/article.php/3765946 It might sound strange but sometimes RSSI can be effected by interference. It's important to run the calcs on what you're running. I have a nice spread sheet if you need a calculator. I really really miss the YDI online one. Nice and simple. Not full of junk no one uses. Oh well. To offer more realistic help here we need to know a lot more info What TX power are the radios set for? How much coax? Lightning arrestors? Amps (db)? What antenna gain? If sectors, what coverage and how are they pointing? (example, customers are 500' lower than the antenna and 1 to 20 miles away. Antenna is downtilted 25*) How long are the cat 5 runs? Any other radio systems on that tower or near by (less than one mile)? If you do an ap scan from each AP (usually have to put them in client mode but some will do so via ap mode) how many other systems do you see and what levels do you see them at? Do you have the ability to run a general spectrum scan? If so, what does it show? (NOTE: If you run this test make sure to run a second test with all of the other AP's that you control turned off.) I turn the power wayy down on almost all of my AP's these days. Most are only putting out 15 to 20 dB. I use slightly larger customer antennas to make up for the AP TX power losses. This has REALLY helped the speeds and stability on my over
Re: [WISPA] Bad radio?
So are you telling me that I can only run 2 or 3 radios in upper 5 GHz without stepping on myself? I was pretty sure nothing was greater than -60 or so before I made these tower changes, but silly me, I didn't bother to document what I saw. Some of those signals don't make any sense, either. Between North and East and North and South are gigantic chunks of metal to where I doubt you could even physically see a foot or two to the side of either sector. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com From: Travis Johnson Sent: Wednesday, October 01, 2008 11:21 AM To: WISPA General List Subject: Re: [WISPA] Bad radio? This really looks like you are causing yourself all kinds of interference. Using the channels you listed: North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 I'm sure they are stepping on each other. The signal doesn't just drop completely off on the edges. I assume 20mhz wide channels (because you didn't specify), meaning North and South edges are right next to each other, but in the same case, on the same card, I'm sure you are stepping on yourself. Same with the East and West ones. So, on the North you are actually using 5775 to 5795. While on the South you are 5795 to 5815. I can tell you right now that even having 10mhz between channel edges isn't going to be enough... you will need 20-40mhz of spacing between the edges of the band to keep from interfering with yourself. Travis Microserv Mike Hammett wrote: The TX is currently set to default in MT 2.9.51, whatever dB that turns out to be. I was going to tune it once the system stabilized. Maybe 20' of LMR-400. No lightning arrestors. The other sectors on the tower sat there for years without them and no damage, so when I upgraded and had the wrong pigtails to use the existing arrestors, I removed them. No amps, just the XR5s... I'd never use an amp. Roughly 16 dBi gain for all 4 sectors pointed North, South, East, and West. They all have roughly 5 degrees of downtilt. No PoE, the MT system is a PC with a 4 slot mPCI adapter. The only other cat 5 is going to an Orthogon Gemini (and it's PoE). This is pointed east and is on the opposite side of the grain leg from the troubled sector. Nothing else within 1 mile that uses upper 5 GHz. Well, there's CPE in unknown bands, but they didn't affect the previous radio\antenna combo. All customers use 19 or 24 dBi RooTennas, depending on distance from tower. I believe everyone is between -60 and -75. For some reason, some of these sectors are much louder than they were previously, though I don't have any documentation as to who saw what before. What do band-pass filters cost and where can I get them? Maybe I ought to invest in some of those. North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 North = 1 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -49 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -37 00156D501709 AB R 00:15:6D:64:0B:59 ICS2 5ghz 5805 -43 00156D640B59 AB RN 00:0C:42:05:51:B7 Walter 5ghz 5260 -71 000C420551B7 South = 2 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -51 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -38 00156D501709 AB R 00:15:6D:64:0B:55 ICS1 5ghz 5785 -29 00156D640B55 East = 4 === Will run a scan during non-peak times. West = 3 === Will run a scan during non-peak times. Mikrotik reports a debatable noise floor reading, which is supposed to represent all non-802.11 systems. It isn't worse than -99 on any sector. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Marlon K. Schafer [EMAIL PROTECTED] Sent: Wednesday, October 01, 2008 9:59 AM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? First, read this: http://www.enterprisenetworkingplanet.com/netsysm/article.php/3765946 It might sound strange but sometimes RSSI can be effected by interference. It's important to run the calcs on what you're running. I have a nice spread sheet if you need a calculator. I really really miss the YDI online one. Nice and simple. Not full of junk no one uses. Oh well. To offer more realistic help here we need to know a lot more info What TX power are the radios set for? How much coax? Lightning arrestors? Amps (db)? What antenna gain? If sectors, what coverage and how are they pointing? (example, customers are 500' lower than the antenna and 1 to 20 miles away. Antenna is downtilted 25*) How long are the cat 5 runs? Any other radio systems on that tower or near by (less than one
Re: [WISPA] Bad radio?
