Class D – let the Battle Begin

Behringer NU12000

Class D – the chip wars

Years back some of the most brilliant scientific minds got together and decided that it was impossible to split the atom. Of course we know what happened in Nevada, Hiroshima, Nagasaki and Bikini Atoll which put paid to that theory.

Behringer NU12000
Behringer NU12000 (2 x 6kW into 2 Ohms)

Likewise in audio circles it was decided that although class-D was feasible, the quality would remain so bad that it would or could only remain a classroom project. Some things have changed since that argument.

Whilst looking around for circuits and projects on the internet which would satisfy my class D curiosity bug I came across a lot of negativity, especially from the audiophile community. Their argument would be that Class A (or SET) is the closest one would come to real quality, the class B approach (actually class AB) is not too bad except for the inherent crossover switching noise and then all things considered SET and then tube push-pull was the only way to go. Class D was not even a consideration.

To prove them all wrong and not wanting to wait for confirmation by doing the intelligent and sane thing, a listening test, I purchased the Behringer 6000 amplifier (without DSP).  I was not disappointed. I do believe that class D amplifiers do not exhibit the performance of class A moving up the frequency spectrum. In my case the higher notes seemed to be lacking somewhat compared to a Proton 1150 which I had laying about.  Driving Peavey 15″ bins with separate tops improved matters. But the powerful gut wrenching bass is what appeals to me and to many DJs out there.  Was  it a good purchase? Bang for buck you won’t get better. Many people use them in their home theatre systems.  Most probably they all live in small apartments as well.

Class D amplifier design and builds are not for the weak though.  Building from a schematic is one thing, designing and constructing a working board another thing entirely.  Although home tinkerers have used 555 timer ICs for sawtooth generation getting  poor to mediocre results, moving up a notch becomes a nightmare.  For most though, fighting with switching noise, EMI, killing MOS devices because of trigger and timing issues and distortion  is a reality.  But kudos to those that have persevered and for especially the DIYer this is a huge feather in one’s cap.

From a marketing perspective though, electronic engineers have moved away from the analogue and linear realm into ultra high speed switching (for audio), digital, DACs, supply PWM regulation and of course class D. Thirty years ago 1/2 and H-Bridge topology was a luxury, an expensive entrance into motor drives, switching supplies and even then, switching amplifiers. Now it’s as commonplace as a barbecue on a Sunday afternoon.

A quick word on modulation:  PWM versus ΔΣ (delta-sigma)

There are many that prefer the delta-sigma (or sigma-delta) modulation process over pulse width modulation but truth be told, the PWM platform is far older and more mature. Another topic, for another day of course, is that fundamentally there should be no difference in audio quality using either topology.

I liked the design in All About Circuits where the author deliberately uses PWM over delta-sigma. The Hi-Lo side driver chip used is the exceptional IR2110 from International Rectifier.  Stage by stage explanation.

Delta Sigma is the process commonly used in ADC or DACs, in class-D we use this process to break down an analog signal into pulses, commonly termed Pulse Frequency Modulation.

Class-D chip manufacturers

Companies such as Texas Instruments have been designing Class D power stages for many years now. They have got it right but they keep on improving. Less cooling, lighter power supplies and now very cost effective chips for  the mass manufacturing of home theatre and stereo systems.  I would be lying if I said that Class D is better than class AB, because theoretically it shouldn’t be by following the straight line rule. But it has unbeatable advantages, cost, space and power efficiency valuable to cell phone and hearing aid technology.   Some say the clarity is not as good as the older and more inefficient systems but in a country like South Africa where the lack of manufacture is a disgrace it certainly keeps shipping costs down. (weight).

The beauty about Class D is that we now have cheap good quality audio chips, using a minimal of components we have a small desktop amplifier capable of lighting up a handful of loudspeakers but the con is sadly the power supply. PC switching supplies have become very reliable and even a good quality properly rated 400W device costs less than R500.  Off the shelf 35V-0-35V supplies in the region of 400W cost 3 to 4 times this, leading us back to toroidals and EI transformers.  Although there are companies in South Africa designing and building SMPS, their expertise usually lends itself to units in the kiloWatt ranges, more margin and possibly bigger demand.  For us to grab all the advantages of class D we need SMPS,  cheaper SMPS. Reverse engineering PC supplies is an alternative but only for the more experienced.

