It started out simply enough….
by Jim Sorensen
Someone asked what the maximum level into a telephone company remote line was allowed to be and this is where we are!
To answer the question… +8 dBm. Maybe.
We got to the point where we decided that a more or less “standard” output level for things like consoles may well be +4 dBm and that’s shown as 0 VU on the console meter and one wonders why.
Ray Rayburn, an accepted expert in the area and a member of the AES Standards Committee, wrote in that the correct identifier for a console meter is a VI meter and not a VU meter but that it was a VI meter calibrated in VU and I thought I’d help to muddy the waters.
VI is the legitimate term for a VU meter with a pointer or an LED or LCD display or for that matter a 6E5 vacuum tube. In any event, a VU meter is a VI meter calibrated so the scratches on the scale are 1 dB each which also displays specific response and ballistic characteristics. You could say it’s a specially calibrated VI…as long as you specify what it’s calibrated to. A VI meter is, as Ray shared, an audio power meter that reads in dB and since we can’t measure power directly in any convenient way, we measure voltage across some resistance. We know, for example that 0.773 volts across a 600 ohm resistor will cause the resistor to dissipate 1 milliwatt of power. If you were to hook a suitable meter up across the audio line with a steady signal of 0.773 volts on it, you’d see the meter rise to some point and rest there. If you could draw on the meter face you could make a mark there and call it “1 dBm” which is the accepted designation for 1 milliwatt being dissipated across 600 ohms. In the industry we call it “0” but with a different label meaning that it’s the reference point between two scales…one where more power is dissipated across our 600 ohm resistor and one where less is dissipated. The confusing part is that on the real meter we mark the junction between those two scales as “0” and we have a -1 dB and a +1 dB mark as well. These are read as a change of “one dB below or above the reference value of 1 dBm.” Leaving off the “m” is significant!
This is confusing because if we called it “0 dBm” we’d really be saying that the signal was off and the meter would not read at all. If we called it just “0 dB” we’d be saying that there was no change! Remember a dB is a relative change, not an absolute value. So we call it “0 VU.” It gets confusing when “4 dBm” is equal to “0 VU” because it seems to increase the size of “0 VU” by one something without other reason or explanation. It flies in the face of several “conservation” laws. Perhaps it make more sense when you consider that it’s “0 VU” because with a steady tone there is no change. At “-1 VU” we’ve seen a change of -1 dB and at “+4 VU” we’ve seen a change of +4 dB.
It’s axiomatic that most all reference values are arbitrary and are either negotiated or done because “that’s how we always did it.”
For generations the accidental reference impedance of a telephone line is by definition 600 ohms when measured at 1000 Hz. If you connected your meter with its little mark on the scale across a telephone line and imposed a signal on it until you got the pointer to go up to the mark you’d find that you had a voltage on that line of 0.773 volts and by so doing you have confirmed that the phone line is indeed 600 ohms.
All VI meters that have scales (not the 6E5 tube which offers a relative indication but no scale) are calibrated in dB but the reference can vary. Those that are calibrated so that 1.228 volts across 600 ohms calibrated in VU and 1 VU = 1 dB. It can be seen that this calibration voltage will cause dissipation of +4 dBm or 6 milliwatts. By placing a pad and a 3600 ohm building out resistor in the circuit the meter itself, a 200 uA d’Arsonval movement with a 3900 ohm internal resistance, is calibrated so that when the pointer is at the dividing line between the two scales the signal level will be +4 dBm which will read as “0VU” on the meter face. Thus, the scale mark called +1 VU is actually also +5 dBm and the mark called -10 VU is also -6 dBm and these two marks are 11 dB apart.
If we were to place this meter across our phone line with 0.773 on it and look at it the pointer would not be at the “0VU” mark but would be 4 dB below it.
VI meters come in a variety of styles and are usually found in test equipment, not in “studio” gear like a broadcast or PA console. Usually a VI meter will tell you something about its reference calibration and often you’ll find that they are calibrated in dBm but that does not necessarily make them the same as a VU meter. The reason is that the typical VI meter is made for steady-state tone measurements, and not for the complex waveforms of voice or music which is the purview of the VU meter.
This points to one of the issues with the VI meter. Since the ballistics or the physical characteristics of how the pointer moves…the stuff like inertia, the mechanics of it if you will…of the VI meter permited so much overshoot without registering it on the meter there were often problems with distortion in recording things since these really big excursions often went unchecked if only because the engineer couldn’t see them. The VU meter went a long way to help with that.