The noise is coming from the actual wireless cards being so close to each other. Not all of the signal is going out the cable... and with cards stacked within inches from each other, more noise bleeds over. Travis Microserv Mike Hammett wrote: So are you telling me that I can only run 2 or 3 radios in upper 5 GHz without stepping on myself? I was pretty sure nothing was greater than -60 or so before I made these tower changes, but silly me, I didn't bother to document what I saw. Some of those signals don't make any sense, either. Between North and East and North and South are gigantic chunks of metal to where I doubt you could even physically see a foot or two to the side of either sector. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com From: Travis Johnson Sent: Wednesday, October 01, 2008 11:21 AM To: WISPA General List Subject: Re: [WISPA] Bad radio? This really looks like you are causing yourself all kinds of interference. Using the channels you listed: North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 I'm sure they are stepping on each other. The signal doesn't just drop completely off on the edges. I assume 20mhz wide channels (because you didn't specify), meaning North and South edges are right next to each other, but in the same case, on the same card, I'm sure you are stepping on yourself. Same with the East and West ones. So, on the North you are actually using 5775 to 5795. While on the South you are 5795 to 5815. I can tell you right now that even having 10mhz between channel edges isn't going to be enough... you will need 20-40mhz of spacing between the edges of the band to keep from interfering with yourself. Travis Microserv Mike Hammett wrote: The TX is currently set to default in MT 2.9.51, whatever dB that turns out to be. I was going to tune it once the system stabilized. Maybe 20' of LMR-400. No lightning arrestors. The other sectors on the tower sat there for years without them and no damage, so when I upgraded and had the wrong pigtails to use the existing arrestors, I removed them. No amps, just the XR5s... I'd never use an amp. Roughly 16 dBi gain for all 4 sectors pointed North, South, East, and West. They all have roughly 5 degrees of downtilt. No PoE, the MT system is a PC with a 4 slot mPCI adapter. The only other cat 5 is going to an Orthogon Gemini (and it's PoE). This is pointed east and is on the opposite side of the grain leg from the troubled sector. Nothing else within 1 mile that uses upper 5 GHz. Well, there's CPE in unknown bands, but they didn't affect the previous radio\antenna combo. All customers use 19 or 24 dBi RooTennas, depending on distance from tower. I believe everyone is between -60 and -75. For some reason, some of these sectors are much louder than they were previously, though I don't have any documentation as to who saw what before. What do band-pass filters cost and where can I get them? Maybe I ought to invest in some of those. North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 North = 1 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -49 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -37 00156D501709 AB R 00:15:6D:64:0B:59 ICS2 5ghz 5805 -43 00156D640B59 AB RN 00:0C:42:05:51:B7 Walter 5ghz 5260 -71 000C420551B7 South = 2 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -51 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -38 00156D501709 AB R 00:15:6D:64:0B:55 ICS1 5ghz 5785 -29 00156D640B55 East = 4 === Will run a scan during non-peak times. West = 3 === Will run a scan during non-peak times. Mikrotik reports a debatable noise floor reading, which is supposed to represent all non-802.11 systems. It isn't worse than -99 on any sector. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Marlon K. Schafer [EMAIL PROTECTED] Sent: Wednesday, October 01, 2008 9:59 AM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? First, read this: http://www.enterprisenetworkingplanet.com/netsysm/article.php/3765946 It might sound strange but sometimes RSSI can be effected by interference. It's important to run the calcs on what you're running. I have a nice spread sheet if you need a calculator. I really really miss the YDI online one. Nice and simple. Not full of junk no one uses. Oh well. To offer more realistic help here we need to know a lot more info What TX
Re: [WISPA] Bad radio?
