South African Music – our musicians, not really but nearly!
Manfred Mann, rather then Manfred Sepse Lubowitz, was born in South Africa in 1940. I would hasten to add that we all like to see him as a South African, like Elon Musc, but truth be told much is driven by influences within the musician’s host country.
Famous as a musician first and foremost, then for the hits “Do Wah Diddy Diddy”, “Pretty Flamingo” and “Mighty Quinn”, then Manfred Mann and now Manfred Mann’s Earth Band was and still is an inspiration to many song-writers.
I have added Manfred Mann to our “SA Music” articles page because he is after all native South African, he did leave South Africa most probably due to political pressures at the time and the article that most gave insight into the Mann we know comes from an interview with Andrew Brel of Keyboard Magazine in 1993.
Oh yes, nearly forgot – Manfred Mann is known for his solo acts on the Minimoog synthesiser. Let that not be forgotten. Please read up on the series based on synth modules used in analogue synthesisers.
And of course, this could not be complete without the famous Bob Dylan track, “Quinn the Eskimo (Mighty Quinn)”
And next, moving on to our top export, “Mchunu’s Bull”. (Coming Up)
The audio synthesiser uses two popular means of modulation, one being the popular VCA or voltage controlled amplifier, also commonly known as a 2-quadrant multiplier and the other, the 4-quadrant multiplier. The 2-quadrant multiplier accepts a unipolar signal, the 4-quadrant, a bipolar signal.
The VCA uses a control voltage (CV) to modulate the amplitude of the signal, this can be anywhere between 0V to max carrier amplitude or Vcc. Bear in mind that the carrier need not be a continuous sine wave of fixed amplitude but also a signal of differing amplitudes and wave shapes.
Above: Kraftwerk co-founder Ralf Hutter used a Minimoog in their highly successful album, Autobahn.
The ring modulator (RM) is also known as the Ring Mixer although purists prefer the term Ring Modulator or 4 quadrant multiplier.
We briefly covered the ring bridge modulators used in SSB and you will notice that the audio and RF modulators amount to the same thing (albeit minus the sometimes tuned LC circuits). The Hugh Davies modulator uses a 1:1 transformer.
The above video is a very good demonstration by an accomplished user in the power of the Minimoog.
Playing around with the side-bands generated in a modulator creates the unique sounds of an analogue RM/synth. As many have discovered, the analogue synthesiser is very much still in demand as can be seen on eBay, with prohibitive pricing. Digital processing has become more the de facto but in tonal quality (and this I really do mean unique to analogue) the analogue VCAs, VCOs and filters have kept the analogue synth very much alive. Price wise, digital is king.
The voltage controlled filter is another very important block in the design of the analogue synthesiser. (Digital control and processing of most blocks within the synth has become more popular because of lower pricing, mass production, comm etc).
The video below covers the history of the Minimoog with some of the artists whom shaped our musical destiny in electronic music.
Passive filtering using capacitors and resistors
Below, figures 2 and 3 show how a capacitor passes high frequencies more readily than the lower spectrum using the Xc or capacitive reactance formula – thus creating high pass and low pass filters in a potential divider network. VR1 would simulate the capacitor C1 in Fig. 2. Note that this is a passive bypass filter.
VCFs are controlled bandwidth frequency amplifiers allowing for high and low frequency cut-off, throughput of a specific bandwidth, notch-filtering and a Q-factor slope control for attentuation.
(Q-factor, or quality factor determines bandwidth of a filter).
Notch filtering is a great way to reduce a frequency which may be causing interference, a deliberate action to cause a specific sound and of course is sometimes used to reduce acoustic feedback. (this is another subject though).
The circuit on the LHS, the High Pass Filter explanation is given below.
All filters will have some Resistance, Inductance or Capacitance component incorporated. A capacitance will pass more current as the frequency rises and in an inductance the opposite is true.