In addition, the VU meter was a specific attempt to correlate the “dB” with what was on the meters and a “VU” is, by convention, equal to “1 dB” if you’re looking at more than a sine-wave. It’s supposed to work that way for complex waveforms, too whereas a different “VI” meter might not. That’s another of those definition things…the VU meter is considered to be the only one that reads complex audio waveforms accurately and to conform to the ways people hear the closest.
The aforementioned overshoot in the recording chain was caused by the fact that the VI meter ballistics were such that the meter typically never settled down for the entire duration of what was being recorded. Remember, VI meters typically didn’t settle down during tests using a 1 sec 1000 Hz tone burst whereas the VU meter settled down in one-third of that time and read acurately for the last 2/3 of that one second burst. Essentially that says that a VI meter will overshoot, sometimes by substantial amounts, perhaps in excess of 60 dB, making the VU meter much more suitable for monitoring recording levels.
In any practical sense it does not matter since most “board ops” now in the business have no idea what that meter is for anyway!
The “typical” or “real” broadcast “VU” meter circuit had a fixed pad and a 3600 ohm building out resistor and used a 200 uA d’Arsonval meter movement with a 3900 ohm internal resistance. This meter usually used a full-wave copper-oxide rectifier that was bulit in and they sometimes ground or filed on the pole-pieces of the meter magnet to get the ballistic calibration right. The idea was to calibrate as needed using the pad to get the meter to read “0” VU…the place where the black and red sections meet. Your goal was to measure 1.228V across a 600 ohm load, which is +4 dBm. Most broadcast consoles had a built in pad that was designed to do just that based on the design of their output amplifier.
The VI meter very typically had a calibrated variable attenuator which offered you the chance to combine the meter reading and the attenuator dial reading to get the value you needed which is indicative of a measuring instrument and not a device for monitoring, for example, an analog recording or broadcast program.
By the way, there were two scales for VU meters, “A” and “B”…someone was up all night thinking that up, huh…and “A” was typically for the recording or studio console and had the “dB” values printed on top whereas “B” was typically for the transmitter console, if the station had one. The reason is that on the “B” scale “0” VU is also 100% modulation and the modulation values were printed on the top of the scale. The use of this kind of metering may require additional amplification and calibration but makes it easier for the transmitter engineer to “ride gain” if he has to. Remember that the older plate modulated AM transmitter usually wanted to see an input audio power of +10 dBm. Gosh, that goes back a ways!
Nowdays you’ll find a “VU” meter scale looking at you if you get a console with “real” meters, but you will most likely find them to be Sifam brand or something that’s an electronic voltmeter with a driving amplifier and calibration pot on the back terminals. These are OK and will give you a good reading at least at the “0” indication. Many of them also use the needed ballistics to give you a good reading on voice and music, but not all in less expensive equipment. The ones that sort of scare me are the stacked LED’s that might or might not be a meter at all. I’m going to get in trouble, in the case of all too many consoles for clubs and HOWs use these meters are more an “activity” display than a monitoring tool if they show anything at all. The guys making them know who they are and so do we!
The telephone company is about the only place you might find anyone interested in a 600 ohm balanced termination…other than some very old PA systems or broadcast audio installations…and I’m not actually sure that even Telco cares anymore. In case they do, it used to be that the most you were allowed to put into a Telco circuit was +8 dBm…dBm being valid because the condition is “Telco circuit” which presupposes a 600 ohm balanced wiring configuration as far as us outsiders are concerned.
The last “loop” gear I saw was an A>D converter that accepted anything that even looked vaguely like audio, balanced, unbalanced and at levels that could be all over the place, so I’m not sure that there’s an issue anymore.
That said, yes, there is a maximum level for these digitizing devices that if exceeded will tear your audio up worse than a Youth Minister in the Spirit so you need to be a little careful and following the old standard is always sound advice because for all their “modernity” Telco hasn’t changed all that much. Ask the Telco guy what level you should send…it’s in the BSP (or whomever’s SP) for the adapter and he can tell you and you can measure your signal and turn knobs until you hit it for him.
If there’s a concluding point to this it’s that in our age of modern digital technology we need to be more concerned with other methods of measurement than the old VU meter but it’s nevertheless a good place to learn about the relative nature of audio and about how the dB works. It’s also a good place to emphasize the need for using proper labels for things.
In a way I’m kind of sorry to see some of this fall off the chart if only because I love a good argument! I also have to admit that I liked strolling down memory lane for a bit where the consoles smelled hot and the transmitters shook the earth when you turned them on and when you hit them hard with modulation!
Keep it out of the red!