I have a genuine interest to know what's going on here. Why wasn't this an issue before and what effect does this have on a radio no longer transmitting as powerfully? I could see it going deaf, but it hears just fine. I'm thinking about ditching the PC and mPCI - PCI adapter and going with 4x RB411AHs instead, providing some increased radio separation. Depending on how I do it, there may be sheet metal between each RB411AH as well. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: Travis Johnson [EMAIL PROTECTED] Sent: Wednesday, October 01, 2008 12:25 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? The noise is coming from the actual wireless cards being so close to each other. Not all of the signal is going out the cable... and with cards stacked within inches from each other, more noise bleeds over. Travis Microserv Mike Hammett wrote: So are you telling me that I can only run 2 or 3 radios in upper 5 GHz without stepping on myself? I was pretty sure nothing was greater than -60 or so before I made these tower changes, but silly me, I didn't bother to document what I saw. Some of those signals don't make any sense, either. Between North and East and North and South are gigantic chunks of metal to where I doubt you could even physically see a foot or two to the side of either sector. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com From: Travis Johnson Sent: Wednesday, October 01, 2008 11:21 AM To: WISPA General List Subject: Re: [WISPA] Bad radio? This really looks like you are causing yourself all kinds of interference. Using the channels you listed: North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 I'm sure they are stepping on each other. The signal doesn't just drop completely off on the edges. I assume 20mhz wide channels (because you didn't specify), meaning North and South edges are right next to each other, but in the same case, on the same card, I'm sure you are stepping on yourself. Same with the East and West ones. So, on the North you are actually using 5775 to 5795. While on the South you are 5795 to 5815. I can tell you right now that even having 10mhz between channel edges isn't going to be enough... you will need 20-40mhz of spacing between the edges of the band to keep from interfering with yourself. Travis Microserv Mike Hammett wrote: The TX is currently set to default in MT 2.9.51, whatever dB that turns out to be. I was going to tune it once the system stabilized. Maybe 20' of LMR-400. No lightning arrestors. The other sectors on the tower sat there for years without them and no damage, so when I upgraded and had the wrong pigtails to use the existing arrestors, I removed them. No amps, just the XR5s... I'd never use an amp. Roughly 16 dBi gain for all 4 sectors pointed North, South, East, and West. They all have roughly 5 degrees of downtilt. No PoE, the MT system is a PC with a 4 slot mPCI adapter. The only other cat 5 is going to an Orthogon Gemini (and it's PoE). This is pointed east and is on the opposite side of the grain leg from the troubled sector. Nothing else within 1 mile that uses upper 5 GHz. Well, there's CPE in unknown bands, but they didn't affect the previous radio\antenna combo. All customers use 19 or 24 dBi RooTennas, depending on distance from tower. I believe everyone is between -60 and -75. For some reason, some of these sectors are much louder than they were previously, though I don't have any documentation as to who saw what before. What do band-pass filters cost and where can I get them? Maybe I ought to invest in some of those. North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 North = 1 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -49 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -37 00156D501709 AB R 00:15:6D:64:0B:59 ICS2 5ghz 5805 -43 00156D640B59 AB RN 00:0C:42:05:51:B7 Walter 5ghz 5260 -71 000C420551B7 South = 2 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -51 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -38 00156D501709 AB R 00:15:6D:64:0B:55 ICS1 5ghz 5785 -29 00156D640B55 East = 4 === Will run a scan during non-peak times. West = 3 === Will run a scan during non-peak times. Mikrotik reports a debatable noise floor reading, which is supposed to represent all non-802.11 systems. It isn't worse than -99 on any sector. -- Mike Hammett
Re: [WISPA] Bad radio?
