On 9/28/2015 2:53 PM, Jerry Stuckle wrote:
On 9/28/2015 2:27 PM, Wayne wrote:


"Jerry Stuckle" wrote in message ...

On 9/28/2015 12:47 PM, rickman wrote:
On 9/28/2015 10:38 AM, Jerry Stuckle wrote:
On 9/28/2015 12:03 AM, rickman wrote:
On 9/27/2015 10:39 PM, Jerry Stuckle wrote:
On 9/27/2015 9:46 PM, Wayne wrote:


From LUNA web site regarding optical measurements which should be no
different from RF...


It "shouldn't be" - but optical measurements are handled differently
than electrical measurements. Fiber Optics have their own way of
measuring loss, reflection and refraction (which doesn't exist in
feedlines).

That's like applying electrician's color codes to electronics. They
both have color codes - but don't hook the electrician's black wire to
ground - or the transformer's green wires to safety ground.

I thought you would claim optical was different. That's why I included
the VSWR vs return loss table link. You didn't comment on that.


# I didn't because I thought it was obvious. But I guess not to you.

# Return loss is calculated with logs. Logs of values 1 are negative.
# And -10db is smaller than -5 db.

# As the SWR approaches 1:1, the reflected power approaches 0, and the
# returned loss approaches NEGATIVE infinity. Note that I said NEGATIVE
# infinity. At the same point, the returned power measured in watts is 0.

Return loss is a positive number for passive networks. The equation has
(P out/P reflected). P out will never be less that P reflected, and
thus return loss will never be negative. (for passive networks)

As the SWR approaches 1:1, the return loss increases in a positive
direction, finally reaching infinity.


No, return loss is calculated as P reflected / P out. P out is the
constant with varying load; P reflected is the variable. The ratio is
always less than one, hence the calculation is always negative DB.

Please point to a reliable source which agrees with you.

I'd like to see a reliable source for either position.

From the references in the wikipedia article...

http://ieeexplore.ieee.org/xpl/freea...number=5162049

https://en.wikipedia.org/wiki/Return_loss

I didn't pay for the article, but it seems pretty clear that the authors
are saying the term is often incorrectly used. I have found other loss
equations in the form of Ls = 10 log (pt/pa) where pt is the input power
and pa is the output power. Even this is not quite the same since it
refers to the measured power at each end rather than the "lost" power.
The point is the input power is in the numerator yielding a positive dB
result for all cases.

I'd be happy to find anything authoritative.

--

Rick