Thermocouple RF ammeters, wattmeters, power output measurement, line efficiency

[k4kyv]

I now have my above-ground coax line installed to link the transmitter to the tuner at the base of the tower. The old buried coax link showed 93% efficiency when I first installed it, using my Mirage Hammy Hambone RF wattmeter.  I know it is off calibration, but it should still be accurate for relative power measurements.  When I re-checked it this summer, the efficiency was down to 75%.

To measure feedline efficiency, I hauled my large 50Ω dummy load down to the tower and connected the coax feed line to it. Then, back at the transmitter I checked SWR, which was something like 1.05:1.  I switched the meter to power output measurement, loaded the transmitter until the permanently mounted rf ammeter I use for reference indicated 3.0 amps, and took a power output reading. Then I removed the wattmeter from the line, went down to the tower, and re-inserted it between the coax and the dummy load, and compared readings between the transmitter end and antenna end of the coax.  I ran the test with both coax lines. Naturally, the transmitter was turned off while the meter was dis-connected and re-inserted into the line.  I got my exercise for the day making multiple trips between shack and tower.

Re-measuring the efficiency of both old and new transmission lines, using the same Mirage meter, both lines showed about the same efficiency @ around 85% this time. So I decided to try measuring efficiency again, replacing the wattmeter with a Weston 0-5 amp thermocouple rf ammeter. The same procedure was used again, with the reference meter reading 3.0 rf amps. With the new coax, the efficiency calculated out to 97% and the old coax to 93%.

The rf ammeter calibration may not be perfect, but it should be more stable and consistent than the Hammy Hambone wattmeter, which does make a good SWR indicator but I wouldn't trust it for power measurements.

While I was at it, I decided to run tests with RF ammeters I have on hand.  Using my permanently mounted thermocouple meter as a reference, I compared readings with the others.  I brought my dummy load back to the shack and connected it in place of the coax to the tower, inserted the meter under test in line, and loaded the transmitter to the point where the reference meter read exactly 3.0 amps each time.

I checked all the meters in my junk box.  Two had open thermocouples and one gave no indication at all.  The rest read between 3.1 amps and 3.65 amps with the reference meter reading 3.0 amps each time.  The meters are like snowflakes and finger prints; to two are exactly alike.  The average reading calculated to 3.386 amps.  That was not surprising, since I am aware that the reference meter reads a little low.

Then I tried checking all the meters with 60~ a.c., using a large resistor and variac, and an Amprobe clip-on a.c. ammeter as reference. Again, I ran the variac up to the point where the Amprobe read 3.0 amps. The rf ammeters read between 3.0 and 3.6 amps, with the average reading  calculated at 3.188 amps.  Apparently, thermocouple meters give a slightly lower reading when measuring 60~ compared to reading rf.  This, despite the fact that some of the meters were stamped with a note that the thermocouple is unaffected by frequency up to 60 mc/s (but they caution that the meters will not work to measure DC). I don't know which reading is closer to correct, since there is no guarantee that the Amprobe is precisely calibrated, either. I tried my Fluke digital VOM in the a.c. amps mode, and it gave a substantially different reading from the Amprobe.

My conclusions are (1) it is very difficult to accurately measure transmitter output power without a recently calibrated professional grade instrument.  Every single one of my meters, thermocouple and the wattmeter, give a different indication.  Those meters make good relative power indicators and the wattmeter makes a good SWR indicator, but I am still  uncertain about my exact output power.  (2) Using the average reading of all my meters indicates a class-C final amplifier efficiency of only about 60%, which is probably not as far off as it might seem.  The tube charts give a nominal efficiency rating of 70% for the HF-300 in class-C, but this does not account for losses in the plate tank circuit. I suspect many hams have an inflated idea about their transmitter output and efficiency. (3) 60~ a.c. is not an accurate way to check the calibration of thermocouple rf ammeters. (4) Given the technical knowledge and measuring instruments at the disposal of to-day's typical ham, output power measurement is a hit-or-miss proposition at best, in most cases giving only a ball-park estimate of what the transmitting power level really is.