Didn't you just "upgrade" to the XR series cards? They put out a lot more power, thus creating issues inside the box. You could turn the power way down or even turn 1 or 2 cards off and see if that fixes the issues. Travis Microserv Mike Hammett wrote: I have a genuine interest to know what's going on here. Why wasn't this an issue before and what effect does this have on a radio no longer transmitting as powerfully? I could see it going deaf, but it hears just fine. I'm thinking about ditching the PC and mPCI - PCI adapter and going with 4x RB411AHs instead, providing some increased radio separation. Depending on how I do it, there may be sheet metal between each RB411AH as well. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: "Travis Johnson" [EMAIL PROTECTED] Sent: Wednesday, October 01, 2008 12:25 PM To: "WISPA General List" wireless@wispa.org Subject: Re: [WISPA] Bad radio? The noise is coming from the actual wireless cards being so close to each other. Not all of the signal is going out the cable... and with cards stacked within inches from each other, more noise bleeds over. Travis Microserv Mike Hammett wrote: So are you telling me that I can only run 2 or 3 radios in upper 5 GHz without stepping on myself? I was pretty sure nothing was greater than -60 or so before I made these tower changes, but silly me, I didn't bother to document what I saw. Some of those signals don't make any sense, either. Between North and East and North and South are gigantic chunks of metal to where I doubt you could even physically see a foot or two to the side of either sector. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com From: Travis Johnson Sent: Wednesday, October 01, 2008 11:21 AM To: WISPA General List Subject: Re: [WISPA] Bad radio? This really looks like you are causing yourself all kinds of interference. Using the channels you listed: North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 I'm sure they are stepping on each other. The signal doesn't just drop completely off on the edges. I assume 20mhz wide channels (because you didn't specify), meaning North and South edges are right next to each other, but in the same case, on the same card, I'm sure you are stepping on yourself. Same with the East and West ones. So, on the North you are actually using 5775 to 5795. While on the South you are 5795 to 5815. I can tell you right now that even having 10mhz between channel edges isn't going to be enough... you will need 20-40mhz of spacing between the edges of the band to keep from interfering with yourself. Travis Microserv Mike Hammett wrote: The TX is currently set to default in MT 2.9.51, whatever dB that turns out to be. I was going to tune it once the system stabilized. Maybe 20' of LMR-400. No lightning arrestors. The other sectors on the tower sat there for years without them and no damage, so when I upgraded and had the wrong pigtails to use the existing arrestors, I removed them. No amps, just the XR5s... I'd never use an amp. Roughly 16 dBi gain for all 4 sectors pointed North, South, East, and West. They all have roughly 5 degrees of downtilt. No PoE, the MT system is a PC with a 4 slot mPCI adapter. The only other cat 5 is going to an Orthogon Gemini (and it's PoE). This is pointed east and is on the opposite side of the grain leg from the troubled sector. Nothing else within 1 mile that uses upper 5 GHz. Well, there's CPE in unknown bands, but they didn't affect the previous radio\antenna combo. All customers use 19 or 24 dBi RooTennas, depending on distance from tower. I believe everyone is between -60 and -75. For some reason, some of these sectors are much louder than they were previously, though I don't have any documentation as to who saw what before. What do band-pass filters cost and where can I get them? Maybe I ought to invest in some of those. North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 North = 1 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -49 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -37 00156D501709 AB R 00:15:6D:64:0B:59 ICS2 5ghz 5805 -43 00156D640B59 AB RN 00:0C:42:05:51:B7 Walter 5ghz 5260 -71 000C420551B7 South = 2 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -51 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -38 00156D501709 AB R 00:15:6D:64:0B:55 ICS1 5ghz 5785 -29 00156D640B55 East = 4 === Will run a scan during non-peak times. West = 3
Re: [WISPA] Bad radio?
I never put more than 2 radio cards in the same box, and then only if they are at least 40MHz apart in the 5GHz band. In the 2.4 GHz band, I simply don't put 2 radios in the same box. Also, the XR's draw more power than the SR cards. Does your 4 card adapter have enough headroom on the power side? Travis Johnson wrote: Didn't you just "upgrade" to the XR series cards? They put out a lot more power, thus creating issues inside the box. You could turn the power way down or even turn 1 or 2 cards off and see if that fixes the issues. Travis Microserv Mike Hammett wrote: I have a genuine interest to know what's going on here. Why wasn't this an issue before and what effect does this have on a radio no longer transmitting as powerfully? I could see it going deaf, but it hears just fine. I'm thinking about ditching the PC and mPCI - PCI adapter and going with 4x RB411AHs instead, providing some increased radio separation. Depending on how I do it, there may be sheet metal between each RB411AH as well. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: "Travis Johnson" [EMAIL PROTECTED] Sent: Wednesday, October 01, 2008 12:25 PM To: "WISPA General List" wireless@wispa.org Subject: Re: [WISPA] Bad radio? The noise is coming from the actual wireless cards being so close to each other. Not all of the signal is going out the cable... and with cards stacked within inches from each other, more noise bleeds over. Travis Microserv Mike Hammett wrote: So are you telling me that I can only run 2 or 3 radios in upper 5 GHz without stepping on myself? I was pretty sure nothing was greater than -60 or so before I made these tower changes, but silly me, I didn't bother to document what I saw. Some of those signals don't make any sense, either. Between North and East and North and South are gigantic chunks of metal to where I doubt you could even physically see a foot or two to the side of either sector. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com From: Travis Johnson Sent: Wednesday, October 01, 2008 11:21 AM To: WISPA General List Subject: Re: [WISPA] Bad radio? This really looks like you are causing yourself all kinds of interference. Using the channels you listed: North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 I'm sure they are stepping on each other. The signal doesn't just drop completely off on the edges. I assume 20mhz wide channels (because you didn't specify), meaning North and South edges are right next to each other, but in the same case, on the same card, I'm sure you are stepping on yourself. Same with the East and West ones. So, on the North you are actually using 5775 to 5795. While on the South you are 5795 to 5815. I can tell you right now that even having 10mhz between channel edges isn't going to be enough... you will need 20-40mhz of spacing between the edges of the band to keep from interfering with yourself. Travis Microserv Mike Hammett wrote: The TX is currently set to default in MT 2.9.51, whatever dB that turns out to be. I was going to tune it once the system stabilized. Maybe 20' of LMR-400. No lightning arrestors. The other sectors on the tower sat there for years without them and no damage, so when I upgraded and had the wrong pigtails to use the existing arrestors, I removed them. No amps, just the XR5s... I'd never use an amp. Roughly 16 dBi gain for all 4 sectors pointed North, South, East, and West. They all have roughly 5 degrees of downtilt. No PoE, the MT system is a PC with a 4 slot mPCI adapter. The only other cat 5 is going to an Orthogon Gemini (and it's PoE). This is pointed east and is on the opposite side of the grain leg from the troubled sector. Nothing else within 1 mile that uses upper 5 GHz. Well, there's CPE in unknown bands, but they didn't affect the previous radio\antenna combo. All customers use 19 or 24 dBi RooTennas, depending on distance from tower. I believe everyone is between -60 and -75. For some reason, some of these sectors are much louder than they were previously, though I don't have any documentation as to who saw what before. What do band-pass filters cost and where can I get them? Maybe I ought to invest in some of those. North, ICS1 = 5785 South, ICS2 = 5805 East, ICS4 = 5765 West, ICS3 = 5745 North = 1 === Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -49 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -37 00156D501709 AB R 00:15:6D:64:0B:59 ICS2 5ghz 5805 -43 00156D640B59 AB RN 00:0C:42:05:51:B7 Walter 5ghz 5260 -71 000C420551B7 South = 2 === Flags: A - active, B - bss, P - privacy, R -
Re: [WISPA] Bad radio?
It would be interesting to switch the freqs around and see what happens. It's probably a bad radio or cable though. -RickG On Sun, Sep 28, 2008 at 1:06 PM, Mike Hammett [EMAIL PROTECTED] wrote: Does this seem like the radio isn't loud enough? I setup the test CPE on the TV tower at my house and pointed it at the tower. Radio Mobile reports the azimuth as 250 degrees, so well within the south sector's coverage and only at a distance of 230'. I am well below the vertical beamwidth of the sectors, explaining the relatively low signals, but ICS2 is horrible. 1 = North, 2 = South, 3 = West, 4 = East. It makes no sense that South is that much worse signal wise than the others, especially considering that it should be on the South sector anyway. 3 and 4 are SR5s while 1 and 2 are XR5s. I just replaced the towers with the XR5s. [EMAIL PROTECTED] Test Platform] /interface wireless scan wlan1 Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG NF SNR RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -72 -99 27 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -68 -99 31 00156D501709 AB R 00:15:6D:64:0B:59 ICS1 5ghz 5785 -77 -99 22 00156D640B59 AB R 00:15:6D:64:0B:55 ICS2 5ghz 5825 -85 -99 14 00156D640B55 Here is a listing of the signals when connected: [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5785MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS1 bssid: 00:15:6D:64:0B:59 radio-name: 00156D640B59 signal-strength: -77dBm tx-signal-strength: -74dBm noise-floor: -107dBm signal-to-noise: 30dB tx-ccq: 58% p-throughput: 5481 overall-tx-ccq: 58% authenticated-clients: 1 current-ack-timeout: 28 wds-link: no nstreme: no framing-mode: none routeros-version: 2.9.51 last-ip: 10.10.1.1 