Bullet M5 Disappoints for Long 802.11a Links

[Ron]

Careful readers may notice that the Bullet 5M datasheet does not list RX sensitivity for 802.11a rates below 24 Mbps. It lists -83 dBm for 24 Mbps. I, and probably others, assumed that it does work at below -83 dBm at lower rates.

Turns out:

***The Bullet 5M does not work at all with signal levels below -83 in 802.11a mode.***

Other posts suggest the same thing but I had to test it out for myself to be sure.
I disconnect a Mikrotik R5H from a dish, where the R5H was getting -89 RX and 1-2 Mbps TCP, and connected a Bullet 5M directly to the dish (no cable or pigtail to cause signal loss) and did a scan, over and over again, the bullet can't even see the beacons from the AP (much less connect)!

Mike & Co, is this a hardware limitation or something that will be fixed in the firmware soon?

[cameron]

Ron,

I am seeing the exact same issue as you with my Bullet M5, see my post regarding M5 signal Performance.

Regards

Cameron

[UBNT-Robert]

Hi Guys --

We do not officially support 802.11a/b/g mode with M series products yet. Under certain conditions, they might work OK, but we cannot guarantee anything right now.

We will provide official 802.11a/b/g support with AirOS V5.1 which is still 1-2 months out.

I know this is not a great answer, but 802.11n and AirMax support is our primary focus right now.

If you need solutions for use with 802.11a/b/g, I would stick with our 802.11a/b/g product line. Even when we do get to V5.1, M series will not provide a performance benefit over the 802.11a/b/g products in 802.11a/b/g application use.

Ron -- comment regarding RF sensitivity performance is not accurate. We will update datasheet to provide 6Mbps sensitivity; it should be -94dBm.

[drwho17]

Well, I'm happy to hear that, thanks for clearing up an issue that has been bewildering to us with our testing of Bullet2M's. We will stop using them as clients until the issue is resolved.

[Ron]

If you need solutions for use with 802.11a/b/g, I would stick with our 802.11a/b/g product line. Even when we do get to V5.1, M series will not provide a performance benefit over the 802.11a/b/g products in 802.11a/b/g application use.

Ron -- comment regarding RF sensitivity performance is not accurate. We will update datasheet to provide 6Mbps sensitivity; it should be -94dBm.

Robert, thanks for dropping by. I know you are busy.

Your 802.11a product line is out of stock. No one seems to have Bullet 5HP. I was hoping Bullet 5M would be at least as good.

Are you saying we can't see 802.11a APs not because the RX sensitivity is poor but because the firmware is not ready yet?

[WHT]

connected a Bullet 5M directly to the dish (no cable or pigtail to cause signal loss) and did a scan, over and over again, the bullet can't even see the beacons from the AP (much less connect)!
Ron....

Can you test your Bullets on the bench. I tested a dozen of mine over the weekend and half of them had a bad RF section that added an additional 10 to 14 dB loss.

I suspected for several months there was a problem with some the Bullet M5 radios, but it was only last week that I finally had the chance to lay them out on the bench and narrow it down. I haven't reported my results to Mike yet, so I may be the only one that knows of this.

Here's my testing setup:

1) I calibrated a Bullet M5 to a known working Bullet as the AP with a known working Nano as the CPE.

2) I used that Bullet M5 as an AP and tested a dozen of my Bullet M5 as CPEs.

3) Most of the Bullet M5 CPE radios showed around -49 dBm received signal level from the AP and the AP reported about -49 dBm received signal level from the CPE (which would be the transmitted output of the CPE)

4) A few radios showed only -56 to -63 dBm received signal level from the AP and the AP reported about -56 to -63 dBm received signal level from the CPE (which would be the transmitted output of the CPE)

While points 3) and 4) together may indicate a simultaneous failure of both the transmitter RF output *and* the receiver front end, but I suspect its more likely a connection from the radio module to the antenna connector.

