[time-nuts] OT: Power level reference

J. Forster jfor at quik.com
Wed Dec 2 07:00:48 UTC 2009


Yeah... I got the model number wrong. I meant the 3403. Digtal, true RMS.
I used it for Y Factor measurements at IF before getting an AIL 2075.

Oops.
-John

=============

> Well John,
>
> Neither do you with regard to the hp-3400A RMS voltneter.  It is spec'd to
> only
> 10 MHz !  From 50Hz to 1 MHz it is +/-1% and at 10 MHz it is rated at
> +/-5%.  So
> at 10 MHz it is no better then the spec of Bird 43 slug type power meter
> (if you
> believe them).
>
> The old hp-434 Colorimetric power meter (i.e., measures heat content of
> applied
> power) is only good for 5%, but than can be improved somewhat with applied
> techniques.  However, this unit has the frequency range of DC to 12 GHz.
>
> The best you can do with an analog scope is 3% measurement of the screen
> display.  If you were to assume the load resistor whose voltage drop is
> being
> measured was perfect, then 3% is the best you could do.  Obviously, such
> is not
> the case so some allowance would have to considered because the resistor
> is not
> perfect, so, again, about the best you can do is in the area of 5%.  The
> newer
> digital displayed readout scopes are really no better.
>
> As someone pointed out, the best you could trust an Amateur type power
> meter
> would be 20% into a perfect "load" without some serious calibration
> efforts and
> recording the offsets at various frequencies.  As most all Amateurs would
> be hard
> pressed to understand and quantify their "load," they are really kidding
> themselves to use the meter for anything more then a relative forward
> power
> indicator.
>
> "J. Forster" wrote:
>
>> > At 08:30 PM 12/1/2009, J. Forster wrote...
>> >>In fact, on the Tek 7704 or 7704A (I forget which) there were "No
>> >>Cost"
>> >>options of maximally flat frequency response OR best pulse response.
>> >>OR!
>> >
>> > Sure, if you're using a 150 MHz to try and measure accurately at 100
>> > MHz, you're not going to get absolute accuracy. I'm not familiar with
>> > that particular scope,
>>
>> Really? The most successful 'scope in history.
>>
>> > but would bet the difference appears as peaking
>> > at the extreme. In the absence of circuitry which deliberately changes
>> > it, the response of an analog scope is (very nearly) Gaussian.
>> >
>> > Can you quantify the difference between those options when measuring a
>> > 50 MHz (1/3 BW) signal?
>> >
>> > Since the OP is a ham, I assume his quest is relative to commonly
>> > available power meters for that market, most of which can't be
>> expected
>> > to do better than 20% accuracy, maybe 5% at the high end (
>> > http://www.telepostinc.com/lp100.html ).
>> >
>> > Sure, you can buy an uncalibrated 3400A (good to ~150W), and get some
>> > unknown amount worse than 5% accuracy.
>>
>> Apparently, you don't know what an HP 3400A is.
>>
>> Stick to political blogs.
>>
>> -John
>>
>> ===============
>>
>> > Or a 432a (good to a whole 10
>> > mW!) for <$100, plus another $100+ for the mount/cable. Now you've got
>> > something which is good to a couple of %, with an _extremely_ limited
>> > range (and practically useless for many ham applications).
>> >
>> > A Gaussian scope is predictably off ~3% @ BW*0.3. A decent scope will
>> > have a vertical amp accurate to a couple of percent. With a 1% load,
>> > you can measure at the 5% level of a $400 dedicated wattmeter (from <
>> 1
>> > uW to 400W, and to the 2000W ham limit if you invest in a 100:1
>> probe),
>> > for little cost.
>> >
>> > "Accurate...simple and inexpensive." Many non-appliance-operator hams
>> > will already have a scope, so the cost is a $10 resistor (for a 100W
>> > rated Caddock, which should be good for a few seconds of full power,
>> > enough to make the measurement).
>> >
>> > Finally, as I've already mentioned, one can make a simple peak
>> detector
>> > using a rectifier, and measure the DC voltage off that.
>> >
>> >
>> >
>>
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
>
>






More information about the Time-nuts_lists.febo.com mailing list