[W2PFY]

Maybe some of the error in the RF amp meters is caused by not having them mounted in a steel panel. Some of my NOS RF meters state the thickness of the metal they were calibrated for.

Still a a very good read and great exercise as you indicated.

[k4kyv]

Some are marked for steel panels.  I tried mounting some in a 1/8" steel rack panel with the 60~ test, and it seemed to make very little difference.

D'Arsonval meter movements have a "magnetic shunt" adjustment that can fine-tune the sensitivity of the movement.  I have never tried to re-adjust one, but that might be a solution.  The problem is that I don't have a trustworthy standard to go by as a reference.

The one thermocouple meter that I had that probably was true reading got turned to a block of charcoal from a lightning strike when I wasn't home and had forgot to disconnect the antenna before leaving. Murphy in action! Nothing else was damaged, except for a few light switches in the house that were welded shut. They were old style switches with ceramic body and easy access to the interior, so I was able to pry the welded contacts apart and didn't have to  replace the switches. 

[W2PFY]

One time on my Westinghouse I was testing it out on 160 meters at 1500 watts output. The meters started acting crazy and I looked up and saw a fire inside the RF amp meter. I shut things down but I could still smell something burning. I went outside to discover that my window line was on fire. I have no antenna tuner for 160 meters but with 265 feet of window line, the impedance was around ten ohms and I have coax going out to the transition to the window line. That's where the short occurred and the resulting blown meter and window line fire. If i didn't go out and look at it when I did, it could have started my cabin on fire

[KA2DZT]

Probably one of the best ways to calibrate amp meters is to use a shunt in series with the amp meter and measure the voltage drop across the shunt and compare it to amp meter reading.

I use a lab type Leeds & Northtrup 0.1ohm shunt (I have three of them).  They are a rather large piece of equipment not like the usual shunt that one may be familiar with.

Simply use a good DVM and measure the voltage drop (in millivolts) across the shunt and compare it to the current reading on the amp meter.

I've use this method on both AC and DC.  I believe the readings to be accurate.

I'm not sure that the shunt method would be accurate at RF, but at least it could be used down at 60Hz

I also have a standard cell which delivers an exact DC voltage to 4 decimal places and my HP DVM reads that exact voltage.  Standard cells are an old piece of equipment that years ago required more Leeds & Northrup complicated equipment to measure (I have some of that junk too).  But today, using modern DVM's that have a very high input resistance, I believe you can measure a standard cell directly.

Fred

[W3GMS]

Don,
What about the idea of getting a scope and read the voltage developed across a 50 ohm dummy load and calculate the current based on the voltage accross the measured R value of your dummy load.  I would thing that would be fairly accurate value for the current.  From that you could see how accurate your RF current meter is.
Joe, W3GMS 

[AB2EZ]

I agree with the method that Joe proposes as being a good method for calibrating the ammeter.

A few additional details to keep in mind when you are talking about measuring the power with an accuracy of a few percent (or better)

1. The simple formula we use to calculate the power flowing into a load of resistance R assumes that the waveforms are sinusoids; and that the voltage is (indeed) exactly in phase with the current. [I.e. the load is an ideal resistor with no reactance]

Even a small amount of harmonic content can disturb the amplitude measurement by several percent... and thus change the value of v**2/R by several percent

(v + d)**2 = v**2 + 2vd + d**2

So... if our measurement of the peak voltage is in error by 1%, our measurement of the power will be in error by 2%.

As an example, if the signal is the sum of a sine wave and some harmonic that is 40dB down, the amplitude of the total (fundamental + harmonic) voltage waveform can be as much as 1% higher or lower than the amplitude of the fundamental alone. If the 2nd harmonic is 30dB down, the amplitude of the total voltage waveform can be as much as 3% higher or lower than the amplitude of the fundamental.

If the measurement of the voltage (attempting to measure the amplitude of the fundamental, but actually looking at the sum of the fundamental and, for example, 2nd harmonic power that is 30dB down from the fundamental) is off by 3%, then the calculated power (using the simple formula) will be off by more than 6%.