Several of the older Bullet M5 radios has a loose N connector center pin that wasn't soldered to the circuit board. But those radios had ZERO TX output and ZERO RX levels, not a mere 10 to 14 dB loss.

[cameron]

WHT What your saying is what my BM5 appears to have, its deaf. I hope Mike can shed some light on this. If I get a chance I'll fire my Bulelt M5 up along side my Mikrotik setup and compare results.

[WHT]

WHT What your saying is what my BM5 appears to have, its deaf.Not really...

1) I'm saying the first forty or fifty or so pre-distribution release radios that I've gotten over the past several months have a problem with some of the radios.

2) Some had a poorly soldered N connector pin and the radios were completely deaf and dumb. Mike said that was fixed with the next batch he sent me and Rory. So I don't think you'll be experienced that problem.

3) Some of the radios have an unidentified problem that reduced TX/RX performance (but not completely deaf and dumb). Since there has never been any reports of this until the past week or so, that might be what you are experiencing.

I'm only suggestion that you try a similar test like I did to see if the distribution radios have the same problem.

If the problem radios have a problem, it would give a similar situation as if you tested a Nano 5 with an external antenna and find a 5 dB loss, but with the Bullet M5 test you may see a 10 to 14 dB loss.

[spainuser]

Hi WHT, can you put photos of the printed circuit, bad and good??


I have similar problem with BULLET2M, bad sensitivity.
Have you noticed any problem with BULLET2M ??


Thanks,


Juan

[WHT]

I never used any 2.4 Gig UBNT gear.

I did take apart one of the Bullet M5 radios, but stupid me...I twisted the plastic case instead of the metal ring and ripped the pin off the circuit board. So its gonna be hard to tell just what was wrong.

My megapixel camera battery is dead...dead...dead. And my new $650 cellphone (Doncha wish you had a Palm Treo Pro..Doncha, doncha?) can't get a good close in shot.

But I just opened up two radios and I cannot see any visual difference in the soldering, nor anything else.

[JustJoe]

Hi Guys --

We do not officially support 802.11a/b/g mode with M series products yet. Under certain conditions, they might work OK, but we cannot guarantee anything right now.

We will provide official 802.11a/b/g support with AirOS V5.1 which is still 1-2 months out.

I know this is not a great answer, but 802.11n and AirMax support is our primary focus right now.

If you need solutions for use with 802.11a/b/g, I would stick with our 802.11a/b/g product line. Even when we do get to V5.1, M series will not provide a performance benefit over the 802.11a/b/g products in 802.11a/b/g application use.

Ron -- comment regarding RF sensitivity performance is not accurate. We will update datasheet to provide 6Mbps sensitivity; it should be -94dBm.

Robert,

UBNT is really kinda screwing with us. When I see a data sheet for B5M that that gives .11a specs, I would have thought you supported it *now*.

I, like many others, waited with pre-orders for B5HP that will apparently never materialize based on your sales newsletters .

I, like others, thought that B5M would be an acceptable alternative to B5HP and run in .11a compatibility. So now I have them in my hands, but if I install them for .11a use and find out they work poorly, the answer from UBNT will be to come back in a couple of months.

:( :( :( :(

[Dave-D]

I just did a Bullet M5 to Bullet 5 link at 14Km
that worked fine. Maybe I'm one of the lucky
ones who happened on the configuration that
worked, but I saw no problems.

Signal strength was exceptional--much better
than I had expected, better than the other
Bullet5 on this PTMP setup.

This link uses only a 5MHz bandwidth, and only
6Mbps data, so it is truly locked down tight.
But within that limitation, it was 100% CCM.

Dave

[WHT]

Yeah....we tried all firmware releases and using ACK on both or just one side won't work.

What we *did* discover was with AP Auto-ACK, the AP timed out before the CPE had a chance to respond. So that didn't work.

