"Richard Clark" wrote in message
...
On 20 Sep 2003 22:23:36 GMT, (Radio913) wrote:
I've heard that the Bird Model 43 is very accurate.

Hi OM,

Now you know different. Besides, NO ONE is accurate without
comparison to a known Standard.

A Bird can be made accurate, better, in fact, than the manufacturer's
claims. I doubt you would spend the money to do it the first time,
much less each time required to keep it accurate (much less more
accurate).

It makes more sense to invest in a good load and calibrate it. It is
not at all difficult, except for the cash-and-carry tech who shuns
benchwork.

73's
Richard Clark, KB7QHC


I'll bet you'd love an invitation to expand on your comments about a "good
load" and "calibrate it."

The only good load is one that presents the desired, purely resistive
impedance. Maybe you can adjust the load, but you don't make it better by
calibration.

Or perhaps you are assuming a good load, and by calibration, you are meaning
some type of thermal calibration to indicate power.

I don't use loads to measure power, I use attenuators (sufficient for the
expected power), and put a bolometer on the end of the attenuator string. Of
course, I have the luxury of other test equipment, so I can check the
attenuators, with low power, against my spectrum analyzer and signal
generators.

All that said, I still think that a Bird 43 (and yes, a good load or
attenuator string [once you get 60 dB or so on the attenuator string, an
open-end reflection is quite minor]) is the best route for a ham. A ham
typically doesn't need even the factory accuracy of a Bird, but the
versatility of forward and reverse power, with multiple power ranges and
multiple frequency ranges, is very nice. Plus, it doesn't even need
batteries, and it's small and light. (Anybody who ever lifted HP gear can
now grin.)

For the 100 MHz region, you can think about a directional coupler. Nice, but
you still need either a bolometric power meter or a calibrated spectrum
analyzer. If you don't already have those two goodies, then the Bird is the
champion choice.

Ed
WB6WSN

On Sat, 20 Sep 2003 18:41:21 -0700, "Ed Price"
wrote:

I'll bet you'd love an invitation to expand on your comments about a "good
load" and "calibrate it."

Perhaps, but very few bench techs are found here.


The only good load is one that presents the desired, purely resistive
impedance. Maybe you can adjust the load, but you don't make it better by
calibration.

Arguable. Calibration is not what you do to something, that is tuning
or adjusting. Calibration is knowledge.

Or perhaps you are assuming a good load, and by calibration, you are meaning
some type of thermal calibration to indicate power.

There are many ways to accomplish the calibration, again, there are
too few bench techs here to care.

I don't use loads to measure power, I use attenuators (sufficient for the
expected power), and put a bolometer on the end of the attenuator string.

The Bolometer is a classic load, there are many others (like a
barreter).

Of
course, I have the luxury of other test equipment, so I can check the
attenuators, with low power, against my spectrum analyzer and signal
generators.

A good method that employs the easier methods of measuring differences
rather than absolutes. However, you must eventually obtain some
standard to make the final determination.

All that said, I still think that a Bird 43 (and yes, a good load or
attenuator string [once you get 60 dB or so on the attenuator string, an
open-end reflection is quite minor]) is the best route for a ham. A ham
typically doesn't need even the factory accuracy of a Bird, but the
versatility of forward and reverse power, with multiple power ranges and
multiple frequency ranges, is very nice. Plus, it doesn't even need
batteries, and it's small and light. (Anybody who ever lifted HP gear can
now grin.)

For the 100 MHz region, you can think about a directional coupler. Nice, but
you still need either a bolometric power meter or a calibrated spectrum
analyzer. If you don't already have those two goodies, then the Bird is the
champion choice.

Ed
WB6WSN


Hi Ed,

I've never run across a need for batteries, except to light the meter.
There are better meters than a Bird. The AN/URM-120 is easily better,
and sells cheaper, has all the same qualities, but for its advantage
it also is larger and presents its readings with the meter horizontal.
If those were the only down-sides, then it is much like personal
choices in tie color.

73's
Richard Clark, KB7QHC

I'll bet you'd love an invitation to expand on your comments about a "good
load" and "calibrate it."

Perhaps, but very few bench techs are found here.


How about a bench engineer? How did you come to the conclusion that the
Bird plugs go bad after 3 months?

Do you have a transmitter of known Pout that you use as a standard? Of
what
power level?

Or do you use known good attenuators, and measure the error of relative
powers?



The Bolometer is a classic load, there are many others (like a
barreter).


I'm not familiar with these, tell me more.




All that said, I still think that a Bird 43 (and yes, a good load or
attenuator string [once you get 60 dB or so on the attenuator string, an
open-end reflection is quite minor]) is the best route for a ham.


Yeah, someone suggested using long lengths of RG-58 (the lossier, the
better) to improve the return loss of a no-so-50 ohm dummy load.






For the 100 MHz region, you can think about a directional coupler. Nice, but
you still need either a bolometric power meter or a calibrated spectrum
analyzer. If you don't already have those two goodies, then the Bird is the
champion choice.

Ed
WB6WSN


Hi Ed,

I've never run across a need for batteries, except to light the meter.
There are better meters than a Bird. The AN/URM-120 is easily better,
and sells cheaper, has all the same qualities, but for its advantage
it also is larger and presents its readings with the meter horizontal.


For the AN/URM-120, i found this:

"On the 25-250 and 200-1000 Mhz slugs 10, 50, 100, and 500 watts
are available."