The NPN or BC549C acts as a buffer amplifier with very high input impedance and low output impedance. The PNP BC559C is where the action happens. The PNP transistor acts as an amplifier where the collector and emitter is coupled to a capacitor which varies in resistance according to the control voltage i.e. the more negative Cv the less resistance. As the resistance becomes less the circuit starts passing only the higher frequency spectrum.
The Q-factor or Quality factor of a circuit determines how tight the bandwidth is controlled. Whereas we add resistances into an LC circuit to reduce the quality factor (Q-factor) in radio circuits we need the LC (inductance and capacitance) Q-factor often to be as high as possible to improve sensitivity.
Equalisers used in audio allow for any part of the audio bandwidth to be cut or boosted, usually +/- 6dB or +/-12dB. Parametric equalisers control the center frequency and bandwidth range as well making for finer and more detailed sound.
There is a lot of confusion over dB and the measurement thereof. The Sengpiel Audio website explains the difference between dBu and dBV very clearly and has an online converter.
Filters, VCOs, PLL, Ring Mod, AM, FM all play an integral role in the audio synthesiser. Modern synthesisers have become extremely sophisticated with the use of microprocessing and the use of digital electronics. Indeed, the circuit board may look so much more sparse than the old Minimoog but such is the way of technology.
The VCA – a Voltage Controlled Amplifier which doesn’t amplify.
The VCA, in this case used in an audio instrument such as an audio synthesiser is usually the last block in the chain. VCAs have a control voltage, on peak control voltage the entire signal is allowed through, on zero volts or lower the signal is blocked. VCAs can also be termed voltage controlled attentuators.
The VCA is usually of 2 quadrant or 4 quadrant type, the output is blocked when the input control voltage is at zero. Although a VCA can be made to amplify an input signal, here the VCA is used and predominantly so as an attentuator. 2 quadrant designs have no output when the CV is at or just below 0V, 4 quadrant designs invert the output signal, and with gain set to an absolute value of CV below 0V the gain then rises. Four quadrant designs are used in amplitude and ring modulation effects. VCAs are therefore ideal for envelope shaping. The MOTM-190 can be switched to either 2 or 4 quadrant mode.
VCAs are often used in circuits for compressing, companding (compressing/expanding to improve S/N) and limiting.
DCAs or digital controlled amplifiers are another variant, describing the controller type which varies the gain.
Although years back VCAs were usually of the discrete type; getting to grips now with most VCAs one would be looking at a high quality, multi-bit controlled unit in an IC package.
The VCA above is gain controlled by FET 2N5457 which acts as a variable voltage resistor. The circuit is not unique, many opamps may and will have better performance but this is a typical design where the linear voltage change across the FET ultimately results in well controlled gain of the LM101 (or other).
Another circuit showing similar principles below.
Although many circuit designers are using operational amplifiers in their circuits there are some that feel with a discrete component approach the output quality can be greatly improved. See below, under further reading.
As discussed before, the VCA is used in limiting, companding and compressing resulting in the use of additional controls which vary attack, delay and sustain, release times. (In synth language ADSR) or AGC (automatic Gain Control).
Above we have the 2 quadrant multiplier or amplitude modulator and the ring modulator both using dual operational transconductance amplifiers, the LM13700 with linearizing diodes and buffers.
For the AM or 2-quadrant multiplier refer to chapters 20 and 21 of the application notes which explains the VCA.
Note the additional path in the RM where a resistor (RM) is installed between out and invert input. RC for Vin2 is a potentiometer.
The VCO – the beginnings of a new era in synthesis
The voltage controlled oscillator is a very common circuit often used in conjunction with a phase locked loop circuit. VCOs often use a varactor diode to control the output frequency of the oscillator. They are cheap to build and are often used in modular synthesiser designs.
This page is not necessarily a continuation of the building blocks in a synthesiser as the VCO in it’s own right makes for a very versatile piece of test equipment.
The VCO can have a control voltage which is fixed to get a single tone or the control voltage can be that from a modulation input which causes shift in pitch synched with the input control.