Texas Instruments

Which brings us on to the TPA3255 Class-D Pure-Path chip, the David of Goliath fame, the mammoth power IC from Texas Instruments. Claiming 300W per channel stereo and 600W  mono parallel bridge tied at 10% THD+N and at only about $10 each this is unbeatable value.

The TPA3255 Class-D (Texas Instruments)
The TPA3255 Class-D (Texas Instruments)

STMicroelectronics – the TDA7498E

Datasheet here – a 160W + 160W Class-D amplification chip.


TDA7293 and TDA7294 Chipamps

TDA7293 and TDA7294

The Chipamp builds continue: The STMicroelectronics TDA7293 and TDA7294 power amplifiers

Quick Spec

Both these audio chips use DMOS or double-diffused metal–oxide–semiconductor output stages.  Supply rails are +/- 60V DC max for the TDA7293 and +/-50V DC for the TDA7294, transformer supply rails AC 32V- 0-32V to 35V-0-35V, recommendation is  32V-0-32V for 8 Ohm operation.

TDA7293 and TDA7294
TDA7293 and TDA7294

Power output is between 50 and 100W, depending on which chip is used, loading and power supply rail voltage.

The amplifiers can be run into 4 Ohm load, recommended +/-30V DC or 20V- 0-20V AC secondary.

THD is 0.1% 1kHz at 50W into 4Ohm. Slew rate is about 5V to 10V/uS, thermal protection cut out at 160 deg C.  (Mute at 150 deg C).

Depending where these chips get purchased, prices vary from R100.00 upwards each, the 7293 being more expensive.

Both these audio chips are used in circuits where the configuration calls for bridging and parallel.  Some of the output powers are absolutely ludicrous, anything from 200W for one chip into 2 Ohms to 700W into 4 Ohms.  This is not going to happen folks.  You either bridge and keep to 8 Ohms using supply rail voltages designed for this or you parallel and drive higher outputs into loads smaller than 4 Ohms. Or of course (not a good idea), you parallel and bridge and drive huge currents through a 1 Ohm load. But never 700W into 4 Ohms.

Functional diagrams from the application notes:

Just like most chip amps these devices use a minimal of external components:

High Efficiency TDA7293
High Efficiency TDA7293 – 80W into 8 Ohms.

The application notes are very good, the circuit above is included as well as a PCB layout.  The more common >100W into 8 Ohms is shown below, a bridged mode amplifier using the TDA7294 but the more common 7293 can be used as well, just keep to below spec limits.

Bridged TDA7294
Bridged TDA7294

Bridging amplifier chips is a very common configuration but DIYers need to know the limitations.  As in the circuit above, running load values below 8 Ohms without the additional cooling and PSU capacity is inviting trouble.



Total Harmonic Distortion meter

Hantek DSO

THD Meters – how they work

Many years back the must haves for anyone building audio equipment would have been the multimeter (DVM), oscilloscope and a function generator. With specialised audio test equipment falling in price over the last twenty years two other pieces of equipment are now also found on the test bench, the THD meter and spectrum analyser.  Indeed your laptop or desktop could be used for many of these functions but it can be cumbersome and/or inaccurate because of sound-card bandwidth limitations.

Hantek DSO
Hantek DSO – 1202B

The THD Meter is definitely a must have, even something for impressing one’s audiophile friends.  Consisting of an oscillator, internal or external, a notch filter to attenuate the fundamental to as little as possible and a voltage / instrumentation amplifier to drive the resultant combination of harmonics and noise to the output meter or video graphics display.

Block Diagram of THD Meter using Wien Bridge Notch Filter
Block Diagram of THD Meter using Wien Bridge Notch Filter

Interesting DIY angle to THD metering from “Wensan” – DIY Audio – Simple THD Meter

Robert Cordell,  American (as in USA) electrical engineer and expert in the audio field kindly publishes his high end THD Analyzer circuit on his website, Cordell Audio.

