Re: dbm to volts (humax F2foxT)

The message <0k4ab31nf9k650nd5gu85jgctfqhh966d6@xxxxxxx>
from brightside S9 <address@xxxxxxxxxxxxxxxxxxxxxx> contains these words:

On Sat, 4 Aug 2007 16:24:58 +0100, Johnny B Good
<jcs.computers***@xxxxxxxxxxxxxxxx> wrote:

The message <hLKsi.5039$vi3.1089@xxxxxxxxxxxxxxxxxxxx>
from "harrogate3" <nospam3@xxxxxxxxxxxx> contains these words:


"Andy Wade" <spambucket@xxxxxxxxxxxxxxxxxxx> wrote in message
news:46b35c00$0$24757$da0feed9@xxxxxxxxxxxxxxxxx
mike wrote:

Conversely, to convert dBm to microvolts follow this example:
<Snip>

You're making hard work of this. To convert from dBm to dBuV in a
75-ohm system just add 109 (or 108.751 if you want to be fussy).

To be _really_ fussy, it's actually 108.7506126 (to 7 decimal places:-)

The
corresponding number for a 50-ohm system is 107 (106.990 if being
fussy).

To be _really_ fussy, it's actually 106.9897 :-)


--
Andy


Can we have yor maths for that please?

0dBuV is 1uV irrespective of impedence

0dBm/75R is 1mW across 75R which by my maths is 273mV

To 3 digit accuracy, it's actually 274mV (to 10 digits, it's
273.8612788mV)

Que?

What? Were you questioning the 273mV figure? :-)


Oh! It seems someone needs to understand significant digits.

ITYMM the difference between rounding to the nearest n significant
digits and truncating to n significant digits (the microsoft way :-)

Stop nit picking and take a look at

http://www.physics.uoguelph.ca/tutorials/sig_fig/SIG_dig.htm

The big difference here is that all the calculations were done by the
use of a scientific calculator which displays to 10 digit resolution
(which doesn't necessarily mean that that's the resolution limit of the
intermediate results produced by the calculator). This leaves the choice
of how many significant digits to quote as the final answer to the
user's discretion (applying the standard rounding rule to correctly
determine the least significant digit value).

That 3 digit value of 273mV is 'correct' only in the Microsoft sense
and wrong by the rules for rounding to the least significant digit in
the answer. As it happens, rounding either of those 3 digit figures down
to 2 significant digits will both, correctly, produce a value of 270mV
provided it is stated to have a 2 digit accuracy (otherwise a 3 digit
accuracy would be implied).

Normally, 2 digit accuracy for db figures suffices for most practical
purposes until we start dealing in +/- db figures greater than 100 where
we then need 3 digits accuracy to at least stay within +/- half a db of
the actual value.

When dealing with a logarithmic based scale, a +/- 5db error remains
equally significant regardless of how many tens of db we are displaced
from the reference level. However, a 2 digit accuracy for voltage (or
current) levels only represents a worst case decibel error of just under
+/- 0.5db (0.4455 db to 4 significant digits) whatever the value.

Just to put it all in perspective, the db error due to specifying 273mV
instead of the slightly less inaccurate 274mV figure represents a mere
0.0318 db ( that's to an accuracy of four decimal places <or 4
significant digits, unless you prefer the form 3.18E-2, which then makes
it accurate to 3 significant digits> :-) Either way, the error is
unlikely to be measurable on any analyser kit available to the aerial
rigging trade. :-)

--
Regards, John.

Please remove the "ohggcyht" before replying.
The address has been munged to reject Spam-bots.

.

Loading