The VCO in a synthesiser is normally tuned to one of either:
Volts per Octave, the common standard used in Moog where e.g. in sequence each volt equals one octave. This is currently the more popular configuration;
Hertz per Volt, commonly found in Yamaha where each octave equals double or half one Volt.
The VCO is very versatile and in modular form will have a multitude of outputs, usually being ramp (sawtooth), triangular, sine and square. VCOs can be connected together through the control voltage, the outputs can be wave-shaped and of course, modulated from an external (or internal source).
One of my favourite circuits is brought to you by Analogue Designs, using the AD654
In a previous article we cover ring-bridge modulation in SSB transmitters to get DSB-SC (double sideband suppressed carrier). Of course not to be outdone it didn’t take scholars of audio synthesisers to add ring mod effects to their synthesisers. But where to first?
Harmonics, pitch and loudness
The first synthesisers were of the analogue variety and yes, to get an audio signal out we will always need some sort of analogue contraption to amplify the signal. Single note synthesisers are no fun of course. The timbre of an instrument is caused by harmonics, pitch and loudness. This can distinguish the difference between a violin and guitar. The lowest note on the piano keyboard is A0 set at 27.5Hz. Middle C (C4) is 261.626Hz and the highest note is C8 at 4186.01Hz. A4 is set at 440Hz and the tuning reference note.
Tuning forks were long used for the tuning of musical instruments because of the high energy tone at the fundamental. Concert pitch A440 tuning forks are therefore very common.
Although there is much to be said about the sophistication of the circuitry used in audio synthesisers, the stability of the oscillators, filters and amplifiers are infinitely more critical in the design and final result.
Polyphony and Homophony
Polyphony is a term used to describe the use of multiple voices or the ability to play multiple notes at a time. The single tone multiple frequency oscillator is an example of homophony. The Minimoog is an example of a monophonic synthesiser.
Of all the scallywags that make up the “voice” of a synthesiser the most important is that of “timbre”. The Minimoog is a classic in that it can reshape, amplify, filter and process the signal to get the most uncanny and bizarre effects. Often the intent to build a multiple oscillator synthesiser becomes a bad choice when one realises that all the effort is lost in the final result – lack of timbre. This is not say that the builder is incorrect, the intent may be to focus on polyphony with fixed frequency tones.
There are many ways to modulate a carrier but possibly the most important one in the scope of this article is the ring-bridge modulator. Ring bridge modulators are famously used in SSB suppressed carrier transmissions and of course, the synthesiser. By applying a 440Hz signal to the modulator input and with a 1kHz carrier input the resulting double sideband is that of 1000Hz + 440Hz (upper sideband) and 1000Hz – 440Hz (lower sideband) with the carrier at the output being suppressed.
But of course there is more to this. Your frequencies can be ramped (sawtooth), square, in fact anything you like.
Some of the best known oscillators – for the beginner
For those interested in electronic circuits it’s almost impossible to stay clear of the humble oscillator. In it’s simplest form the oscillator generates an alternating current from direct current. It is used most often in radio receivers and transmitters, digital circuits (possibly timing, clock), audio effects and of course in function generators where the output is almost universally sine, sawtooth and square waves.
Relaxation with an Oscillator
In my opinion the easiest oscillator to make would be the free-running astable multivibrator using two transistors or using the 555 IC. The output is square wave and is usually running in the audio spectrum (but not always).
Astable (running free)
In the interests of this page let’s look at the 2 transistor Astable multivibrator. Why astable? There are 3 basic configurations used in multivibrators: astable, bistable (which has two stable states duh!) and the monostable. Guess which one only has one stable state? The astable multivibrator is free running with a square wave output. It has two transistors with a positive feedback path made up of the cross connected capacitors, Q1 collector to Q2 base and Q2 collector to Q1 base. If Q1 conducts first due to possibly having a higher gain than Q2, capacitor C1 is charged through R2 which forces the Vbe of Q2 to rise until conduction which in turn charges C2. This process makes up the basics of an oscillator. The positive feedback is what enables these transistors to switch on and off, components R2, C1 form the time constant for Q1 low state (in this case) and R3, C2 form the time constant for Q2 high state (in this case).