Analysing the Waveform – skirting the FFT

The analysis of signal waveforms is rather mystical to most of us and becomes a highly complicated mathematical subject when breaking a signal or rather batch of frequencies down to a function of time. FT, or Fourier Transform converts the waveform data in the time domain into the frequency domain each containing individual signals of phase, amplitude (magnitude) and frequency.

Modern computer software makes this previously time consuming and highly mathematical task accessible to the home experimenter using what is now commonly known as FFT or Fast Fourier Transform.

Digital Signal Processing

Digital signal processing in the modern computer has made decomposition of signals in audio a common discussion point on the forums, possibly not always in our best interests because of the complexity and hence confusion.

FFT and Nyquist are of interest to the engineer, especially now in the DSP domain where bit rates and sampling frequencies have become an integral part of our lives.

Our modern (and even older) hardware can run software for analysing and changing the audio signal – and sometimes it’s free. See Audacity and it’s features.

Shit in Shit Out

Technically speaking it is not probable that an ill designed amplifier using inferior components will sound good. 🙂

It is probable that with the assistance of application notes, spec sheets and white papers from most manufacturers to build a quality product with the minimum of fuss.

Hantek  – USB oscilloscope, multi-function

I have four oscilloscopes, three of them analogue. I like the ease of triggering and set up. Coming from the analogue era certainly assists – the digital scopes don’t appeal from a GUI perspective until one needs to record an event or compare functionality to price. They become indispensible. Just look at the hand-helds.

The digital scope I use is the Hantek 3062. It has a logic and spectrum analyser, frequency counter and of course it’s a scope.  It cost about R4K five years back from Kmeasure, based in Pretoria. Very good service. I mention this as an aside, there are always sensible, cost effective solutions in our pinched South African market.

Lastly, of course we need a good quality function generator.  I find Instek to be of exceptional quality and reliability. Available at Mantech Electronics. (Cape Town, Johannesburg, Durban).

Further Reading:

Cordell Audio:  Robert Cordell Build your own THD Meter

KMeasure – Professional measurement systems (based in Pretoria)

Signal Lab – SigView Signal Analysis software

The VCF or Voltage Controlled Filter



The Wien Bridge Oscillator

The Merits of Wien Bridge Oscillator types

The Wien Bridge oscillator is an oscillator type used in many circuits which rely on a sine wave to drive or act in part on a circuit designed for another purpose.

The Wien Bridge oscillator is possibly one of the more versatile and easier to build circuits, used often to generate a 20Hz to 20kHz sine wave for testing audio equipment.

The Wheatstone Bridge

But first of all, let’s look at the Wheatstone Bridge, another very important circuit and used to accurately measure very low (or high) resistances.  Most Ohm Meters or resistance checkers are not accurate when going down to below a few Ohms and the Wheatstone Bridge is the solution.

Wheatstone Bridge
Wheatstone Bridge

The circuit above works on the theory that if the potential difference across a Volt meter (galvonometer, a device which measures electric current), there would be no deflection of the needle as no current flows through the meter.

Now if there are two known resistors in circuit plus a carefully calibrated potentiometer, VR1 we can use the above circuit to calculate an unknown resistance.  If this was a device used to measure very low resistances, in the order of 0.1 Ohms we would have a current limiting resistor in the circuit, possibly in the (+) line of the PSU.

E.g R1 = 4.7 Ohms, R2 = 2.2 Ohms and the meter shows no deflection when VR1 is set to 4 Ohms the formula above would give an output, read off the potentiometer indicator as 1,87 Ohms.   Read up on Kirchoff’s Law.

This circuit is used for impedance measurement as well. Below is another variation, the Wien Bridge:

Wien Bridge (for capacitor check)
Wien Bridge (for capacitor check).

