V-U Meters and SSB Power Measurement

[k4kyv]

The 1958 (second edition) of the ARRL publication Single Sideband for the Radio Amateur states the following in regards to the SSB power limit:

Some of the single-sideband gang who use big tubes have wondered what the FCC interpretation on the SSB power limit is.  We quote from a Commission letter addressed to ARRL:

The following ... may be considered as a presently acceptable method for determining the d.c. plate power input to the final r.f. stage of a single-sideband amateur transmitter:

The maximum d.c. plate voltage input to the radio frequency tube or tubes supplying power to the antenna system of a single-sideband suppressed-carrier transmitter, as indicated by the usual plate voltmeter and plate milliammeter, shall be considered as the "input power" insofar as Sections 12.131 and 12.136(d) of the Commission's rules are concerned, provided the plate meters utilized have a time constant not in excess of approximately 0.25 second, and the linearity of the transmitter has been adjusted to prevent the generation of excessive sidebands.  The "input power" shall not exceed one kilowatt on peaks as indicated by the plate meter readings.

According to globalspec.com,

VU meters measure the volume (intensity) of analog audio signals. They display signal levels in volume units (VU), a measure of average volume level. By definition, a volume indicator (VI) reading of 0 VU describes a 600-ohm resistance in which 1 milliwatt (mW) of sine-wave power flows at 1000 cycles per second (CPS). Most VU meters have a scale from -20 to +3, and rise and fall times of 300 milliseconds (ms). If a constant sine-wave of 0 VU is applied, the meter will achieve 0 VU in 300 ms. Readings on the VU scale are approximately log to the base 10 of the power ratio referenced to that level. For a pure 100-Hz sine wave, 0 VU equals 0 dBM or 0.775-V root mean square (RMS). In practice, however, VU meters contain an isolating resistor so that 0 VU equals +4 dBM for a pure sine wave.

Note the similarity between the time constant of the standard VU meter and the FCC's standard for determining power input for SSB transmitters.

There is a predictable relationship between the d.c. input to a SSB final amplifier and average output power. OTOH, the relationship between d.c. input and p.e.p. output is highly unpredictable, since it may vary widely depending on the voice characteristic of the speaker.

A satisfactory standard for measuring average output power from a SSB transmitter would be simply to use an average reading rf output meter, using an analogue meter movement with the proper ballistics to cause it to have 250-300 millisecond time constant. This would give an indication of the average volume level of a SSB signal, since a SSB transmitter is nothing but a frequency converter that transposes voice frequencies up to the rf range.  This would give a more meaningful approximation to the effective output power of a SSB transmitter, than what is achieved by measuring peak envelope power.

Bird wattmeter specifications do not give a ballistics characteristics or time constant for the meter movement. The model 43 and most other wattmeters available today were designed to measure power of constant amplitude, sinusoidal waveforms. The model APM-16 was designed to measure average power output of complex, composite RF signals with 5% accuracy, but no time constant figures are given.

Average Reading RF Power Meter

[flintstone mop]

It looks like a peak reading, calibrated, LED readout would give any accuracy.
My Bird only shows 20 watts average when I speak into a SSB transmitter, when the actual P.E.P. is 100 watts out.
VU meters for broadcast are fairly accurate but engineers and knowledgeable board ops know to only run the levels about mid-scale. Running right up to the '0' may acutally be +3 or more. That's why broadcast, or any decent audio equipment, should have lotssa headroom for the bozos in the sutdio.

Fred

[k4kyv]

It looks like a peak reading, calibrated, LED readout would give any accuracy.  My Bird only shows 20 watts average when I speak into a SSB transmitter, when the actual P.E.P. is 100 watts out.

The peak reading LED readout indicates peak output power, not average.  It would give exactly the same reading for a 20 watt transmitter with 100 watt peaks that dwell for only a half-millisecond and occur at intervals a half second apart, as it would for a transmitter that puts out a full 100 watts steady carrier.  But we know there is a big difference in power output from those two transmitters.

VU meters for broadcast are fairly accurate but engineers and knowledgeable board ops know to only run the levels about mid-scale. Running right up to the '0' may acutally be +3 or more. That's why broadcast, or any decent audio equipment, should have lotssa headroom for the bozos in the studio.

You want it to hang mostly at somewhere between -10 and -7, and hit mid scale only on voice peaks.  -10 on the VU meter is 33% on the per cent scale, and -7 is 44%.  If the meter is set to modulate 100% with a sine wave tone at Zero VU (100%), the percent scale will indicate modulation percentage.  But the ballistics of the d'Arsonval meter movement will cause it to average out the level, integrated over several audio cycles, so that the meter measures the perceived loudness of the signal, but it won't respond to short duration peaks.  It is this averaged out level that affects the loudness of a signal, whether it is at audio frequency, or translated to an RF frequency in the form of a SSB signal.  That is precisely what the VU meter was designed to do.