Stu

[W7SOE]

Don,
    Why not get rid of all your power/efficiency measurement tools except one.  Then you would know exactly what the power/efficiency is.

Rich

If I can't be helpful, I'll just be a smart-ass.

[k4kyv]

I normally just tune for maximum reading of the rf ammeter.  I would like to know more precisely what the efficiency is, and correct any deficiencies in the final if it is abnormally low. Plus or minus 20% of power won't make a perceptible difference to the signal.  It's like agonising over a 1.3:1 SWR reading because you can't get it down to a perfectly flat 1.000:1.000.

I'm keeping the extra thermocouple meters as spares, since good ones are getting harder to find and I have a real knack for f**king things up.

I have never trusted methods like using the scope and resistor.  Too many variables involved (scope calibration and resolution, resistor tolerance, ohmmeter calibration, etc), and the outcome is usually no better than what I started out with.  What I really need is to borrow a meter that has recently been certified as calibrated, and use it as a standard to calibrate the rest of mine.

[W2PFY]

51 watt Fred said.

I'm not sure that the shunt method would be accurate at RF, but at least it could be used down at 60Hz

Some background. My Westinghouse is a SW transmitter, not a converted BC rig and were extensively used by the USAF and the Technical manuals were written for it in Syracuse, NY.

The test procedure for setting up the limiter and and clipper circuits involved a tone generator, a load resistor to match the output of the 833 modulator secondary and a RF amp meter.( Scope too) The test tone was 1000 cps.

The only reason I have posted this post is to let new people to Am know that an RF amp meter can be used to test high power audio circuits. In the service manual they make it very clear that you have to be very careful doing these tests as the HV can kill you

[KA2DZT]

I have a RF ampmeter built in my 51 watt xmtr.  I used a RF ampmeter from a ART-13 xmtr.  The meter itself is really a 250ma movement that works off a transformer made of a few coils in the ART-13.  The scale is 0-5 amps.

I made another xfmr from a 4-turn silver plated coil 1" diameter on a 1" form.  I added a 2-turn pick-up coil that runs to the meter.

I calibrated the meter by adjusting the distance between the two coils.  I used my Bird meter and 50ohm dummy load and set it to read 2 amps on the 0-5 amp scale.  So, a 2 amp reading is really only 1 amp to indicate 50 watts.  I did it this way to get better resolution.  The meter will also read the output current from the 813, second power level of the xmtr.

The 813 stage is progressing along as I have been working on it most everyday.  Can't do much of anything else outside because of the snow.

Fred

[K5UJ]

I'm glad I finally got around to reading this thread; I have been meaning to for several days.  Don, tnx for the great report; it was very interesting, especially the part about efficiency of class C PAs. 

I have one or two RF current meters but I have not yet tried them out.  One is an old military thermocouple meter, around 4 or 5 amps that I got at a hamfest last year for a few dollars.  I am assuming one way to  use it is to mount it in a metal box in series between the centers of a couple of UHF jacks?   Insert it in a coaxial feedline that way?

Don, what is your new and old coaxial feedline?  You may have said once before but I forgot.  I know you want to eventually try some kind of open wire line but what are you running above ground for now?

Rob

[k4kyv]

Don, what is your new and old coaxial feedline?  You may have said once before but I forgot.  I know you want to eventually try some kind of open wire line but what are you running above ground for now?

Cable Xperts inc. direct-burial mil-spec rated RG-213.  Old and new coax off the same roll, picked up at a hamfest some years ago.

[N8ETQ]

Yo'

RG-213... Good stuff indeed.

   In my never ending state of confusion I have to wonder what these
are.  Are they usefull?  Are these the devices my old "Chief" used to
bitch about if we didn't push back in the shunt after taking base current
readings ?

73
/Dan

[W2PFY]

Those are RF meter shunts, now all you need to do is find out how many micro-volts they put out at full scale and then match them up with a meter.

