Hello,
i have a question regarding the tx-power and the signal strength…
We have the following configuration:
RB532 with DCMA81 and 28dbi grid antenne on both ends, about 8km with clear line of sight, 40m height at one and about 15m on the other end. Connected each end with 6m Ecoflex 10 cable.
Channel: 5500 MHz
manual tx-power, all rates fixed
If i test the link with 15dbm tx power then i get a signal of 65-67 on both ends.
But when i reduce the tx power to 13dbm, i get a signal of 75-78
How could this be???
For my opinion, I should get a Signal of about 69, because i only reduce tx power with 2dbm!!!
Anyone who can give me a hint?
kind regards!
that´s really crazy:
with the following tx-power settings i get the following Signal levels:
10dmb → 81
13dbm → 77
14dbm → 75
15dbm → 66
Why is there such a big difference between 14 and 15 dbm?
It happens on both sides (same Card, DCMA81)…
Remember db is logaritm scale not linear.
Card power referred to 1mW:
10dbm = 10^(10/10)= 10 mW
13dbm = 10^(13/10)= 19.95 mW
14dbm = 10^(14/10)= 25.11 mW
15dbm = 10^(15/10)= 31.62 mW
16dbm = 10^(16/10)= 39.81 mW
20dbm = 10^(20/10) = 100 mW
With your configuration you have.
28dbi gain antenna absolute amplify the input signal :
28dbi = 10^(28/10) = 630
So you are irradiating with 13 dbm in input:
19.95 mW * 630(antenna gain) - cable loss(to calculate) = 12,5 W
while with 15 dbm:
31.62 mW * 630(antenna gain) - cable loss(to calculate) = 19.9 W
between 13 and 15 dbm your signal has about 7 W of differents.
thanks for your reply…
your calculations are absolutely right -but i think the rx-signal i get from the mikrotik is also an logarythmic value - isn´t it?
incoming signal is NOT meassured in mW - it´s meassured in dbm!
so my opinion is, that if i reduce tx power at one side with 3db, than rx signal on the other side should also be reduced by 3db!
please tell me if there is somethin that i missunderstood… 
With log powers (dBm) you ADD the values: i.e. a -3db change in TX power affects the receiver by -3dB also.
The results published suggest some non-linearity, most likely someones card is not reporting correctly.
Also note, TX power varies with modulation rate in most cases, so a “good” link with good SNR at high modulation rate may show a “lower” RX power at the receiver than a bad one …
Regards
okay - so my calculation was right…
So, what could be the case for the significant rx power loss on my link?
A bad working wireless card?
A bad working mikrotik driver, that cannot realy affect the tx power?
i also tried it on another link 5 minutes ago - following results:
17dbm → 68
15dbm → 72
13dbm → 78
Nearly the same - if i reduce the tx power with 4dbm, i loos 10dbm rx loss…
How could that be? Can anyone confirm this results?
Interesting results on a third link with CM9 card:
if i change the tx power ±4dbm, my rx power on the other side changes also by ±4dbm
Could it be that there is a problem with DCMA81 cards (Atheros 5413)?
With new Atheros 5006x chipset I’m verifyng other strange issue.
I think that driver needs to be update… I’m waiting
Yes you have reason about tx and rx dbm scale.
so - it seems that i can nothing do in this case??
just have to wait since mikrotik releases a better working driver for this card?
Would be interessting if MT could confirm that there are problems with this card and the tx power…
Hi all
I think it’s best to quote the exact card (make/model) on each end of your link, with the results, otherwise MT are not going be able to make much sense of it.
AFAIK each radio card has a different look-up table, set by the manufacturer and some of them might not be inaccurate.
Am sure MT have tested with the well-known ones (CM9, R52, SR5) but others might not necessarily have been tested.
Regards
CableFree Solutions
hello,
we have dcma81 on each end of this link.
On another link we have dcma81 at one end and cm9 on the other.
always the problem with the dcma81…
This card has the Atheros AR5413 Chipset - so i think it should be a well-known card…