Community discussions

MikroTik App
 
Bitto
Frequent Visitor
Frequent Visitor
Topic Author
Posts: 94
Joined: Wed May 02, 2012 10:15 am

RB435G voltage related radio throughput problem and solution

Sat Jun 29, 2013 9:43 pm

Hello Guys,

We installed two boxes of rb435G on a radio tower we use 12V 10A power adapters with 12V cooling fans. we used cat7 sftp cables (10 meters lenght) between nodes and switch with gigabit poe injectors.
When we boot up devices i try to test bw with built in tool total throughput was spiky unstable and unexpectedly low (30 mbit up 10 mbit down). I suspect it was a frequency or fresnel issue, i triple check everything. While i was checking my configuration, i noticed health monitor shows 9V DC with 7.4 Watt power consumption.
Day after we take a 24V 2.5 Amp adapter and climbled the tower.
First we increase output voltages of adapter to max (voltage increased 12.5V), throughput almost dubled after, we tried to give power directly to board this would increased voltage to 13.7V throughput increased futher and became more stable then, we decided to use 24V adapter. We get 90-80Mbit upload 50Mbit download we remove fans and tried 12V adapter with max power. board received 15V test was little slower but almost same 80Mbit up 60-45Mbit down and stable.
Sake of a beter understanding, i repeat tests and monitored health monitor; in the begining, consumption was 7.4 Watt "with bandwith test running" with max power and direct connection consumption increased to 14 watts, but it reduced 9.4 watts when we remove fans.

Any ideas why is this happening with 120Watt PSU?
24V 60Watt PSU is far more better i think mikrotik does not work well under 14V input voltage if you are using high power cards and active cooling.

So i am curious why is that mikrotik recomends 12V PSU with 12V fans in 435G's PDF
..
 
onnoossendrijver
Member
Member
Posts: 487
Joined: Mon Jul 14, 2008 11:10 am
Location: The Netherlands

Re: RB435G voltage related radio throughput problem and solu

Sat Jun 29, 2013 10:01 pm

Maybe this is not a voltage issue, but an issue with your power supply.
Most PSU's need some load before they give a stable output.
I think you just need some more load on that 120 watt PSU to operate in a stable fashion.
The 60 watt PSU probably operates in a more comfortable region, and therefore gives a more stable output.

You should always choose the right PSU for the task. A PSU is most efficient at about 90% load. However for safety reasons I would choose a PSU of about 2x the max. power usage.
 
Bitto
Frequent Visitor
Frequent Visitor
Topic Author
Posts: 94
Joined: Wed May 02, 2012 10:15 am

Re: RB435G voltage related radio throughput problem and solu

Sun Jun 30, 2013 12:13 pm

I do not think so because,

If we increase output of PSU throughput become more stable board consumption shows 14 watts if we discoonect fans at highest PSU output consumption reduces 9.4 watts and we get best throughput. Plus we are using 24V 8.5A 200 Watt PSU without a problem one of 435G's on this installation site (best throughput).
 
n21roadie
Forum Guru
Forum Guru
Posts: 1949
Joined: Fri Aug 07, 2009 10:36 pm
Location: Limerick,Ireland

Re: RB435G voltage related radio throughput problem and solu

Sun Jun 30, 2013 6:05 pm

@Bitto
How are you doing this throughput test as it is advised to test throughput you set a point before and after the item being tested, for example running a test from point A to point D would give the throughput between point B and C?
 
Bitto
Frequent Visitor
Frequent Visitor
Topic Author
Posts: 94
Joined: Wed May 02, 2012 10:15 am

Re: RB435G voltage related radio throughput problem and solu

Mon Jul 01, 2013 11:52 am

@n21roadie

I made this test between 2 nodes 10km apart with clear los.
While making the test, we increased and decreased voltagge and observed througput change with voltagge accordingly.

Who is online

Users browsing this forum: No registered users and 16 guests