When we set the CPE to Auto-ACK with a 22 mile shot, the CPE would auto set the ACK to like ummm, I think it was 1.7 miles. BUT when you started passing data though it, the ACK would go up to 21 miles. After the data stopped, it dropped down to 1.7 miles again.

[Dave-D]

Yep, I had to manually set ACK to 15km. The
auto-ACK gave me 3% CCM and effectively
no throughput. I don't see the value of an
auto-ACK setting anyway, because the
distance is known. It's not like you're
chasing down the road with an AP. Dave

[george]

Robert,

UBNT is really kinda screwing with us. When I see a data sheet for B5M that that gives .11a specs, I would have thought you supported it *now*.

I, like many others, waited with pre-orders for B5HP that will apparently never materialize based on your sales newsletters .

I, like others, thought that B5M would be an acceptable alternative to B5HP and run in .11a compatibility. So now I have them in my hands, but if I install them for .11a use and find out they work poorly, the answer from UBNT will be to come back in a couple of months.

:( :( :( :(

Frankly, with a couple of small exceptions (where's that White Paper on AirMAX Robert?) UBNT has been pretty open about the beta process that got us to this point.

You know that the primary beta was limited to two users, you know who they are, you know what kind of environments they were testing in. You know their results.

Anyone who wasn't aware that we are in the USUAL second stage beta that ALL early adopters of a new technology go through, simply wasn't paying attention.

All our pre-order (and its a big one) is about testing and developing and working with UBNT until we get a stable, usable product. We aren't rolling out anything anywhere that our business will depend on to survive until that time.

That's why my signature says 'coming soon' rather than 'sign up here'.

George

[lncommunications]

All our pre-order (and its a big one) is about testing and developing and working with UBNT until we get a stable, usable product. We aren't rolling out anything anywhere that our business will depend on to survive until that time.

That's why my signature says 'coming soon' rather than 'sign up here'.

George

Amen george, amen. I fully back that statement, you must be crazy if you were to put up any new-to-the-market device straight up and expect it to work 100%.

[JustJoe]

Frankly, with a couple of small exceptions (where's that White Paper on AirMAX Robert?) UBNT has been pretty open about the beta process that got us to this point.

You know that the primary beta was limited to two users, you know who they are, you know what kind of environments they were testing in. You know their results.

Anyone who wasn't aware that we are in the USUAL second stage beta that ALL early adopters of a new technology go through, simply wasn't paying attention.

All our pre-order (and its a big one) is about testing and developing and working with UBNT until we get a stable, usable product. We aren't rolling out anything anywhere that our business will depend on to survive until that time.

That's why my signature says 'coming soon' rather than 'sign up here'.

George

I am sorry, but your statement is pure hogwash.

I am perfectly aware that there will be *bugs* in new products. And I am perfectly willing to provide every bit of documentation for any bug I run across as I have in the past.

But don't go telling me that a feature for which the product spec sheet says it was tested, is not really "officially" supported and will not be until the next version 2 months from now. That means .11a support is *not even beta test* stage. Especially aggravating if I have to catch this little tidbit of news in the forums as they fly by from day to day.

The product boxes and sales flyers are marked AirOS V 5.0 not V 5.0 Beta.

[george]

You can whine until you're blue in the face, but the fact these products were only announced a few weeks ago would tell any reasonably experienced network operator that they were going to be an early adopter, and any reasonably experienced network operator would know what being an early adopter means. :icon_mrgreen:

You're entitled to your opinion of course, just don't expect it to be universally shared...
:ubnt_banana:

[UBNT-Mike.Ford]

Hey Guys,

I just ran a MikroTik/XR5 Ap to both a Bullet 5 and Bullet M5. The Bullet M5 was able to maintain higher through in 802.11a mode then the Bullet 5 ( due to the higher tX power). The recieve sensitivities for 802.11a rates were reletivly the same and I tested down to -90@6mbps.