802.1x-port-enabled: yes compression: no current-tx-powers: 6Mbps:24(24),9Mbps:24(24),12Mbps:24(24),18Mbps:24(24),24Mbps:24(24),36Mbps:22(22),48Mbps:20(20),54Mbps:19(19) notify-external-fdb: no [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5825MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS2 bssid: 00:15:6D:64:0B:55 radio-name: 00156D640B55 signal-strength: -86dBm tx-signal-strength: -76dBm noise-floor: -107dBm signal-to-noise: 21dB tx-ccq: 59% p-throughput: 5535 overall-tx-ccq: 58% authenticated-clients: 1 current-ack-timeout: 167 wds-link: no nstreme: no framing-mode: none routeros-version: 2.9.51 802.1x-port-enabled: yes compression: no current-tx-powers: 6Mbps:24(24),9Mbps:24(24),12Mbps:24(24),18Mbps:24(24),24Mbps:24(24),36Mbps:22(22),48Mbps:20(20),54Mbps:19(19) notify-external-fdb: no -- [Q quit|D dump|C-z pause] [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5745MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS3 bssid: 00:15:6D:50:16:C6 radio-name: 00156D5016C6 signal-strength: -74dBm tx-signal-strength: -70dBm noise-floor: -106dBm signal-to-noise: 32dB tx-ccq: 59% p-throughput: 5518 overall-tx-ccq: 59% authenticated-clients: 1 current-ack-timeout: 28 wds-link: no nstreme: no framing-mode: none routeros-version: 2.9.51 last-ip: 10.10.3.5 802.1x-port-enabled: yes compression: no current-tx-powers: 6Mbps:24(24),9Mbps:24(24),12Mbps:24(24),18Mbps:24(24),24Mbps:24(24),36Mbps:22(22),48Mbps:20(20),54Mbps:19(19) notify-external-fdb: no [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5765MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS4 bssid: 00:15:6D:50:17:09 radio-name: 00156D501709 signal-strength: -69dBm tx-signal-strength: -69dBm noise-floor: -106dBm
Re: [WISPA] Bad radio?
I changed the freq a bit and didn't see any significant change. I climbed and swapped the pigtails with North and changed all the settings so they were fully swapped. Everyone on the south sector (which did have the questionable radio) is now happy. Now the North sector is showing the same symptoms on my test CPE. It's a good thing everyone except two are on East and West until I get this sorted out. Time to contact Streakwave about a bad radio... -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: RickG [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 1:05 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? It would be interesting to switch the freqs around and see what happens. It's probably a bad radio or cable though. -RickG On Sun, Sep 28, 2008 at 1:06 PM, Mike Hammett [EMAIL PROTECTED] wrote: Does this seem like the radio isn't loud enough? I setup the test CPE on the TV tower at my house and pointed it at the tower. Radio Mobile reports the azimuth as 250 degrees, so well within the south sector's coverage and only at a distance of 230'. I am well below the vertical beamwidth of the sectors, explaining the relatively low signals, but ICS2 is horrible. 1 = North, 2 = South, 3 = West, 4 = East. It makes no sense that South is that much worse signal wise than the others, especially considering that it should be on the South sector anyway. 3 and 4 are SR5s while 1 and 2 are XR5s. I just replaced the towers with the XR5s. [EMAIL PROTECTED] Test Platform] /interface wireless scan wlan1 Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG NF SNR RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -72 -99 27 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -68 -99 31 00156D501709 AB R 00:15:6D:64:0B:59 ICS1 5ghz 5785 -77 -99 22 00156D640B59 AB R 00:15:6D:64:0B:55 ICS2 5ghz 5825 -85 -99 14 00156D640B55 Here is a listing of the signals when connected: [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5785MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS1 bssid: 00:15:6D:64:0B:59 radio-name: 00156D640B59 signal-strength: -77dBm tx-signal-strength: -74dBm noise-floor: -107dBm signal-to-noise: 30dB tx-ccq: 58% p-throughput: 5481 overall-tx-ccq: 58% authenticated-clients: 1 current-ack-timeout: 28 wds-link: no nstreme: no framing-mode: none routeros-version: 2.9.51 last-ip: 10.10.1.1 802.1x-port-enabled: yes compression: no current-tx-powers: 6Mbps:24(24),9Mbps:24(24),12Mbps:24(24),18Mbps:24(24),24Mbps:24(24),36Mbps:22(22),48Mbps:20(20),54Mbps:19(19) notify-external-fdb: no [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5825MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS2 bssid: 00:15:6D:64:0B:55 radio-name: 00156D640B55 signal-strength: -86dBm tx-signal-strength: -76dBm noise-floor: -107dBm signal-to-noise: 21dB tx-ccq: 59% p-throughput: 5535 overall-tx-ccq: 58% authenticated-clients: 1 current-ack-timeout: 167 wds-link: no nstreme: no framing-mode: none routeros-version: 2.9.51 802.1x-port-enabled: yes compression: no current-tx-powers: 6Mbps:24(24),9Mbps:24(24),12Mbps:24(24),18Mbps:24(24),24Mbps:24(24),36Mbps:22(22),48Mbps:20(20),54Mbps:19(19) notify-external-fdb: no -- [Q quit|D dump|C-z pause] [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5745MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS3 bssid: 00:15:6D:50:16:C6 radio-name: 00156D5016C6 signal-strength: -74dBm tx-signal-strength: -70dBm noise-floor: -106dBm signal-to-noise: 32dB tx-ccq: 59% p-throughput: 5518 overall-tx-ccq: 59% authenticated-clients: 1 current-ack-timeout: 28 wds-link: no nstreme: no framing-mode
Re: [WISPA] Bad radio?