No 1000 or 2000 watt slugs for 25-250 MHz?


Slick

On 21 Sep 2003 20:45:52 GMT, (Radio913) wrote:

How about a bench engineer? How did you come to the conclusion that the
Bird plugs go bad after 3 months?

This is the standard calibration cycle. I did not come to the
conclusion, it is commonplace schedule subscribed to by registered
laboratories performing calibrations traceable to the NIST.

The necessity is driven by experience of the industry exhibiting that
need. In other words, for 100 Birds and 100 URM-120's coming into the
lab for calibration or qualification, 70 to 90 of the Birds were out
of tolerance; and 70 to 90 of the URM-120's met spec. 10 to 30% out
of tolerance hits are not acceptable for FAA flight line work (the
Birds are much worse).

If you waited a year instead of 3 months, the Bird that was out of
whack would still be out of whack, but probably at the same level.
You could say pretty much the same for 20 years later (barring smoke
curling up at hot connectors). Repeatability is accurate only when
you've calibrated it. In other words, if I had simply noted what
Power was being read (like 60W) for an applied Power (like 100W); then
it would probably stay that way until judgement day (evidenced by
smoking connectors). If someone lost that calibration certificate,
then for those many years after, they would have been repeatedly off
by a considerable margin (and probably trying in vain to force any amp
to do what it was already doing). This is called "clueless." They
would sneer at the vendor of the amps, change their sources, and
snuggle up to their meter each evening.

Do you have a transmitter of known Pout that you use as a standard?

Of course. And a reflection free system.

Of what power level?

Full Scale and at each Cardinal point to calibrate scale linearity.
As I said, even the Bird can be calibrated to accuracies that exceed
their specification.

Or do you use known good attenuators, and measure the error of relative
powers?

Attenuators do nothing for the required input. The reference to
attenuators offered by Ed was for the detector head in a Directional
Coupled system.

The Bolometer is a classic load, there are many others (like a
barreter).
I'm not familiar with these, tell me more.

Maybe because I mis-spelled it: Barretter.

Anyway, it is a fine wire detector (that's right, just a wire). The
most accessible form, and most common item is found in a 10 mA fuse.
Another form is called a Wollaston wire. The Barretter is used in a
lot of HP Power Meters (the premium stuff). They are usually found in
pairs with one being half of a bridge with a steady bias, and the
other being the termination to a coupler connection into the
transmission line feeding the load. Only takes about +10dBm to wipe
out one of these detectors (be prepared to spend a couple hundred to
fix it too).

All that said, I still think that a Bird 43 (and yes, a good load or
attenuator string [once you get 60 dB or so on the attenuator string, an
open-end reflection is quite minor]) is the best route for a ham.

Yeah, someone suggested using long lengths of RG-58 (the lossier, the
better) to improve the return loss of a no-so-50 ohm dummy load.

That, too, is suitable. Calibration, as I pointed out, is not making
adjustments, it is gathering information. You might optimize
performance as an ancillary function of this; and to maintain
equipment in good order (plus keep customers); but if you offer no
information, then no calibration has been performed.

Calibration would also reveal that this suggestion of lossy cable was
a loser proposition if you didn't check for connector mismatch. If it
was considered trash cable, it probably has trash connectors.

For the AN/URM-120, i found this:

"On the 25-250 and 200-1000 Mhz slugs 10, 50, 100, and 500 watts
are available."

No 1000 or 2000 watt slugs for 25-250 MHz?


Slick


Hi OM,

I wouldn't know off-hand. Not really important anyway as Ed pointed
out. Directional Couplers work just as well too, plus they make the
bench more versatile than one instrument with several plug-ins. You
can use Direction Couplers for many things, the plugs-ins for only
one.

Power can be measured by calorimeters (HP made one decades ago that is
still better than any suggestions here). Or measured by a Lamp
(photometrically, as long as you are matched to its Z). Or measured
by a Lamp (by calorimetric method, as long as you are matched to its
Z). Or measured by a Thermistor (another calorimetric method). Or
measured by Thermocouple (guess what, another calorimetric method).
Or measured by substitution (using any of the above as non-matched
loads as long as both sources are fully matched). By measuring the
resistance of the load (especially carbon) as long as the power
changes the resistance to a matched condition (or by substitution if
not).

73's
Richard Clark, KB7QHC

Richard Clark wrote in message . ..

If you waited a year instead of 3 months, the Bird that was out of
whack would still be out of whack, but probably at the same level.
You could say pretty much the same for 20 years later (barring smoke
curling up at hot connectors). Repeatability is accurate only when
you've calibrated it.


Then it sounds like you could re-calibrate after 3 months, and not
have to worry about it for a while.


Slick

"Radio913" wrote in message
...

SNIP

All that said, I still think that a Bird 43 (and yes, a good load or
attenuator string [once you get 60 dB or so on the attenuator string, an
open-end reflection is quite minor]) is the best route for a ham.


Yeah, someone suggested using long lengths of RG-58 (the lossier,
the
better) to improve the return loss of a no-so-50 ohm dummy load.

SNIP


Slick


Well, I wouldn't go quite that far. The "loss" of RG-58 is due to a
combination of loss in the transmission line and leakage of energy through
the relatively poor shielding. That radiated energy hanging around your test
bench is more than likely cause you new measurement errors.

Ed