Thanks,

Mike

[WHT]

I, like others, thought that B5M would be an acceptable alternative to B5HP and run in .11a compatibility. It was designed ground up to be used as part of a complete AirMax system. Forcing it work work in an 802.11a environment is your own choice. No one ever said it was supposed to be an acceptable alternative. You took it upon yourself to assume it was.

[JustJoe]

It was designed ground up to be used as part of a complete AirMax system. Forcing it work work in an 802.11a environment is your own choice. No one ever said it was supposed to be an acceptable alternative. You took it upon yourself to assume it was.

OK, let's back the truck up. When I was buying B5Ms, we weren't even being told what "AirMax" was other than it could be turned on and off.

There were *plenty* of forum posts of people asking about what would happen if B2&5Ms were forced to associate with .11a/b/g devices. The answer each time mentioned only one *penalty*, which was falling back to the a/b/g throughput levels and losing the .11n aggregation efficiencies. To me (and apparently others) that implied that they were intended to inter-operate with only that limitation. Neither you nor any UBNT people said don't do it because it is not supported until the next version.

If I go to buy an SUV and it is rated with a certain towing capacity in the *specs*, I assume that means it comes with an engine and a transmission.

[WHT]

UPDATE......

Mike just called me and verified my PRE-RELEASE BETA units had problems.

These were from the original batch of radios.

The low signal levels were caused by a problem with the N connector soldered to the circuit board.

The dead radio was caused by a shorted capacitor.

ALL OF THE PROBLEMS HAVE BEEN RESOLVED IN THE PRODUCTION UNITS CURRENTLY AVAILABLE FROM DISTRIBUTORS.



Can you test your Bullets on the bench. I tested a dozen of mine over the weekend and half of them had a bad RF section that added an additional 10 to 14 dB loss.

I suspected for several months there was a problem with some the Bullet M5 radios, but it was only last week that I finally had the chance to lay them out on the bench and narrow it down. I haven't reported my results to Mike yet, so I may be the only one that knows of this.

Here's my testing setup:

1) I calibrated a Bullet M5 to a known working Bullet as the AP with a known working Nano as the CPE.

2) I used that Bullet M5 as an AP and tested a dozen of my Bullet M5 as CPEs.

3) Most of the Bullet M5 CPE radios showed around -49 dBm received signal level from the AP and the AP reported about -49 dBm received signal level from the CPE (which would be the transmitted output of the CPE)

4) A few radios showed only -56 to -63 dBm received signal level from the AP and the AP reported about -56 to -63 dBm received signal level from the CPE (which would be the transmitted output of the CPE)

While points 3) and 4) together may indicate a simultaneous failure of both the transmitter RF output *and* the receiver front end, but I suspect its more likely a connection from the radio module to the antenna connector.

Several of the older Bullet M5 radios has a loose N connector center pin that wasn't soldered to the circuit board. But those radios had ZERO TX output and ZERO RX levels, not a mere 10 to 14 dB loss.

[UBNT-Mike.Ford]

OK, let's back the truck up. When I was buying B5Ms, we weren't even being told what "AirMax" was other than it could be turned on and off.

There were *plenty* of forum posts of people asking about what would happen if B2&5Ms were forced to associate with .11a/b/g devices. The answer each time mentioned only one *penalty*, which was falling back to the a/b/g throughput levels and losing the .11n aggregation efficiencies. To me (and apparently others) that implied that they were intended to inter-operate with only that limitation. Neither you nor any UBNT people said don't do it because it is not supported until the next version.

If I go to buy an SUV and it is rated with a certain towing capacity in the *specs*, I assume that means it comes with an engine and a transmission.


Hey Guys,

I just ran a MikroTik/XR5 Ap to both a Bullet 5 and Bullet M5. The Bullet M5 was able to maintain higher through in 802.11a mode then the Bullet 5 ( due to the higher tX power). The recieve sensitivities for 802.11a rates were reletivly the same and I tested down to -90@6mbps.

Thanks,

Mike

[rconaway]

I never used any 2.4 Gig UBNT gear.

