Good points. Likely I will not touch it again unless it breaks. I'll
try to get a make on the "good" cable, but I know the cheap stuff I ran
yesterday is the arc wireless shielded, flooded, drainwire, I picked up
6 months ago when it was $69 a roll.
Brian
Tom DeReggi wrote:
So the question that arises, is why did that fix it?
I see two possibilities....
1) Poor quality cable or cable shields. (Loss running Ethernet data
parallel to power)
2) Sharing a CAT5 jack on the 532 main board for Power and Data.
Travis previously talked about the horrid RF interferrence that the
532 board generated when using 48V, due to the 532 onboard power
converter/supply. I'm wondering if the distortion/loss was at the
board itself apposed to cable?
It would have been interesting to know, if you used one cable for both
data and power, but terminated the data pairs to a different Ethernet
port instead of the POE port used for power.
What also would have been interesting would have been to know wether a
18V power supply would have worked on a shared single cable.
Different ethernet chipsets do have different characteristics and
ranges. So it is possible that just the different chip made the
difference based on compatibilty or characteristics of chip. But the
other reasons are just as probable.
What brand (not just shield type) cable were you using? I realize that
you would not likely pursue additional tests as you found a fix
already, but it would be interesting to know, just so we can keep
collecting data should we experience similar problems in the future.
We had a similar situation that was due to chipset. We ran 10 mbps
ethernet 550 feet to our subscriber. (different radio brand). We used
a slightly higher power voltage to make up for cable loss. Our
laptops worked great over the link. The customer's 3 identical
routers could not stay connected for long. We were not sure if it was
a speed autodetection issue, or the distance for the chip to work. We
installed a 10mbps Cisco Switch in between their router and our cable
dmarc in their premise, and it all worked.
Tom DeReggi
RapidDSL & Wireless, Inc
IntAirNet- Fixed Wireless Broadband
----- Original Message ----- From: "Brian Rohrbacher"
<[EMAIL PROTECTED]>
To: "WISPA General List" <wireless@wispa.org>
Sent: Friday, September 15, 2006 9:20 AM
Subject: Re: [WISPA] MT power supplies THE SOLUTION
I started with.... RB 532 on tower. It comes down 265 feet to
poe injector to router. Major packet loss.
2) switched RB 532 out. No change.
3) Created test setup on ground with "bad board" and it looked fine.
(from laptop--6ft cable----poe----265 ft---RB)
4) Blamed it on the cable, and got a cable certifier from a friend.
5) Right before climb, I re did the test setup on the ground. This
time I plugged the 265 feet into the actual router instead of my
laptop. The problen was back. (I was bummed)
6) One final test. Get another 265 foot cable. I used 265ft for
power and 265ft for data to eth 2 or 3. Problem solved. I can only
speculate that the chipset on RB 532 poe port is diffrent from the
chipset on eth 2/3.
And for whatever reason it was not compatable with cable, hardware,
ect.....setup.
I may never know for sure why, but I have the workaround. Good
enough for me.
FWIW I ended up pulling 2 new cables (all 3 certified fine). I used
the original cable for data (it has "real" shield) I used my new 2
(cheapo foil shield) for power and slapped the other into eth3 for
the heck of it.
Lessons learned for next time. Measure cable, crimp, and power up on
ground using the EXACT same everything as what the final deployment
will have. And then test.
Hope that sums it all up.
Ok to directly answer your question. Yes. I did this on the ground
test unit.
Brian
Rohrbacher
Paul Hendry wrote:
Brian,
Just out of interest, did you try running both power and data over
the new
cable and did you still see the same issue?
P.
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of Brian Rohrbacher
Sent: 15 September 2006 02:43
To: WISPA General List
Subject: Re: [WISPA] MT power supplies THE SOLUTION
First off. I'm back to a 48v 420mA power supply.
To the solution.
I ran another cat5 up the tower and plugged it into the RB 532.
Now I have one cable for poe and one cable for data, and it all
works fine.
And check this. My headache went away as soon as the problem did. :)
Problem solved. NEXT!
Brian
Tom DeReggi wrote:
Amps don't mean a thing without disclosing Volts, Consider Watts
instead. 1300mA at 3V is much different than 1300mA at 18V.
The mPCI slot (SR5) is 3.3V. Power to the Motherboard is from
12-48V. W=V*A
Tom DeReggi
RapidDSL & Wireless, Inc
IntAirNet- Fixed Wireless Broadband
----- Original Message ----- From: "Mark McElvy" <[EMAIL PROTECTED]>
To: "WISPA General List" <wireless@wispa.org>
Sent: Wednesday, September 13, 2006 8:19 AM
Subject: RE: [WISPA] MT power supplies
I am surprised no one has mentioned this. I looked up power
consumption
on the SR5 and it shows 800 to 1300 mA each. You state your power
supply
is 700mA. I did not look up power consumption for the RB532 but I
would
think you would need at least a 3A supply.
-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On
Behalf Of Brian Rohrbacher
Sent: Tuesday, September 12, 2006 11:51 AM
To: WISPA General List
Subject: Re: [WISPA] MT power supplies
So, does anyone know if it looks like I would be fine on the power
side
of things?
I have tweaked the ethernet port settings for no gain.
Next step is to get climbing 280ft to replace board, but I'd like to
confirm power first.
Brian
Brian Rohrbacher wrote:
I have a RB 532 on 300 foot of cat 5 with 2 sr5.
I'm using poe 48v .700a power supply.
I'm seeing weirdness.
Do I have enough "juice"
Brian
--
WISPA Wireless List: wireless@wispa.org
Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless
Archives: http://lists.wispa.org/pipermail/wireless/
--
No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.1.405 / Virus Database: 268.12.4/448 - Release Date:
9/14/2006
--
WISPA Wireless List: wireless@wispa.org
Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless
Archives: http://lists.wispa.org/pipermail/wireless/