Here is one I purchased from epay. The set was built in 1939. Yours may be older. (I see they were made in 1922) Really Old Buzzard

[KA2DZT]

I'm glad I finally got around to reading this thread; I have been meaning to for several days.  Don, tnx for the great report; it was very interesting, especially the part about efficiency of class C PAs. 

I have one or two RF current meters but I have not yet tried them out.  One is an old military thermocouple meter, around 4 or 5 amps that I got at a hamfest last year for a few dollars.  I am assuming one way to  use it is to mount it in a metal box in series between the centers of a couple of UHF jacks?   Insert it in a coaxial feedline that way?

Don, what is your new and old coaxial feedline?  You may have said once before but I forgot.  I know you want to eventually try some kind of open wire line but what are you running above ground for now?

Rob

Rob

If it's a 5amp meter, take a close look at the face plate,  my 5amp military meters are really only .250ma meters with a 5amp scale.

Fred

[Gito]

Hi.

Maybe the Be Metal used in all these different  RF Ampere meters(different be metals?) has a difference response to the same RF frequency.

I think  RF ampere meter is sensitive to the RF Frequency  of the transmitter.

 A low frequency  make the shunt in the Ampere meter less warm than a high frequency output?
 
It's a simple way to measure the output of a transmitter using a dummy load,Heatkit Cantenna(dummy load)  used 2 resistor a diode and a C  and a VOM

Gito

[k4kyv]

Here is one I purchased from epay. The set was built in 1939. Yours may be older. (I see they were made in 1922) Really Old Buzzard

I have several of those; they all have that date.  That is the patent date, not necessarily the manufacture date.

[K5UJ]

Rob

If it's a 5amp meter, take a close look at the face plate,  my 5amp military meters are really only .250ma meters with a 5amp scale.

Fred

Okay, I went and looked at my two meters.  The first one has a round face about 1.5 inches diameter.  It is a Weston and says "thermocouple type Signal Corps Type 1S-128 model 507."  Its scale is 0 to 3A so I was mistaken about the range.  The other one is 0 to 3A also.  It is a Simpson, much newer and makes no mention of a thermocouple.  Just says "RF Amperes" on the scale.  The box says Model 1337 Cat.No. 03010.

Rob

[KL7OF]

I now have my above-ground coax line installed to link the transmitter to the tuner at the base of the tower. The old buried coax link showed 93% efficiency when I first installed it, using my Mirage Hammy Hambone RF wattmeter.  I know it is off calibration, but it should still be accurate for relative power measurements.  When I re-checked it this summer, the efficiency was down to 75%.

To measure feedline efficiency, I hauled my large 50Ω dummy load down to the tower and connected the coax feed line to it. Then, back at the transmitter I checked SWR, which was something like 1.05:1.  I switched the meter to power output measurement, loaded the transmitter until the permanently mounted rf ammeter I use for reference indicated 3.0 amps, and took a power output reading. Then I removed the wattmeter from the line, went down to the tower, and re-inserted it between the coax and the dummy load, and compared readings between the transmitter end and antenna end of the coax.  I ran the test with both coax lines. Naturally, the transmitter was turned off while the meter was dis-connected and re-inserted into the line.  I got my exercise for the day making multiple trips between shack and tower.

Re-measuring the efficiency of both old and new transmission lines, using the same Mirage meter, both lines showed about the same efficiency @ around 85% this time. So I decided to try measuring efficiency again, replacing the wattmeter with a Weston 0-5 amp thermocouple rf ammeter. The same procedure was used again, with the reference meter reading 3.0 rf amps. With the new coax, the efficiency calculated out to 97% and the old coax to 93%.

The rf ammeter calibration may not be perfect, but it should be more stable and consistent than the Hammy Hambone wattmeter, which does make a good SWR indicator but I wouldn't trust it for power measurements.

While I was at it, I decided to run tests with RF ammeters I have on hand.  Using my permanently mounted thermocouple meter as a reference, I compared readings with the others.  I brought my dummy load back to the shack and connected it in place of the coax to the tower, inserted the meter under test in line, and loaded the transmitter to the point where the reference meter read exactly 3.0 amps each time.