My megapixel camera battery is dead...dead...dead. And my new $650 cellphone (Doncha wish you had a Palm Treo Pro..Doncha, doncha?) can't get a good close in shot.

Boy you should see the quality of the camera in my Windows phone. It even has a close up mode.

[cameron]

I am still having issues with my bullet M5

Connected to my AP 10km away, CCQ sits at 35% low I know but not sure why so low receive it does 6mbps but as soo nas you try test send it link drops out.

I plugged my Bullet 5 in and tested CCQ 35% 5mbps both directions no problems.

I am not sure why the BM5 drops the link when you try load it up on send :S

I am also not sure why the CCQ is so low seeing I have a mikrotik and XR5 on the same mast connected to a diff Radio on the same site 10km away and it is 100% ccq but using a panel instead of a Dish.

[rconaway]

What is your signal level? What is the gain of your antennas?

[cameron]

Signal level is -74 and -98 Noise, Antenna is a 27dbi Dish bullet on 10dbm.

Other end is a 15dbi Waveguide Sector with a R5H.

Manually set the ACK to 97 for 11.4km as the link is 10.6km. It was when I first powered up the unit sitting at 96% ccq but after 10 minutes was sitting on 11% and not moving but today theire is little to no throughput but pings are >=1ms

I tried both A & N same issue.

Cheers

Cameron

[UBNT-Robert]

Hey Cameron --

For M series, we are fully concentrated on resolving the remaining 11n mode and airmax issues right now. As soon as that is on its way, we are going to complete our 802.11a/b/g compatibility testing and release AirOS 5.1 with official 802.11a/b/g support thereafter. We are moving as fast as we can.

For now, it is what it is. We have heard reports of M series working well in 802.11a/b/g mode in some environments. We have also have heard reports of issues similar to yours.

Just to repeat -- we have not completed M series 802.11a/b/g compatibility testing yet and do not have answers for any 802.11a/b/g issues with M series.

I strongly recommend to stick with our 802.11a/b/g products for 802.11a/b/g applications at least until AirOS 5.1 is released.

Robert

[cameron]

Robert,

I understand this but the problem is also there when i switch to N only on the AP using the Mikrotik unit with a R52n.

I also note that any power level lower than 15dbm on the B5M will not connect to the AP at all, it can see it but will not connect see's it with signal -77

I'll post some stats tomorrow from the link to compare & reference.

Cameron

[rconaway]

What is an R5H?

[Beninspain]

I strongly recommend to stick with our 802.11a/b/g products for 802.11a/b/g applications at least until AirOS 5.1 is released.

Robert

Are you still making and shipping these. I cant get them for love nor money. Sitting here with a pile of NanoM5s that wont work with our existing network.

What i gotta do? start shopping elsewhere? MikroTik all is forgiven

[UBNT-Mike.Ford]

Are you still making and shipping these. I cant get them for love nor money. Sitting here with a pile of NanoM5s that wont work with our existing network.

What i gotta do? start shopping elsewhere? MikroTik all is forgiven


Hello,

Yes we are still building and selling these. Also why wont your M5's work with your exisiting network? Are you usng WEP encryption?

Thanks,

Mike

[Beninspain]

Hello,

Yes we are still building and selling these. Also why wont your M5's work with your exisiting network? Are you usng WEP encryption?

Thanks,

Mike

Well you main european distibuter has had no stock for some months
And no. WPA2 PSK aes ccm. tried the new firmware but we just cannot get it to register to our APs. tried stacks of options and like most people on here, burnt hours doing so.

At 20 feet we can connect, but over longer links we can clearly see and select in the scan. Its a H/V issue as laying it on the side, we can get bettter results, but really need to be able to switch H/V. existing network consists of mainly Mikrotik APs. with r52 cards and Vertical Pol antennas. no joy connecting, seeing timeouts in the NanoM5 logs.

Main screen shows 2x2 chain (we dont want) and H/V signal strenght like -55/-96
im guessing that the problem.