The Wien Bridge, known after Max Wien is similar to the Wheatstone Bridge where the potential divider networks need to have an equal ratio to null the network. This is often used to determine, capacitive, inductive (Maxwell-Wien), frequency and resistive values. In an AC circuit things become more complicated and a damned sight more interesting than DC. Both capacitive (Farad) and inductive (Henry) values need a known frequency to determine the capacitive or inductive reactance, resp. Xc or Xl. Xc = 1/(2πfC) and Xl = 2πfL.

In the circuit above there are two variable components, the capacitor C1 and VR1 in parallel. The component being checked is a capacitor which also has a high internal resistance.  By adjusting VR1 and C1 a null will be reached where the values of the two variable controls are read off and the parallel impedance calculated, and likewise the values of unknown components calculated through the null value balanced equation Z1/Z2 = Z3/Z4.

The Wien Bridge Oscillator

Basic Wien Bridge Oscillator
Basic Wien Bridge Oscillator – note the series R1C1 and parallel RC networks R2C2

The Wien Bridge oscillator is based on the bridge circuit.  The circuit above is the classic student’s build using a small incandescent bulb in the circuit to control the amplitude. (AGC).

The capacitors and resistors are of the same value. The result of this is C1 = C2 = circuit C and R1 = R2 = circuit R.  Frequency of oscillation is calculated as being 1/(2 * π * R * C).  Rf  is the positive feedback resistor.

A stable oscillation output is determined when the bulb filament resistance is equal to the feedback resistance / 2 or Br = Rf/2

Sine wave output is generally of good quality, undistorted and even amplitude. Limitation is generally only the lower frequency range, making it good for testing audio equipment.

Next:  Total Harmonic Distortion meter for audio equipment





Tripath Amplifiers – are they analogue?

Lepai Class T 20W per channel

Tripath amplifiers – analogue switching amplifier TA2020

Note: As of 2007 Cirrus Logic currently own the original Tripath intellectual capital and as a result the companyTripath is no longer in existence. The Lepai and the legendary TA2020 are still available. Lepai and Tripath were not the same company. Lepai, Lvpin is now sold as Lepy. There are three chip ranges that you should be aware of:  2020A (Yamaha) 2020+ (Tripath) 2020TI (Texas Instruments).

Class D / Tripath class T

Class D – super efficient, mobile friendly to head stomping 12kW power amplifiers.

Lepai Class T 20W per channel
Lepai Class T (registered Trademark) 20W per channel – eBay or Aliexpress

Going back a few steps we need to look at the original class-D designs. The first attempt at a commercial “working” Class D amplifier came through Sinclair Radionics, brainchild of Sir Clive Sinclair, genius inventor and entrepreneur.  The commercial failure of this amplifier, the 10W ‘X10’ was more due to the piss poor frequency response and switching characteristics of the Germanium transistors to be had at the time.

Although on paper the class D amplifier looks nothing more than a high frequency switch modulated by audio with a Pulse Width Modulated output there were some technological hiccups.  Although tubes are linear devices and operate well in class A, AB, C configurations they are also very inefficient as switches. There are devices which were designed for this purpose such as the thyratron but they only offer a modest switching speed.

It was only in the 90s that the work from some companies forking out on a Class D solution started paying off.  One such company was ICEpower,  a combination of resources from team Karsten Nielsen and the Bang & Olufsen audio group. ICEpower remains one of the top end suppliers of class D audio. Tact or later,Lyngdorf was another company specialising in high end class D and this as a result of their variable supply and digitally controlled PWM amp, the TacT Millennium.

The TA2020

Specification sheet for the Tripath TA2020

Tripath, named after Adya Tripathi the CEO,  was the first company to mass manufacture a class-D chip, the TA1101. It was only when the TA2020 was released did the audio fraternity sit up and take notice – 20W into 4 Ohms at a fraction of the price of equivalent chipamps of class-AB design. After the demise of Tripath the market became flooded with the 2020. The TA2020 is rated as one of the top 20 IC breakthroughs ever, in line with 555 timers, ARM1 and Intel’s 8088.  A fabulous testament to the short lived Tripath name.