Although oscillators come in many shapes and forms audio oscillators use predominantly resistances and capacitances (RC) in the tuning circuit compared to radio frequency oscillators which use inductances and capacitances (LC).
Positive versus negative feedback
Now in the real world when designing an audio amplifier the last thing we want is positive feedback which can cause the amplifier to go into oscillation. Positive feedback is when the output is fed back into the input but in phase. Negative feedback is out of phase.
In radio telecommunications it is guaranteed that the student will have the Colpitts, Hartley and Crystal Oscillator thrown into their studies. These are extremely popular due to their high performance, the Xtal (Crystal) being the most superior but locked to resonance of the Xtal.
Note the Hartley oscillator split inductance / capacitor configuration as opposed to Colpitts, split capacitance / inductance.
The crystal oscillator is a high performance oscillator, highly stable and very accurate. Although Xtal oscillators boast temp stability, this is not entirely true. Early Xtal oscillators were mounted in temperature controlled ovens to reduce drift. These oscillators were often used in PLL circuits as the reference oscillator.
Operational Amplifiers demystified – starting with the uA741
As a writer of some of the articles here we do try to keep the topic off common ground, throw in a bit of humour but give the reader some insight into components and circuit design. Op-amps we treat no differently.
Today we look at op-amps in general and specifically look at the brain child in 1968 of British Fairchild engineer, Dave Fullagar, the incredible 741 op-amp.
In simple terms an operational amplifier is a DC coupled voltage amplifier which has a differential input and a single ended output.
GAP/R George A Philbrick Researches – currently Microchip
The world’s first commercial op-amp was known as the K2-W, manufactured from 1952 until 1971.
The first tube op-amps were known as general purpose, high gain, DC coupled, inverting feedback amplifiers. Although many experimenters see these devices as being rather simple it’s only when one looks at the GAP/R P65, an early discrete component solid state op-amp designed and built by Alan Pearlman, co founder Nexus, that we note the complexity of the circuit. Pricing for the later P45 was in the region of U$120.00 each.
The company veered clear of Germanium transistors and only started pushing solid state once silicon transistors proved reliable.
About the time of GAP/R startup Robert Page Burr and Thomas Brown started a company, Burr-Brown (of great audio fame) which was purchased out by Texas Instruments in 2000. Burr-Brown is known to manufacture some of the best quality operational amplifiers used in analog circuitry and DSP / electronic signal processing.
So what’s the interest in Op-Amps?
Well, recently we ran an article on the not-so-dreaded 4558. I mean what’s the point of shooting down a 1970s chip in 2017 for poor performance? An Op-Amp has made out life easier, it’s made designing cheaper, it’s made audio simpler and there’s very little one cannot do with it. And it’s analogue. Well mostly, of course.
We can multi-vibrate, add, subtract, differentiate (or), compare (comparator), integrate (or), filter, chop, amplify, invert and …. the list goes on. Op-amps are used in digital circuits as well as analogue. Op amps are temp stable. Gain calculations are simple, in the op-amp (741) image above the gain is equal to R2/R1. The output is in phase. Input to the (-) negative terminal inverts the phase.
What more can one want from a device which in open loop mode (without feedback resistors, gain can be anything from 10 000 to 100 000 (or higher).
Many manufacturers now give open loop gain for an op-amp as V/mV. E.g. 10V/mV = 10 * 1000 or 10 000. The 741 shows a typical open loop gain of 200 000.
Some historical and other facts
Tube negative feedback devices were in use pre 1940s.
The op-amp 709 was plagued by issues, instability, latching and self-destruction was not uncommon. The 741 was the fix.
The uA741 was commercialised in 1968, possibly the largest selling op-amp. They are also amongst the the cheapest – about R8.00 at Yebo.
The opamp cannot live in a world without feedback.