At 20 ft - 2Km we have had some sucsess but in the real word its just not working for us.
would love to talk you through our issues see if we can make something work.
Feel free to contact me direct anytime ben(at)nerjasolutions.com

[cameron]

What is an R5H?

There is another Card and Antenna on the same board which is a R52N I just switch between them instead of reconfiguring the AP as people are connected to them.

[rconaway]

We had similiar issues with the Nano 2's and IBM laptops a while back that got cleared up in firmware. I believe it was a timing/compatibility issue. I'm guessing you have the same type of problem in the sense that it's definitley in the firmware. I would wait until the next version of firmware comes out before retesting. There isn't much in your control until then.

[UBNT-Mike.Ford]

Well you main european distibuter has had no stock for some months
And no. WPA2 PSK aes ccm. tried the new firmware but we just cannot get it to register to our APs. tried stacks of options and like most people on here, burnt hours doing so.

At 20 feet we can connect, but over longer links we can clearly see and select in the scan. Its a H/V issue as laying it on the side, we can get bettter results, but really need to be able to switch H/V. existing network consists of mainly Mikrotik APs. with r52 cards and Vertical Pol antennas. no joy connecting, seeing timeouts in the NanoM5 logs.

Main screen shows 2x2 chain (we dont want) and H/V signal strenght like -55/-96
im guessing that the problem.

At 20 ft - 2Km we have had some sucsess but in the real word its just not working for us.
would love to talk you through our issues see if we can make something work.
Feel free to contact me direct anytime ben(at)nerjasolutions.com


Some more points:

1. There are (2) chains for Nano/Rocket M series. With our existing firmware, associations are done on Chain 0 only when using 802.11a/b/g protocol. Since Nano Chain 0 antenna is horizontal, then it does follow a connection can only be made (with a Vertical polarized AP) when rotating Nano M 90 deg. so horizontal antenna becomes vertical. We will need to change this in firmware.

As stated before in other threads, we unfortunately have all resources now on 11n / Airmax issues first. Once we get those under control, we'll move quickly to 802.11a compatibility and AirOS 5.1 We are moving as fast as we can


2. Getting the higher MCS rates and faster speeds we have found are dependent on two criteria: the first of course is signal strength. The second is the multipath environment. Best results are when the units are placed at some distance with line of sight (or near NLOS) or in a multi-path rich environment (maybe even with walls in the way). In room testing and close range testing sometimes does not produce great results. I do not know why this is; it's beyond our understanding at this point.

[rconaway]

We didn't see this problem with the Bullet 5's, with distance ranges from inches to 15'. We were able to get maximum link quality. However, AutoACK has to be off.

[Dave-D]

You wouldn't, Rory. Bullet M5 is 1x1.
And no internal antenna..... Dave

[rconaway]

I understand. That was my point. This issue may be related to a cross-polarization or timing issue.

[JustJoe]

Hmmm, just to confirm, what if I have a B5M attached to a V-pol sector and config'ed as an AP ...

Should a number of N5M stations correctly link and run normally to that B5M ??? The answer should be yes.

(Of course that limits all the links to 1x1)

[UBNT-Mike.Ford]

Hmmm, just to confirm, what if I have a B5M attached to a V-pol sector and config'ed as an AP ...

Should a number of N5M stations correctly link and run normally to that B5M ??? The answer should be yes.

(Of course that limits all the links to 1x1)


Hello,

yes this will work without issues.

Thanks,

[WHT]

I do not know why this is; it's beyond our understanding at this point.Threshold of determination comes to mind...

Back in the late seventies, I was called in to resolve a persistent problem with a telephone company's VHF mobile phone voting receivers' problem. It took me ten minutes to change the threshold from 3 dB to 6 dB. Amazing how $2,000 of outside thinking can do to fix a problem.

[Dade_Marfi]

Hi Guys --

We do not officially support 802.11a/b/g mode with M series products yet. Under certain conditions, they might work OK, but we cannot guarantee anything right now.