If there is no significant distortion in the audio chain, running the audio level as described above will generate peaks that approximate 100% modulation at frequent intervals, while the pointer of the meter follows at a syllabic rate.

So, your Bird is indicating about right.  If you set 100 watts to be 100% (this could be the headroom limit of the final amplifier, analogous to 100% modulation of an AM transmitter), 20 watts is 45%.  Remember, power has a square-law relationship to voltage amplitude.  It takes four times as much audio power to 100% modulate a transmitter is it does to modulate it 50%.  Without a lot of signal processing or overdriving the linear amplifier, 100 watts p.e.p. will give about 20 watts average power.

Bird claims the Model 43 power meter is designed to indicate steady carrier output, and that it loses accuracy measuring the average power of a fluctuating signal such as a SSB transmitter, so they offer the APM-16 Average Power Meter, which is supposed to be accurate to 5%.  I am not sure what is the exact difference, but I know that some rf meters read average voltage while others read r.m.s.  An r.m.s voltmeter will kick upwards about 22% when an AM transmitter is tone modulated at 100%, while an average-reading voltmeter will stand still under modulation.  The same holds true for current; that's why the plate current of an AM transmitter is not supposed to deflect upwards or downwards with modulation. 

"Mean" power is defined as r.m.s. voltage multiplied by r.m.s. current.  RF power meters are basically rf voltmeters calibrated in watts when working into a defined load resistance, usually 50 ohms.  So, to measure mean power, the meter must be disigned to indicate r.m.s., not average rf voltage.  The thermocouple rf ammeter reads r.m.s. current, and the needle will kick up slightly when reading the line current of an AM transmitter when it is modulated.  A diode rectifier type of field strength meter reads average rf voltage, and the pointer should not move when the transmitter is modulated.

Mean power is the same thing as average power, but I have used the term "mean" to avoid confusion with "average" (vs r.m.s.) voltages and currents.

I do not own a Bird 43, so I am not sure how it reacts to modulation when measuring the output of an AM transmitter.  Perhaps someone who owns or has recently used one could tell us whether the needle remains stationary under modulation (with the peak function turned off if it is equipped with one).

That could explain why Bird says the Model 43 meter is not accurate for measuring average power, and therefore it might logically follow that the power reading from a SSB transmitter would tend to be low.

[K5UJ]

I use a 43 to monitor my carrier output.  I look at mod. envelope on the scope to see when I'm near 100%.  When this is going on the bird meter wiggles very slightly.  I'd say the end of the needle moves a couple of millimeters gently.

Rob K5UJ

[Jim, W5JO]

I use a 43 to monitor my carrier output.  I look at mod. envelope on the scope to see when I'm near 100%.  When this is going on the bird meter wiggles very slightly.  I'd say the end of the needle moves a couple of millimeters gently.

Rob K5UJ

The needle of the Bird 43 that I had would move slightly downward with modulation at times.  I also had a Bird 4411 which includes the peak metering circuit and even it had a formula to calculate true peak.

I sold both and bought a Telepostinc.  It gives me a better picture and reads return loss.

[KD6VXI]

I use a 43 to monitor my carrier output.  I look at mod. envelope on the scope to see when I'm near 100%.  When this is going on the bird meter wiggles very slightly.  I'd say the end of the needle moves a couple of millimeters gently.

Rob K5UJ

The needle of the Bird 43 that I had would move slightly downward with modulation at times.  I also had a Bird 4411 which includes the peak metering circuit and even it had a formula to calculate true peak.

I sold both and bought a Telepostinc.  It gives me a better picture and reads return loss.

There is a billion PEP circuits for the Bird now.  Dennis Ostrowski has the best one I've found, single ended and calculates true PEP.  Unfortunately, he now has no interest in marketing it and I haven't been able to find my notes where he sent it to me.  KE7TRP <<MIGHT>> have one of them, I can't remember if he bought one or not.

The 4314 is the PEP reading version of the 43. 

ALL Bird's that I've run across (which, admittedly is a limited amount) have a disclaimer of sorts about measuring AM Pout.  I don't think they are accurate at all unless measuring a single tone, or two unrelated tones at 100 percent modulation.

From what I remember, the Bird shouldn't move much, if at all, when modulated at 100 percent positive and negative.  When starting to do negative peak clipping and other signal mangling, you can get it to do some pretty wild things.  My current homebrew AM TX will key 75 watts and modulate to 750 on the PEP side of the 4314.  It does almost 700 on the AVG side.  (yes, LOTS of controlled carrier).  If I turn off the compressors, limiters, etc in the audio chain, I get more along the lines of 75 modulating to about 77 or so....  Really, the amount of positive needle flicker is about unmeasurable.  On the scope, I have little to NO negative modulation. 