The input (voltages) will try to remain the same through the feedback path, if not, the output will swing towards the polarity of the most positive input e.g. if invert is high, then the output goes negative and vice versa. This forms the basis of many configurations.
In this article we cover basic modulation and tone generation circuits using the 555 IC for the DIYer. Many synth based projects are based on the 555 IC and young hobbyists are encouraged to read, experiment and play with these fascinating chips.
The generation of electronic signals which are converted to sound is possibly the easiest description of the synthesiser (UK) or synthesizer (USA). Being pedantic this would also describe an electric guitar which is not a synthesiser, or is it?
Moog and DX7 – analogue to digital
Most people in the music industry would have heard of Moog and the Yamaha DX7. When we think of Moog we think of two people, Robert Moog and Rick Wakeman. I would like to describe Rick Wakeman as possibly one of the most technically advanced musicians of our time, an absolute legend. Youngsters of today should switch off the rap n crap, pop and listen to The Six Wives of Henry VIII and Journey to the Centre of the Earth. Wakeman has played with the big names in industry, including Cat Stevens. Oh, Yes!
As an aside but big news if you are a Trevor Rabin fan (ex-Rabbitt, Yes) we have the new formed group, ARW – Anderson, Rabin and Wakeman. AWR Tour.
Robert Moog was king of the music scene in the early 60s with his analogue synthesiser development which was used by most of the big bands at the time. Where Moog was king of analogue, Yamaha became king of digital with the DX7. And no, Moog’s analogue models are still popular today and fetch a pretty penny on eBay. Look for one of his most famous models, the Minimoog of the 70s.
What’s in it?
In our previous articles on Ring Bridge modulation and FM Receivers it’s no wonder that it didn’t take engineers long to start using FM synthesis to generate unique sounds. Opportunities create wealth, right?
The common modules found in synthesisers are:
VCO – voltage controlled oscillator
LFO – low frequency oscillator, usually sub hearing <20Hz
VCA – voltage controlled amplifier, 2 quadrant multipler
VCF – voltage controlled filter
Gates – not a module but the electronic switching
Arpeggiator – sequences the notes of a chord (minimum 2 notes). Up, down, up and down, randomise- after defining a chord.
Sequencer – Plays notes in a series, mostly user defined.
Ring Bridge Mixer – 4 quadrant multiplier, suppressed carrier AM
Building blocks – modules
Music synthesisers form great building blocks for DIYers and programmers. The initial build could carry for example two VCOs and a ring bridge mixer with additional modules added later to generate more effects. (I include unique tones, even noise here as an effect).
Your hobbyist free standing synthesiser with out any form of midi controller input would not need to follow any specific design criteria if the objective is just to get a unique sound, possibly generated through voice (vocoder). Your synthesiser modulation or carrier wave does not have to use sine waves either – triangle, sawtooth and square with variable mark space ratios are indeed great for changing frequencies to get unique sounds.
The Nuts and Bolts – FM Synthesis
Note that the circuit above was only simulated on LT-SPICE XVII and not tested. For best results see TI literature on the 555/556 chips.
Not intending to duplicate what one can find on the internet in droves, write-ups and warnings of sophistication on FM Synthesis striking fear into the hearts of many. It’s not to be, this is a DIY page, the hobbyists page. Buy a DX7 if you want but when fiddling there is nothing like having a Function generator on your workbench. Since you are working with audio frequencies, building your own 10Hz to 100kHz FG is not expensive neither difficult. You will need the frequencies to be stable though.
See Yebo Electronics – Audio Function Generators Kit52S – I have not built this circuit. Ask the shop assistant to check the diagram before shipping. Sometimes it’s not that legible. Pricing around R450.00 ex shipping.
The most practical circuits to build when experimenting usually incorporates the 555 or dual 555, the 556 from Texas Instruments.
Every budding young DJ has visions of grandeur, exposing their talents to millions of listeners throughout the globe. Strangely enough older technology did lend itself easier to this pre-Shoutcast boom if one had the cash to purchase a high powered AM transmitter. The quality of sound would have been pretty awful though and the weird beards would scratch your eyeballs out because of the inefficiencies in this technology.
Shoutcast Vs FM Radio
FM Radio could well be the cheapest way if it had the range. A good quality rig could be had for about R10 000 but then we only have line of sight reception so now we need to be looking at investing in repeaters. Streaming and Shoutcast has made international reach well possible but those interested need to know the hidden costs. It doesn’t come cheap!
FM transmission is nothing new. FM does not suffer from static interference because any amplitude distortion is removed at the receiving end. (using limiter circuits and dbx). Or at least that’s the theory. Because of the higher frequencies we have a wider bandwidth available for audio which although acceptable is certainly not high quality. FM transmissions carrying audio needs the signal to be pre-emphasised, that being to boost the high frequencies before radiation and then de-emphasised at the receiver. The more modern stereo broadcasts consist of left and right audio channels modulating a 38kHz double sideband suppressed carrier. (DSB-SC). This 38kHz frequency is regenerated in the receiver. A 19kHz pilot tone (1/2 the 38kHz) in the transmission is used to define and correct phase shifting at the receiver. Like the NTSC television standards where colour should not create problems on monochrome receivers (or PAL for that matter), mono FM receivers would not be affected by stereo broadcasts.
South African content
Much has been said about the SABC COO Hlaudi Motsoeneng’s decision to air 90% South African content, much like the blanket ban the National Party had on the airwaves upto 1994. I was fortunate enough to travel overseas through the 80s and my God were we oppressed. Who could forget the best station in the world, rock nogal, WPLJ in New York. In South Africa the closest thing we had to a good radio station was LM Radio (Lourence Marques, now Maputo), which yep, you have it, wasn’t South African but in Mozambique. LM Radio then became Radio 5, then 5FM. Although claiming to be the best, it’s still a far cry from some of the USA’s stations.
The question as to whether 90% local content has improved anything is up to the believer – I am of the opinion that it’s total madness. Does anyone know how much revenue a rock n roll station in this country would generate? And isn’t this what it’s all about. Watch driver and passenger faces when Ryan o’ Connor of KFM fame in Cape Town plays naughty and spins a U2 track. How often do we hear U2 these days? Nada. Good for the soul? But enough about the music and the stations, why FM?
But why the interest in FM in any event?
Most folk starting out in the electronics field will be interested in building their own transmitter at some stage to understand the basics and beyond. FM transmitter kits are abundant and give us a better insight into the different stages used. Although most are mono there are stereo kits available. Most home built kits are prone to frequency drift. The big people’s kit will have PLL correction and virtually no drift. And yes, FM has another use as well, in music synthesisers. Hetrodying is a common term used by radio people where two signals at different frequencies are mixed to get an intermediate frequency. This is also commonly used in synthesisers which is where we are headed.
PLL or Phase Locked Loop
A relatively new concept in terms of evolution in radio circles, this circuit comes in many forms. The simple explanation is that of a variable frequency oscillator (usually a VCO or voltage controlled oscillator) which has it’s output phase compared to that of a reference oscillator in a phase shift detection circuit. Any shifts in phase results in a low frequency or DC output which is then applied to the VCO causing minute changes to oscillation frequency and locking the output to the input. The VCO output is usually frequency divided to have an output many times lower than the input and at the same frequency as the reference. This means a small shift in phase has a much finer control over the VCO output.
Many years back all AM transmitters were crystal controlled which amounted to the user having to switch between crystals to change frequency. The PLL circuit allows for synthesising, a much more practical and cheaper solution. Here the PLL could be made to change an entire band e.g. 4MHz to 8MHz and allow for incremental steps, accurate to a few Hertz. The PLL circuit makes FM at it’s higher VHF range of frequencies 88MHz to 108MHz more accurate and stable. The output frequencies could be read off through a frequency counter.
Frequency modulation does not only have it’s merits in RF work. The Synclavier Digital Synthesiser was a very interesting piece of work and used FM synthesis to generate distinctive sounds. More about this later.