We will provide official 802.11a/b/g support with AirOS V5.1 which is still 1-2 months out.

I know this is not a great answer, but 802.11n and AirMax support is our primary focus right now.

If you need solutions for use with 802.11a/b/g, I would stick with our 802.11a/b/g product line. Even when we do get to V5.1, M series will not provide a performance benefit over the 802.11a/b/g products in 802.11a/b/g application use.

Ron -- comment regarding RF sensitivity performance is not accurate. We will update datasheet to provide 6Mbps sensitivity; it should be -94dBm.

please, when will 5.1 fw ? ... I would like use nanostation M5 in network in only A mode.

[ANGEL_SANCHEZ]

Hi Guys --

We do not officially support 802.11a/b/g mode with M series products yet. Under certain conditions, they might work OK, but we cannot guarantee anything right now.

We will provide official 802.11a/b/g support with AirOS V5.1 which is still 1-2 months out.

I know this is not a great answer, but 802.11n and AirMax support is our primary focus right now.

If you need solutions for use with 802.11a/b/g, I would stick with our 802.11a/b/g product line. Even when we do get to V5.1, M series will not provide a performance benefit over the 802.11a/b/g products in 802.11a/b/g application use.

Ron -- comment regarding RF sensitivity performance is not accurate. We will update datasheet to provide 6Mbps sensitivity; it should be -94dBm.

So, If I need full performance We have to use B5M to B5M?

[UBNT-Mike.Ford]

Hello,

Yes AirMax was always intended to be M series to M series to achieve the maximum performance from your links.

Thanks,

Mike

[hao]

Hm... yesterday(17Nov) I did my pre-testing with bm5hp 10.6km away(actual) (+30dBi grid Ant)
CCQ sits at 35%~80%
signal level -68~-70
datarate jumping like a kiddo 48, 54....108...120
tested with Btest, 2x PCRouter mikrotik get throughput hardly 1/2(half) mega~8mbps max...:icon_question:
playing with CW, fq to max matching load w/ ant.. result a poor outcome!

seem the promising receive sensitivity doesnot showing its character.
also setting TX to its max(25dB), look a distorted output too

ask the question 3 times, why?... why?, why?... answer is poor quality inspection... this is what we pay for the low pricing

also these why I had been suggesting in compliant test(AP mode) they should have exact same full blown ch list as to be list in scan list(station mode)

likely these bm5hp does prone to interference well than R52H / XR5, low pricing doesnot gimme a good quality that I am looking for. if UBNT prod/ sell at higher cost but the quality is none to the send best.. I will still buy it...
we are believing its hyped.... the first class true carrier

I am still having issues with my bullet M5

Connected to my AP 10km away, CCQ sits at 35% low I know but not sure why so low receive it does 6mbps but as soo nas you try test send it link drops out.

I plugged my Bullet 5 in and tested CCQ 35% 5mbps both directions no problems.

I am not sure why the BM5 drops the link when you try load it up on send :S

I am also not sure why the CCQ is so low seeing I have a mikrotik and XR5 on the same mast connected to a diff Radio on the same site 10km away and it is 100% ccq but using a panel instead of a Dish.

[rconaway]

Hm... yesterday(17Nov) I did my pre-testing with bm5hp 10.6km away(actual) (+30dBi grid Ant)
CCQ sits at 35%~80%
signal level -68~-70
datarate jumping like a kiddo 48, 54....108...120
tested with Btest, 2x PCRouter mikrotik get throughput hardly 1/2(half) mega~8mbps max...:icon_question:
playing with CW, fq to max matching load w/ ant.. result a poor outcome!

seem the promising receive sensitivity doesnot showing its character.
also setting TX to its max(25dB), look a distorted output too

ask the question 3 times, why?... why?, why?... answer is poor quality inspection... this is what we pay for the low pricing

also these why I had been suggesting in compliant test(AP mode) they should have exact same full blown ch list as to be list in scan list(station mode)

likely these bm5hp does prone to interference well than R52H / XR5, low pricing doesnot gimme a good quality that I am looking for. if UBNT prod/ sell at higher cost but the quality is none to the send best.. I will still buy it...
we are believing its hyped.... the first class true carrier

Hao, are you running Bullet 5M to Bullet 5M?