What's really interesting to note is that when running a CB shop about a decade ago, I noticed the high level AM modulated radios would cause the Birds to do more needle flicker during modulation than the low level modulated ones (IE, collector mod would cause increase in Pout, whereas a Balanced Mod radio wouldn't).  PEP values would be darn near the same.

So, there's another monkey in the works....  Signal processing can effect the meter a LOT, as well as mod method?

I did just hook up a bone STOCK high level modulated Uniden AM CB type radio.  3.5 watts into a dummy load, and needle flicker under modulation, but about the same pos and neg.....  The radio I just tested is a Pro 510 XL.  Transformer modulated am radio.

--Shane

[k4kyv]

I did a Google search for the APM-16.  Several companies have them in stock, but none of them give the price.  They all act like it is Top Secret or something.  This tells me that these meters are expensive, which is what I would suspect anyway.  I did find a place in UK that offered used ones at £650 or so, and with all accessories the price was in the £900's!  But sometimes stuff like is way overpriced in UK, especially if it was imported from the US, so they would probably be less costly here.

Long before Bird and others came out with their wattmeters, some hams were building their own.  If the circuit principle Bird uses for the APM-16 is available or can be figured out, it might be easy to homebrew an accurate average-reading wattmeter.

Regarding the Bird 43, if the needle does not kick up when the AM transmitter is modulated 100% with a sine wave tone, that means that the meter is reading average rf voltage, and the average power reading on the watts scale will be erroneous.  Average power is the product of rms voltage X rms current.  Isn't that confusing?  That's why it is perhaps better to use the term mean power instead of average power.

A meter that reads average or mean power rms current through the transmission line, should kick up 22.5% when an AM transmitter is modulated 100% with a sine wave tone.  How is that derived? Take a 100 watt carrier.  At 100% sine wave tone modulation, an additional 50 watts is added to generate the sidebands, so your average power is now 150 watts.  Using the formula P= I²R,  I²= P/R,  I=√P/R.  Since R is assumed to be constant, I is proportional to √P.  With 100% tone modulation, P=1.5, therefore √P=1.224744871, which rounds off to 1.225, or 22.5% above 100.

With voice modulation, it is quite a different matter.  With most human voices, it takes a lot less than 50 watts average power to modulate that carrier 100%, due to the high peak-to-average ratio of the voice waveform.  10 watts would be closer to typical. Therefore, when a 100 watt transmitter is voice modulated to 100%, the average power output is more like 110 watts, and the rms line current kicks up only about 5% (√1.1=1.0488). so an accurately reading average power meter, which is really an rms voltmeter or ammeter, would not show much of a kick upwards.

A thermocouple rf ammeter measures true rms current, but it is so sluggish that with voice modulation, there is hardly any kick upwards at all with voice modulation.  In AM broadcast work, rf current readings are often taken without paying any attention to the program material.  However, certain loud musical passages may kick the average line current up even past 22.5%, if the waveform happens to resemble a square wave.  Modulating a 100 watt carrier 100% with a pure square wave yields 200 watts average power!

[KD6VXI]

I did a Google search for the APM-16.  Several companies have them in stock, but none of them give the price.  They all act like it is Top Secret or something.  This tells me that these meters are expensive, which is what I would suspect anyway.  I did find a place in UK that offered used ones at £650 or so, and with all accessories the price was in the £900's!  But sometimes stuff like is way overpriced in UK, especially if it was imported from the US, so they would probably be less costly here.

Chuck Martin carries them.

Old style, cream case = 399.00

New style, darker case = 599.00 (used).  998 for a NIB unit.

All in US Dollars.

--Shane

[k4kyv]

How many cans of dark paint could I buy for $200?   

It's probably a much lower demand item than the 43.  If the volume of sales went up, they would probably come down in price.  A good item to keep an eye out for on the used market, ePay and hamfests.

Their ads warn that the slugs are NOT interchangeable.  A new 43 with a set of slugs is up there in price, too.

I have seen rectifier type a.c. voltmeters that read true r.m.s. while others read average.  There must be a simple difference but I have not dived into detailed theory of how these types of meters work.

[Tom WA3KLR]

A meter that reads average or mean power should kick up 22.5% when an AM transmitter is modulated 100% with a sine wave tone.  How is that derived? Take a 100 watt carrier.  At 100% sine wave tone modulation, an additional 50 watts is added to generate the sidebands, so your average power is now 150 watts.  Using the formula P= I²R,  I²= P/R,  I=√P/R.  Since R is assumed to be constant, I is proportional to √P.  With 100% tone modulation, P=1.5, therefore √P=1.224744871, which rounds off to 1.225, or 22.5% above 100.

