[Yaesu] FT-ONE - low output.

[Dr. David Kirkby]

On 06/ 1/11 07:37 PM, Danny Pease wrote:

>> Unfortunately I've never tested any ham meters in a lab, so I don't know
>> how good/bad they are.
>
> Then how do you KNOW some Birds won't do 10% accuracy?

Since I've tested a large number of slugs, and know some slugs simply don't. 
Military users were sending slugs in with a notice that +/- 10% of FSD should be 
considered a pass, not a failure. But even with that, some failed.


> Some copies of the 1956 vintage Monimatch are often sold for CB. Near
> all including Bird and Collins wattmeters have an adjustable resistor
> for calibration. Bird pots are hidden in the slugs generally. Many CB
> SWR bridges with watts indications have a serious sensitivity to applied
> frequency.

The problem is they vary with frequency, so simply adjusting a pot is not the 
solution. I had that done on one element for 432 MHz, as that was the only 
frequency I was interested in. But on HF, that is not practical.

>> I've often wondered how difficult it would be for hams to make a water
>> calorimeter. For high power, that's how its measured in a standards lab.
>> Of course, this would require that you transmit for a time sufficient to
>> reach thermal equilibrium, which would exceed the duty cycle of most ham
>> rigs. Making one would be a major project, but an interesting one and
>> within the capabilities of a dedicated constructor.
>
> There are two schemes for water calorimetry. One used occasionally for
> primary standards uses controlled water flow through a well insulated
> load and measures the temperature rise of that water flow.

That was the type I was thinking of, since I know how it works. That was the 
system used in the standards lab. Note this was not directly used for measuring 
Bird meters. It is too slow for testing lots of devices.

> A better ham solution uses the load in a well insulated water volume.
> With that water stirred to keep the temperature uniform, its easy enough
> to compute the rate of temperature rise by the power applied. 3.412 BTU
> per watt. I believe the fundamental is that 1 pound of water rises one
> degree F for each applied BTU, though that may off by a factor of a
> thousand. However its easy enough to measure the rise in water bath
> temperature vs time with applied RF power. Then to let the water cool,
> and apply easily measured DC or power line frequency AC power to the
> same load and adjust for the same rate of water bath temperature rise.
> The insulation can be a $3 foam picnic cooler, the load can be a coil of
> small diameter coax immersed in the water in the cooler. A rapid
> response dial thermometer cost me $7 in the hardware store, should I
> want to do that I keep a couple on hand for checking my vehicle's AC
> performance and when farming for checking soil temperature when planning
> for planting. By calibrating with DC, the thermal mass of the load and
> the water is combined so the experimenter need not measure the volume of
> the water precisely. But the technique gets RF power measurement down to
> the fundamentals of measuring time and temperature. And with the DC
> calibration the precision is limited mostly by the operator care and the
> calibration of the DC instruments which can be traced to primary world
> standards if need be.

Of course the specific heat capacity of water would also depend on impurities in 
it, though I doubt that would be the major source of error. But there are lots 
of little errors. But ultimately you could do a lot better job than a Bird, at a 
lot less convenience.



>> I think for me a spectrum analyser is the easiest solution. It may not
>> be the best, but is good enough. I could calibrate a ham meter against
>> that.
>>
> Being that the spectrum analyzer is essentially a logarithmic scanning
> receiver with tolerances of a few dB, I'd not consider that precise
> enough for a calibration standard. I much prefer a good (TEK or
> HP/AGILENT only) oscilloscope. I own a Tek 475, the 465 is great for HF,
> like its direct competition the HP1740. I'm less of a fan of digital
> scopes, but the makers don't build analogue scopes anymore.

> 73, Jerry, K0CQ

A decent spectrum analyzer is a lot better than "a few dB".

> Bird 43 meters are only spec'ed at full scale, anything else and their
> spec's are out the window. Beyond that, the weak and variable link is the
> element itself, which is where the real inaccuracy lies.

Agreed.

> Unless you have
> access to a calibrated power meter, such as a calorimeter, you never know
> what the accuracy of a Bird element is at less than full scale.

But their spec is +/- 5% of FSD. But they don't do that. Even full scale, they 
are not +/- 5%.

> Using a 2500
> watt element to measure 1500 watts is very common, but how many have the
> element calibrated at 1500 watts?

I suspect few. The claimed accuracy is 5$ of FSD which would be an error of 125 
W, but in practice they are rarely that good. Most will be accurate to +/- 250 W 
(+/- 10% of FSD), if you are unlucky +/- 375 W (15% of FSD)

> Not only that, is your element calibrated
> at every ham band you use?

Exactly. My limited experience is when I told the distributor I wanted to use it 
only at 432 MHz, they got if within 5% of FSD at that frequency. I wish I could 
recall what accuracy it actually had, but I know it met the spec.

But that was a brand new element. I've measured many elements which have never 
been in service and we put in the standards lab before being used. Many were out 
of spec from day 1.

> Have you ever applied too much power to a lower
> power element?

Yes, elements do get abused.

> Have you ever dropped an element?

Yes. Can't say I have ever quantified the error that introduces.

> Have you ever used a low
> power element for reflected reading and exceeded the 10:1 ratio. I know from
> experience it is very easy to damage a Bird element by using, say a 5 watt
> element in the reverse but run 100 watts through the meter and watch the
> reflected reading creep up. That creeping is heat causing the diode to leak
> and/or the load resistor overheating, both affecting calibration. If you
> antenna system (load) is not a non-reactive 50 ohms, Bird 43 elements and
> meters accuracy are again suspect in accuracy.



> It is very easy to know your meter is accurate and any one power level and
> frequency by comparing to a known good meter (not one you think it accurate)
> but it is much tougher to know a meter is accurate over a large range of
> frequency and power levels.

Agreed. There is a practical point though - we rarely need high accuracy for 
power measurement.

> Measuring PEP is basically a guess unless you
> use an accurate Oscilloscope and a really good dummy load. These "peak
> reading" meters do not measure PEP, they just add a capacitor to hold the
> meter deflection closer to the top of the meter swing. It is much cheaper
> and much more accurate to use an oscilloscope and a dummy load/attenuator to
> measure power than to buy a really good wattmeter. I have also seen a few
> big name oscilloscopes that were not very accurate when measuring radio
> frequencies, so again, you are back to having to have a known good reference
> to know where you are.

Measuring RF power accurately is not an easy task.

> One place I worked at with its own calibration lab calibrated Bird 43
> elements against a Bird standard at the power levels and frequencies we
> tested our transmitters at and made up a cheat sheet for each element
> showing what fudge factor to add or subtract from the meter reading, power
> vs. frequency.

I can understand that would be possible, but for military users, it was a simple 
pass/fail.

> This was a lot of work for the calibration techs, so we
> finally went to Bird 4420 meters (+/- 3% at all power levels within the
> frequency range of the sensor, in our case 25 MHz to 1000 MHz), that solved
> almost all of our accuracy issues.

I've never used the 4420 series. I guess these are more expensive that the Bird 
43s that some people worship like God.

> Making your own thermal meter is really pretty cheap, just not convenient to
> use for most of our applications.

I've never done it, but I think it would be doable too. I can't say I have much 
inclination to do it myself.

> All that said, 10% is more than accurate enough for amateur use.

Agreed, which is why I'm not going to sweat over a few percent. But I dislike 
the Bird 43 mainly since it has a specification far higher than it actually 
achieves.

-- 
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?