[WHT]

Golly gee, guys....

Lets use my 22 mile link as a base line. It has been working flawlessly since April (except for two weeks where I had to use a Mikrotik system because one of my pre-distribution beta release Bullets had an N connector problem). I'm running 28 dBi wire grid antennas with pigtails at both ends.

I have -69 down to -72 dBm signal levels both way (however it dropped to -73 when it dropped to near freezing last night), and CCQ is solid at 98 to 100%. Usually I'm seeing 65 Mbps throughput, but on occasion its only 58.5 - so I'll only claim its reliable for 58.5. Several very heavy thunderstorms passed through my link and it never dropped the signal or CCQ.

Other than drop my TPO down by one db, set my network settings, and used manual ACK for 16 miles I did nothing else to optimize the link.

Now...you're running a link one fourth the distance and not anywhere getting the results I'm getting. Ergo...It has to be something not right at your deployment. Try playing with the ACK in quarter mile increments, that has solved some issues.
http://www.ubnt.com/forum/showthread.php?t=14741

"Answer is poor quality inspection" would apply to the UBNT motherboard, not likely the Atheros RF engine. And I don't think (within the scope of the discussion) UBNT has any Q.A. issues.

[rconaway]

I kind of have to agree here. I have 4 different setups that have been running for months, 24x7, with no issues. 2 are PTP and 2 are PTMP. Athough I will finish more tomorrow and I'm sure the firmware is not quite ready for primetime, it has been working fine for these links.

[hao]

2 WHT, thank you I will looking/ learning into tuning ACK range

its bullet M5 HP... bro rconaway, do you prod/ sell bullet m5 only?...:icon_lol:

ouch... but where do you live... here.... using ptp is very common thus you can imagine the interference over the 5GHz spectrum, may be its due to interference its unable to get the right spec as much I can hope.
2nd.. the way of pointing/aiming... gotta ^_^ into these in the field location at 13rd floor (- zzz -) 40th floor (ptp)

anyway... the point is, I am/ we are hoping UBNT always generate the very extreme good Rx sensitivity but also excellent Tx performance quality... :icon_mrgreen:

question.... when enabling airmax feature, I am getting quality ~70% vs cap(forgotten) how is it impact to is throughput performance correlate w/ signal/ ccq? enabling/ disable ?% throughput

PS: how about a little sneaker to other player http://www.tp-link.com/simulator/TL-WA501G/userRpm/index.htm
I like the Bridge feature: wireless/wireless mode/Bridge

[rconaway]

I understand it's a Bullet 5M. I had several in beta testing and they are still running. I'm not sure of your question concerning if that is all I sell. I have hundreds of radios up across the whole product line and I'm still installing several per week.

[SaltyTiger]

I kind of have to agree here. I have 4 different setups that have been running for months, 24x7, with no issues. 2 are PTP and 2 are PTMP. Athough I will finish more tomorrow and I'm sure the firmware is not quite ready for primetime, it has been working fine for these links.

This is to backup rconaway posting, we have some links up running for a while now and very stable at the moment and even more so with the 5.02 firmware.

[rconaway]

Man I'm dense. I sell motorola, bridgewave, skypilot, etc.... I deal with very expensive and very inexpensive equipment. Believe me when I tell you I make far more money on high-end equipment. However, I find the best product for the project and cost/benefit analysis is part of that. If the radios didn't work, I wouldn't install them and I just committed them to a multimillion dollar project.

[hao]

good to you guys, especially bro rconaway.... thus its good for us too... :icon_lol: