The Peak Programme Meter
An article on the development and design of the PPM plus a review of one of Bulgin's meter kits with some astounding Special Offers!
The following article is an explanation of the design of Peak Programme Meters, followed by a look at a PPM kit, the 402/K from Bulgin Electronics (Soundex) Ltd.
Since the inception of mixing desks and recording equipment in general, there has been a need for some kind of recording level meter to monitor the signals being processed. Record at too low level and on replay the system noise consisting of tape 'hiss', mains hum and general amplifier noise will be present. Record too high, to avoid the noise and there is a danger of running out of head room into distortion.
The most common type used is the Volume Unit (VU) meter, however, the difficulty in interpreting the needle's action led to the development of the Peak Program Meter (PPM). This type of meter which was pioneered by the BBC over forty years ago is most valuable for setting the correct recording level and resolving visually the ear's perception of the sound level.
From Figure 1 it can be seen that a PPM has four main sections: a full wave rectifier, time constant, processing amplifier and meter.
The full wave rectifier is required due to the nature of the audio signal. If this was a pure steady tone, half wave rectification would suffice, since both positive and negative parts of a sine wave are the same. However, most audio signals lack this symmetry as shown in Figure 2, and therefore require full wave rectification.
As well as being able to rectify all types of audio signal, the rectifier must also be free from any threshold effects. A simple diode will not suffice because each of the diodes requires a voltage threshold to be reached before it begins to conduct. This would mean any meter connected to such a system suddenly coming alive as the audio level increased beyond a certain point.
To overcome this problem, an active rectifier is used with the diodes placed in the feedback loops of operational amplifiers (see Figure 3). This has the effect of linearising the diodes by removing their threshold voltage for conduction. In addition this type of rectifier can be used down to very low signal levels and has the advantage of low output resistance.
One other important characteristic of the rectifier is that it must have a flat frequency response throughout the whole audio spectrum.
Having obtained a rectified or unidirectional version of the audio signal, it is not enough to simply apply it directly to the input terminals of the meter. In the case of a moving coil instrument, the average of signal rather than peak value, would be indicated and although a light column may indicate the peak of the signal, readings would be very confusing. It may be wrongly assumed that an instant response by the meter is the ultimate goal of a PPM and that this is reflected in the general popular market shift away from mechanical meters to light columns, however, two things should be taken into account when setting the speed of response of a meter, these are: the ear's perception of distortion and the perceived level of pulsive sounds.
Firstly, the ear will only recognise distortion if it lasts longer than a certain period of time. Hence a click will always sound like a click even if it is over the top. Secondly, the perceived level of pulsive sound depends on the duration of the event as much as the real level. A super-fast meter can present totally confusing information. The two totally different signals shown in Figure 4 may be perceived by the ear to be the same level simply because duration as well as level is being taken into account.
One way in which the PPM may be made to take into account the ear's response to pulsive sounds is to be given an attack time constant. A resistor and capacitor as in Figure 5 will do this function. Audio events of sufficient duration will always charge C1 to the voltage on the input side of R1, but shorter duration events, because of the time taken for C1 to charge through R1, will be over before C1 has time to fully charge. The voltage on C1 is therefore a measure of the duration as well as the amplitude of the incoming signal. R2 is used to leak away the voltage stored on C1 and will determine the return time of the meter needle when the audio is removed. A fairly long return time, 2 seconds or so from full scale deflection, does much to relieve eye fatigue when watching the meter.
When audio is not of a pulsive nature, but more steady, the perceived level is almost entirely dependent upon amplitude (within the normal frequency response of the ear). The relationship between perceived and real level is logarithmic, so for example, two real level changes of 10mV to 100mV and then 100mV to 1000mV are interpreted by the ear as being two equal changes, even though the actual range of voltage covered by the second change is much greater than the first. To put it another way, the ear perceives as equal changes what are in fact equal ratio changes (10:1 in each case) as shown in the graph Figure 6.
The temptation is, when faced with this logarithmic relationship, to reflect the perceived level using a linear response meter calibrated with non linear scale markings. The VU meter is classic in this respect but close scrutiny reveals the scale to be far from ideal, since one third of the scale is taken up with a red overload region, where because it presumably means what it says, the needle is not supposed to go; another third where most audio is never consistent enough to remain; and lastly a lower third where the scale markings are cramped and lack resolution, this being part of the scale where most resolution is required.
The PPM overcomes scale calibration problems by processing the applied input voltage to the meter in an amplifier that, as the input level increases, reduces gain logarithmically, see Figure 7. The logarithmic processing is generally achieved piece-wise by a series of straight line approximations that allow minimisation of errors individually at various points along the scale as shown in the graph Figure 8. The actual error allowed at each point of the scale is closely specified, and at important points, despite the scale's apparent crudeness, is as low as +0.2dB.
The 1-7 PPM scale preferred by British broadcasters is shown in the photograph. It represents 24dB in 4dB steps, but there are many other acceptable scales, some of which cover different ranges of levels. It should be said also that the meter movements used for PPM systems are of special design to provide fast reliable response with minimal overshoot. The electrical characteristics of the drive circuits may differ slightly for these various meters but all function to the same basic principles as previously described.
We are not able to present this in our normal project style since certain components of the circuit must be selected and the meter must match the associated electronics. However, Bulgin Soundex have made available to us three versions of their range of PPMs at reduced prices, see the special offer details at the end of this article.
The specification for the 402/K is given in Table 1. From this it can be seen that it will work with a wide variety of supply voltages and covers the complete audio spectrum. The last three categories: overshoot, response to tone bursts and decay time indicate the action of the meter's needle — very little overshoot but a fast reaction to sudden bursts and because of the 1.75s decay, no rapid swinging back and forth that is characteristic of VU meters.
The complete circuit for the PPM 402/K is shown in Figure 9. RV1 determines the sensitivity of the circuit. IC1 is a preamp. IC2 and 3 with the associated resistors and diodes form the active rectifier as described earlier. R14, 15 and C7 form the meter time constant. The needle attack time is approximately 0.6ms and the decay time 1.75s.
IC4 and associated components form the processing amplifier. As described earlier, a series of straight line approximations are made using transistors to form the real and perceived sound levels. In the circuit for the 402/K only one approximation stage is employed. This is formed by resistors and diode D5. Diodes can be used in the same way as the transistors, viz: with an increasing signal more diodes begin to conduct.
The PCB photograph illustrates the straightforward construction of the PPM 204/K and the following text explains the procedure required to build the kit.
Check against the parts list that all components have been supplied in the kit. Identify all the components and observe their layout on the PCB. Note that IC2 and 3 may be supplied as either eight pin DIL or metal can types. Solder the components to the PCB ensuring that the polarity of C2, C7 and the diodes is correct. Also, note that IC2 will be coloured red so ensure that this is inserted in the correct location. R14 and R15 are the two resistor elements of the time constant circuit and to facilitate later alterations to the constant it is advisable to mount these on pins.
Observing polarity, mount the meter on the board using the meter studs, nuts and washers provided. Connect the power leads to pins 24V DC POS and NEG. Connect audio leads, screen to LO INPUT (same as 24V DC NEG) and signal to HI INPUT. The meter will then be ready for calibration.
The sensitivity of the meter is adjustable over the range 0 to -20dB. To calibrate the instrument against an existing VU meter (this being the simplest method since most people will have VU meters) a source of steady tone is required, say a 1kHz sine wave. With the PPM across the output of the equipment sending the tone, the gain of the recorder is adjusted so that its VU reads 0 with the PPM reading -6dB below reference. The reference level at 0dB on the PPM is 0.775V RMS. Mark 6 on a 1-7 PPM is the reference so the PPM is adjusted to indicate 4V4. Note on the Soundex 402/K the -6dB point is marked on the scale. When using 1-7 scale PPMs, the peak audio level is indicated by the 6 mark.