Measuring power, especially over a substantial range of frequencies, has
always been a difficult task.
If it is done by measuring voltage or current, a 1% error in voltage represents
2% error in power.
Taking that into account and the calibration of the element plugged into it
which has a frequency response
which limits its accuracy. 5% as an guaranteed maximum error is not that bad.
Just look at the analog AC volt meters.
If you get one that is specified to have 2% of full scale accuracy you are
doing real good and that translates to
4% in power. We have not even thrown in temperature concerns. If you were
trying to measure the power into
a purely resistive load at 60Hz with a pair of good old reliable analog Simpson
260 meters, one to measure voltage and
the other to measure current via a resistor in series with the load, you would
be lucky to get a result within
5% or the actual value as measured by a calorimeter. Errors creep in from all
directions and when they all add up
you find your measurement way off. It was a long time before hams started
actually measuring output power.
Accurate power meters were just too costly for most. I would have loved to have
a good power meter in the
60's so I could determine the efficiency and power gain of my amplifiers.
73
Bill wa4lav
___________________________________
From: Amps [amps-bounces@contesting.com] on behalf of Jim Thomson
[jim.thom@telus.net]
Sent: Monday, May 04, 2015 7:47 AM
To: amps@contesting.com
Subject: [Amps] Time for New Power Meter
Date: Sun, 3 May 2015 21:38:01 -0700
From: Bill Turner <dezrat@outlook.com>
To: Amps group <amps@contesting.com>
Subject: Re: [Amps] Time for New Power Meter
------------ ORIGINAL MESSAGE ------------(may be snipped)
On Mon, 04 May 2015 08:09:47 +0800, you wrote:
>Might I be allowed to suggest that Accuracy +/- 5% of Full Scale Power"
>means that the best accuracy they guarantee is that 5% at full scale.
>
>At lower readings, it may still be 5% or it might be a little more.
>Whatever the figure might be, to cover their ass, the manufacturer is
>not going to make any guarantees other than at full scale.
>
>To suggest +/- 125 Watts anywhere on the scale, for a 2500 Watt range,
>simply does not make sense and it would make a total mockery of the
>using the instrument as a power measuring device.
>
>73, Alek.
>VK6APK
REPLY:
Perfectly stated. Thank you, Alek.
73, Bill W6WRT
## Nice try Bill...it doesn’t work that way. Its +/- 5% of full
scale.....anywhere on the scale. Even W8JI harps
on upon that fact. If you try and measure say 100 watts..using a 2.5 kw
slug, you will get BS readings. You need to use
the smallest element that will do the job. That’s why I have elements for
100-250-500-1000-2500-5000-10,000 and 25 kw.
## from birds site...from the Bird Q+A url the fellow posted..which u
obviously didn’t read.
Question / Issue
Bird 43 wattmeter with 250H slug element. Does the 5% +/- accuracy at
full scale mean it also has the same 5% accuracy at any point along that scale,
or is it more accurate for levels below 250w? I believe you understand my
question. Thanks!
Answer / Solution
With the Bird 43 the +/- 5% accuracy full scale means at 250 watts the
accuracy is based on the 250 W element in use. So using a 250W element the
accuracy is 250 watts + 5% and 250 watts -5%, or 250 +12.5 watts or 250 -12.5
watts. The accuracy window is therefore 237.5 to 262.5 watts.
If you try to measure 200 watts using the 250 watt element then the SAME
+12.5 or -12.5 accuracy window is still in effect, as it is BASED ON FULL SCALE
OF THE ELEMENT.
If you need the accuracy to track the lower power then the accuracy spec
you need to look for is +/- 5% of display. We also manufacture that type of
metering.
## Note, in my experience, if you use an element that is too big for the
job, the results will be on the high side..but not by a helluva lot. Bird sez
to use the smallest slug to do the job. My L4B in low power mode indicates
700 watts..using a 5 kw slug. It indicates 650w, using a 2.5 kw slug. It
indicates 625 watts..when using a 1 kw slug. The 625 result is close to dead
on. Some of it will depend on what freq you are on. The elements cover 2-30
mhz... and I still don’t know what freq they are calibrated at...when they
leave the factory. They read BS on 160m too...way too high. They make a 500
khz to 2.2 mhz slug that may be a bit more accurate on 160m. Also, the
readings will drift to the high side when the room temp is higher. IE: 60
deg F in the morning..vs 85 deg F in the late afternoon. The array
solutions power master meters are not affected by room temp. Array solutions
sez they hand calibrate each coupler to NIST standards. If you want to
calibrate it urself, go for it. The frwd and rvs calibration points can be
tweaked in 1% increments...from -15% to +15%. They mark the calibration
settings to use on each coupler....once for 160-10m...and again for 6m band.
Worse case scenario is if a 1% calibration point straddled the actual real
power. In which case the most it can be off is .5%.
## 1.5 kw is 5.477 amps into a 50 ohm load. = 273.861 volts rms. =
387.2398 V peak. = 774.4797 peak to peak. Its not rocket science to
calibrate a scope to read 387.2398 volts. IE: make sure a real 400V Peak
indicates 400 V peak on the scope. As noted b4, if the scope is off by just
1%..the calculated power results will be off by 2%...or double. Since power =
V squared / 50 ohms.
## Everybody that I know that uses an AL-1500 / Om etc, etc, brags how
they run em at 2.5 kw pep out on 80m. That has been going on for years. Ok,
so they are 2.2 db more than 1.5 kw...who really cares. Measuring po at the
back of the amp is silly anyway. What really counts is ERP...which will be a
function of feed line loss plus gain of the ant plus po of the amp. 1.5 kw
into a stacked array is...ok, meanwhile 1.6 kw into a dipole is a no-no...gimme
a break.
Jim VE7RF
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|