Wrong again Don.  A meter that truly reads average or mean r.f  power should kick up 50 % in value when an AM transmitter is modulated 100% with a sine wave tone!  A bolometer meter like the HP 431 C on an rf generator will demonstrate this nicely.

Thermal-based or high speed sampled multiplication and integration will do the job accurately.  You can get by without having to sample both the voltage and current if the impedance at the measuring point is close to the design value and unreactive. (For a Bird 43 to read accurately, they are assuming 51 + j0 at that point in the system.)

No super simple analog technique will provide an averaging r.f. wattmeter for AM.  An attempt to peak detect the r.f. voltage (or current as a 43 does) will ultimately yield a PEP meter.  The 43 is looking at tx line current and “squaring” by meter scale calibration.  Also included in the scale is the diode nonlinearity.

To provide an average r.f. power meter by an analog solution, the voltage presented to the averaging filter must represent power.  This means that the voltage or current sample made must be sent through an analog “voltage value squarer” amplifier circuit first, not to be confused with a “waveform squaring” circuit.  This is probably the technique for that other Bird meter which I have not studied.  This meter would have an upper modulation frequency limit.

[k4kyv]

A meter that reads average or mean power should kick up 22.5% when an AM transmitter is modulated 100% with a sine wave tone.  How is that derived? Take a 100 watt carrier.  At 100% sine wave tone modulation, an additional 50 watts is added to generate the sidebands, so your average power is now 150 watts.  Using the formula P= I²R,  I²= P/R,  I=√P/R.  Since R is assumed to be constant, I is proportional to √P.  With 100% tone modulation, P=1.5, therefore √P=1.224744871, which rounds off to 1.225, or 22.5% above 100.

Wrong again Don.  A meter that truly reads average or mean r.f  power should kick up 50 % in value when an AM transmitter is modulated 100% with a sine wave tone!  A bolometer meter like the HP 431 C on an rf generator will demonstrate this nicely.

You're right, I didn't state that first sentence correctly.  The RMS current through the transmission line to the antenna kicks up 22.5% at 100% sine wave tone modulation.  The average power kicks up 50%, which, if you read carefully, you will see is exactly what I  said:
"At 100% sine wave tone modulation, an additional 50 watts is added to generate the sidebands, so your average power is now 150 watts.  Using the formula P= I²R,  I²= P/R,  I=√P/R.  Since R is assumed to be constant, I is proportional to √P.  With 100% tone modulation, P=1.5, therefore √P=1.224744871, which rounds off to 1.225, or 22.5% above 100."

In-line wattmeters are actually rf voltmeters or rf ammeters,  with a scale calibrated in watts under the (usual) assumption of a 50Ω  nonreactive load.  So the needle in the wattmeter kicks up 22.5%, but the scale printed on the meter scale is nonlinear (squaring), calibrated to read a 50% increase in power.  But this requires an r.m.s. reading ammeter, not an average reading one. 

No super simple analog technique will provide an averaging r.f. wattmeter for AM.  An attempt to peak detect the r.f. voltage (or current as a 43 does) will ultimately yield a PEP meter.  The 43 is looking at tx line current and “squaring” by meter scale calibration.

A thermocouple rf ammeter is an example of a simple RMS reading analogue meter.  It could be equipped with the squaring scale to make it read watts into a known resistive load. Most thermocouple meters would be too sluggish to measure average power from a voice modulated transmitter, but it would work for transmitters modulated with a sustained tone.  Typical diode rectifier rf meters read average, not rms voltage or current. 

I'll correct that first sentence and the error that follows in the previous message.

[ke7trp]

I have a Bird 43 with the optional pep kit. This is a modified Dennis O PEP kit. The board is signed and dated inside. 

I have used it for 12 years.

I agree. The Bird meter is not the best for measuring SSB power in stock form. I have found that if you hold a steady tone and give the meter enough time, It will read SSB signal as pep or the same as pep.  The example of 20 watts on the bird for the 100 watt pep radio:  HOLD a steady tone and you shall see 100 watts or there abouts.

With my modified meter, I can accuratly measure PEP for SSB and for AM signals.  The RMS of an AM signal works out almost perfectly every time with this meter.  If you take the fully modulated AM reading in the non PEP mod and add it by 2.828 times, You will arrive at the pep reading of the meter when the PEP function is engaged.

After 10 years of using this meter, I can attest that it is accurate and I can arrive at a percentage of modulation by simply reading the unmodulated carrier and the pep power on the meter. When checking against my O scope, It almost always works out to exactly what the Bird is showing.

I say almost as we know that a dirty signal, or reflected power will destroy the meters accuracy.  I have no idea what Dennis did to modify this meter. I wish I did.