I checked all the meters in my junk box.  Two had open thermocouples and one gave no indication at all.  The rest read between 3.1 amps and 3.65 amps with the reference meter reading 3.0 amps each time.  The meters are like snowflakes and finger prints; to two are exactly alike.  The average reading calculated to 3.386 amps.  That was not surprising, since I am aware that the reference meter reads a little low.

Then I tried checking all the meters with 60~ a.c., using a large resistor and variac, and an Amprobe clip-on a.c. ammeter as reference. Again, I ran the variac up to the point where the Amprobe read 3.0 amps. The rf ammeters read between 3.0 and 3.6 amps, with the average reading  calculated at 3.188 amps.  Apparently, thermocouple meters give a slightly lower reading when measuring 60~ compared to reading rf.  This, despite the fact that some of the meters were stamped with a note that the thermocouple is unaffected by frequency up to 60 mc/s (but they caution that the meters will not work to measure DC). I don't know which reading is closer to correct, since there is no guarantee that the Amprobe is precisely calibrated, either. I tried my Fluke digital VOM in the a.c. amps mode, and it gave a substantially different reading from the Amprobe.

My conclusions are (1) it is very difficult to accurately measure transmitter output power without a recently calibrated professional grade instrument.  Every single one of my meters, thermocouple and the wattmeter, give a different indication.  Those meters make good relative power indicators and the wattmeter makes a good SWR indicator, but I am still  uncertain about my exact output power.  (2) Using the average reading of all my meters indicates a class-C final amplifier efficiency of only about 60%, which is probably not as far off as it might seem.  The tube charts give a nominal efficiency rating of 70% for the HF-300 in class-C, but this does not account for losses in the plate tank circuit. I suspect many hams have an inflated idea about their transmitter output and efficiency. (3) 60~ a.c. is not an accurate way to check the calibration of thermocouple rf ammeters. (4) Given the technical knowledge and measuring instruments at the disposal of to-day's typical ham, output power measurement is a hit-or-miss proposition at best, in most cases giving only a ball-park estimate of what the transmitting power level really is.

To quote you Don, from another thread....."Throw those worthless pieces of crap away and purchase a Bird APM-16 wattmeter.   

[W2PFY]

To quote you Don, from another thread....."Throw those worthless pieces of crap away and purchase a Bird APM-16 wattmeter.   

I think they are great. Not worthless. Bird meters are for you rich folks

[K5UJ]

answered my own question.  turns out the sig. corps meter is a thermocouple meter, S.S. Nebraska wants $60 for one just like it, but the Simpson is a mystery.  Nothing about it on the Simpson website or anywhere else.  I'm pretty sure it has to be used with an external thermocouple.  That's okay, i can do plenty with the other one.

rob

[Gito]

Hi

Thermocouple meters  Schematic diagram and how it works.

Looking at it maybe different meters have a small difference in reading,since it is depended on the heat  generated in the joint end the thermocouple(be metal) ,that generates a voltage proportional  to the current which produced the heat.

since it's really a heat type meter,it depents on the resistance wire,and the be metal(thermocouple

Gito

[cx2ua]

Hello my cordial greetings, very good post on the rf ammeters, suppose that the difference in reading an ammeter to another

may be related to the parasitic capacity with this type of ammeter she can influence the behavior of the line or dummy load,

you will not be 50 but would j0 ohms capacitive component desert.
this is one of several reasons which modern rf ammeters are of the toroidal transformer having very low parasitic capacity etre other advantages.
I would recommend you contact the chief technician of some medium-wave broadcasting in your town to see if you have an rf

ammeter as the Delta series TCA them are factory calibrated at a frequency of 1 MHz, not 60 hz.
Sure to contact the technical staff of some broadcastin taste can compare them with your instrument against the station

broadcasting it all depends on the power to handle this is not clear who set fire to his instrument if the issuer is of the

class of 50 kw.
Remember as the saying "the worst management is not done"

Greetings.

[KB2WIG]

GIOVANNI,

welcome to amFone


klc