NTP time, Jitter shown

Actually i have a lot of questions about PTP100 series. Short introduction, i’m an engineer from one of the WISP in Malaysia. We covered our network by using PTP100 and PTP400Lite/Full. There are couples of link we had established, and aware that got some issues for Canopy. I listed down:

First - NTP
Although NTP server is synchronized, but the time always be -8 instead of +8 that showing on BHM. Graph attached.


Second - Jitter (BHM/BHS BH20)
BHM Version


/---------'
C A N O P Y

Motorola Broadband Wireless Technology Center
(Copyright 2001-2008 Motorola Inc.)

Telnet+> ver
Software Version : CANOPY 9.0 BH20-DES
Software Boot Version : CANOPYBOOT 3.0
FPGA Version : 061708
PLD Version : Unable to get PLD Version on platform 9
Frequency Band : 5.7GHz
Hardware Platform : 9
Hardware Minor Revision : 0
Device Type : CANOPY

Telnet+>


BHS Version

/---------'
C A N O P Y

Motorola Broadband Wireless Technology Center
(Copyright 2001-2008 Motorola Inc.)

Telnet+> ver
Software Version : CANOPY 9.0 BH20-DES
Software Boot Version : CANOPYBOOT 1.0
FPGA Version : 061808
PLD Version : 11
Frequency Band : 5.7GHz
Hardware Platform : 10
Hardware Minor Revision : 0
Device Type : CANOPY

Telnet+>


Both BH20 radio running 2X rate, but the Jitter and Powel Level shown different. Is this bug?

Yikes. While I can’t comment on the signal variations, the NTP thing is a known “lack of feature” - there is no timezone support. Going on about 4 years here… viewtopic.php?f=4&t=517

setup your own ntp server with the correct GMT settings and point AP’s/ BH’s to you ntp server rather than an ntp.org address.

nucoles wrote:
setup your own ntp server with the correct GMT settings and point AP's/ BH's to you ntp server rather than an ntp.org address.


Yep, that's what we do.

Are your TX power levels set the same on both your masters and slaves and are you using bare or dished units? I have slight variances between mine, but not as great as you're seeing. Could be some kind of multipath or something on one end.

Since the high end threshold is -40dB, I’m surprised these are even working. How far apart are these and are they dished?

wifiguy wrote:
[quote="nucoles":n3mwdaxf]setup your own ntp server with the correct GMT settings and point AP's/ BH's to you ntp server rather than an ntp.org address.


Yep, that's what we do.

Are your TX power levels set the same on both your masters and slaves and are you using bare or dished units? I have slight variances between mine, but not as great as you're seeing. Could be some kind of multipath or something on one end.[/quote:n3mwdaxf]

Yes, we have our NTP server installed, query from it we getting correct date/time. But canopy doesn't. There is no way to allow me to set Time Zone +8

:(
Jerry Richardson wrote:
Since the high end threshold is -40dB, I'm surprised these are even working. How far apart are these and are they dished?


2 miles apart with both reflectors installed. Just wondering why the BHS is mis-displayed.

It may be due to overloading the receiver.

with 18dB dishes at both ends, you could reduce the power to 3dB and they would still hook up.

Jerry Richardson wrote:
It may be due to overloading the receiver.

with 18dB dishes at both ends, you could reduce the power to 3dB and they would still hook up.



What is the recommended Tx for each link? If you set to lower power, will it affects link quality?

+20 dB over the noise floor is ideal.

So, if your noise floor is -82, then you would want your links at -62. With reflectors, you won’t be able to reduce the power low enough.

My recommendation would be to remove the dish from one end which would remove 18dB of gain from the system. Then see what your power levels look like, and if necessary you can reduce Tx power on the master and slave to hit your target.

With RF less is more. It’s good practice to only use as much power as you need - anything above that just pollutes the RF spectrum.