You don't keep spare radio cards in stock? That's probably something you should consider. Travis Microserv Mike Hammett wrote: I changed the freq a bit and didn't see any significant change. I climbed and swapped the pigtails with North and changed all the settings so they were fully swapped. Everyone on the south sector (which did have the questionable radio) is now happy. Now the North sector is showing the same symptoms on my test CPE. It's a good thing everyone except two are on East and West until I get this sorted out. Time to contact Streakwave about a bad radio... -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: "RickG" [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 1:05 PM To: "WISPA General List" wireless@wispa.org Subject: Re: [WISPA] Bad radio? It would be interesting to switch the freqs around and see what happens. It's probably a bad radio or cable though. -RickG On Sun, Sep 28, 2008 at 1:06 PM, Mike Hammett [EMAIL PROTECTED] wrote: Does this seem like the radio isn't loud enough? I setup the test CPE on the TV tower at my house and pointed it at the tower. Radio Mobile reports the azimuth as 250 degrees, so well within the south sector's coverage and only at a distance of 230'. I am well below the vertical beamwidth of the sectors, explaining the relatively low signals, but ICS2 is horrible. 1 = North, 2 = South, 3 = West, 4 = East. It makes no sense that South is that much worse signal wise than the others, especially considering that it should be on the South sector anyway. 3 and 4 are SR5s while 1 and 2 are XR5s. I just replaced the towers with the XR5s. [EMAIL PROTECTED] Test Platform] /interface wireless scan wlan1 Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG NF SNR RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -72 -99 27 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -68 -99 31 00156D501709 AB R 00:15:6D:64:0B:59 ICS1 5ghz 5785 -77 -99 22 00156D640B59 AB R 00:15:6D:64:0B:55 ICS2 5ghz 5825 -85 -99 14 00156D640B55 Here is a listing of the signals when connected: [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5785MHz tx-rate: "6Mbps" rx-rate: "6Mbps" ssid: "ICS1" bssid: 00:15:6D:64:0B:59 radio-name: "00156D640B59" signal-strength: -77dBm tx-signal-strength: -74dBm noise-floor: -107dBm signal-to-noise: 30dB tx-ccq: 58% p-throughput: 5481 overall-tx-ccq: 58% authenticated-clients: 1 current-ack-timeout: 28 wds-link: no nstreme: no framing-mode: none routeros-version: "2.9.51" last-ip: 10.10.1.1 802.1x-port-enabled: yes compression: no current-tx-powers: 6Mbps:24(24),9Mbps:24(24),12Mbps:24(24),18Mbps:24(24),24Mbps:24(24),36Mbps:22(22),48Mbps:20(20),54Mbps:19(19) notify-external-fdb: no [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5825MHz tx-rate: "6Mbps" rx-rate: "6Mbps" ssid: "ICS2" bssid: 00:15:6D:64:0B:55 radio-name: "00156D640B55" signal-strength: -86dBm tx-signal-strength: -76dBm noise-floor: -107dBm signal-to-noise: 21dB tx-ccq: 59% p-throughput: 5535 overall-tx-ccq: 58% authenticated-clients: 1 current-ack-timeout: 167 wds-link: no nstreme: no framing-mode: none routeros-version: "2.9.51" 802.1x-port-enabled: yes compression: no current-tx-powers: 6Mbps:24(24),9Mbps:24(24),12Mbps:24(24),18Mbps:24(24),24Mbps:24(24),36Mbps:22(22),48Mbps:20(20),54Mbps:19(19) notify-external-fdb: no -- [Q quit|D dump|C-z pause] [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5745MHz tx-rate: "6Mbps" rx-rate: "6Mbps" ssid: "ICS3" bssid: 00:15:6D:50:16:C6 radio-name: "00156D5016C6" signal-strength: -74dBm tx-signal-strength: -70dBm noise-floor: -1
Re: [WISPA] Bad radio?
Oddly enough, I have spares for the clients, but not for the towers... never had a bad tower radio before. This one could be classified as DOA since it hasn't even been up there a week before it started doing this. There was only 1 wireless client (is a repeater) total between North and South sectors... I just replaced the PacWireless sectors, SR5s, and u.fl pigtails with MTI sectors, XR5s, and MMCX pigtails. Just didn't have the coverage I was experiencing with the East and West sectors, which have significantly more people. -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com From: Travis Johnson Sent: Sunday, September 28, 2008 3:02 PM To: WISPA General List Subject: Re: [WISPA] Bad radio? You don't keep spare radio cards in stock? That's probably something you should consider. Travis Microserv Mike Hammett wrote: I changed the freq a bit and didn't see any significant change. I climbed and swapped the pigtails with North and changed all the settings so they were fully swapped. Everyone on the south sector (which did have the questionable radio) is now happy. Now the North sector is showing the same symptoms on my test CPE. It's a good thing everyone except two are on East and West until I get this sorted out. Time to contact Streakwave about a bad radio... -- Mike Hammett Intelligent Computing Solutions http://www.ics-il.com -- From: RickG [EMAIL PROTECTED] Sent: Sunday, September 28, 2008 1:05 PM To: WISPA General List wireless@wispa.org Subject: Re: [WISPA] Bad radio? It would be interesting to switch the freqs around and see what happens. It's probably a bad radio or cable though. -RickG On Sun, Sep 28, 2008 at 1:06 PM, Mike Hammett [EMAIL PROTECTED] wrote: Does this seem like the radio isn't loud enough? I setup the test CPE on the TV tower at my house and pointed it at the tower. Radio Mobile reports the azimuth as 250 degrees, so well within the south sector's coverage and only at a distance of 230'. I am well below the vertical beamwidth of the sectors, explaining the relatively low signals, but ICS2 is horrible. 1 = North, 2 = South, 3 = West, 4 = East. It makes no sense that South is that much worse signal wise than the others, especially considering that it should be on the South sector anyway. 3 and 4 are SR5s while 1 and 2 are XR5s. I just replaced the towers with the XR5s. [EMAIL PROTECTED] Test Platform] /interface wireless scan wlan1 Flags: A - active, B - bss, P - privacy, R - routeros-network, N - nstreme ADDRESS SSID BAND FREQ SIG NF SNR RADIO-NAME AB R 00:15:6D:50:16:C6 ICS3 5ghz 5745 -72 -99 27 00156D5016C6 AB R 00:15:6D:50:17:09 ICS4 5ghz 5765 -68 -99 31 00156D501709 AB R 00:15:6D:64:0B:59 ICS1 5ghz 5785 -77 -99 22 00156D640B59 AB R 00:15:6D:64:0B:55 ICS2 5ghz 5825 -85 -99 14 00156D640B55 Here is a listing of the signals when connected: [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5785MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS1 bssid: 00:15:6D:64:0B:59 radio-name: 00156D640B59 signal-strength: -77dBm tx-signal-strength: -74dBm noise-floor: -107dBm signal-to-noise: 30dB tx-ccq: 58% p-throughput: 5481 overall-tx-ccq: 58% authenticated-clients: 1 current-ack-timeout: 28 wds-link: no nstreme: no framing-mode: none routeros-version: 2.9.51 last-ip: 10.10.1.1 802.1x-port-enabled: yes compression: no current-tx-powers: 6Mbps:24(24),9Mbps:24(24),12Mbps:24(24),18Mbps:24(24),24Mbps:24(24),36Mbps:22(22),48Mbps:20(20),54Mbps:19(19) notify-external-fdb: no [EMAIL PROTECTED] Test Platform] /interface wireless monitor wlan1 status: connected-to-ess band: 5ghz frequency: 5825MHz tx-rate: 6Mbps rx-rate: 6Mbps ssid: ICS2 bssid: 00:15:6D:64:0B:55 radio-name: 00156D640B55 signal-strength: -86dBm tx-signal-strength: -76dBm noise-floor: -107dBm signal-to-noise: 21dB tx-ccq: 59% p-throughput: 5535 overall-tx-ccq: 58% authenticated-clients: 1 current-ack-timeout: 167 wds-link: no nstreme: no framing-mode: none routeros-version: 2.9.51 802.1x-port-enabled: yes compression: no current-tx-powers: 6Mbps:24(24),9Mbps:24(24),12Mbps:24(24),18Mbps:24(24),24Mbps:24(24