Tripath Schematic 20W+20W into 4 Ohms
Tripath Schematic 20W+20W into 4 Ohms

There are lots of add-on circuits for these amplifiers on the web but mostly one would be looking at pre-amps etc. The circuit itself may not lend itself to much better output quality due to the self-contained 2020 (ever tried to open a chip) but there is a definite improvement by using a higher current PSU and according to DIY audio (the site), recapping and resoldering should make a difference. These amplifiers were sold in South Africa for about R250.00 and to date the cheapest is about R350.00 on Bid or Buy. For that price don’t expect a Premier 5, do expect the best value for money, especially if driving from a PC.

One day the original chip will become a collector’s item – so box and stash away.

Oh yes, I believe the Texas Instruments chip has the best figures. More about this later.

Class D is here to stay.

The Golden Ear vs X-Ray Vision

Fact versus Fiction – our hearing is pretty hairy

There is a gem of a website call Audio Check which allows you to test your sound system, DACs, headphones as well as hearing. I recommend all those interested in audio to pop over there, light a cigar, grab a cup of tea or your favourite brew and do some reading.

Another great read is “Testing Audiophile claims and myths“.

To sum it up:

Look, it has all but gotten pretty boring over the years.  When in my early 20s and in the 80s I had a sound system which I would rate as pretty powerful, about 100W per channel into a 4  Ohm load. When it came to parties and the venue owner’s system ran out of steam we’d plug in this baby and crank the volume. The quality of the sound was proportional to the amount of alcohol.  It probably deserved to have a rating of about 2/10 compared to modern lower powered systems.

What I have learnt through the ages is i) to take your time to build an amplifier ii) use the best components possible in the power supply, cooling, consumer controls, jacks and plugs iii) we now have a much wider access to and better quality patch cables and connectors iv) electronic consumer goods have become much more reliable generating some pretty exotic powers. Oh, this list goes on.

Good looks until it opens it’s mouth

A good looking amplifier doesn’t mean it was well built. It’s under the hood that counts. Also, knobs and good quality pots cost good money.

Don’t be hood-winked into buying an amplifier or any component for that matter without doing your homework. In South Africa, unless it’s very high end audio, the retail store is not going to do a delivery and then a collection again after you have done a test over the week-end.

Brands we know and trust

Most of the better known amplifiers (and pre-amps) which would include Yamaha, NAD, Harmon-Cardon, Denon, Onkyo, Rotel are well designed, are great for normal listening and more.  Get something which likes your speakers and the speakers like the amplifier. Reading up about ‘which’ loudspeaker is crucial and is always the best bet – it’s going to cost.  Be warned again, a set of loudspeakers may sound good with one amp but not the other. In South Africa Mission is often partnered with NAD.  I don’t know about this match but there’s thousands of people out there raving about the dimension, the colour, the warmth, the depth, the ooh, ooh, the orgasm of things.


Headphones are a touch more personal – each one to their own of course. I don’t believe most of the marketing hype.  A set of LG headphones which I received as a gift about 15 years back astonished me in the quality (excellent). I like Senheisser. That does not mean you have to.

Bitrates and Sampling

Too many articles pointing to the hazards of dubious marketing – our hearing is only 20Hz. to 20kHz. Mine is maybe 30 Hz to 14 kHz. 24 Bit / 192kHz may not sound better than 44.1kHz at 16 Bits. Maybe 8, but not 16.

Milking the snake of all it’s oils

What does all of this mean of course?  Have you seen the advertisements adorning Hi-Fi mags promoting audio and power cables?  R 1 000/m or more. In a magazine years back the fundis tested numerous cables and the one they thought sounded best was telephone cable. (yep ma’am, that cable running to your house, not to your telephone).  Now we get told how good CAT-5 cable can sound, both on the input and output. Kettle plugs?  Just make sure the L is Live. Some aren’t and don’t ask me where they come from.


Spectacles may improve one’s eyesight but a high powered amplifier is not going to improve your hearing. Most of what we want doesn’t always follow the rule of logic. Do read “Testing Audiophile claims and myths” and Audio Check!

Phase Splitters – driving Tubes and Transistors

Maplin Bridging Adapter for the 75W - 100W amplifier

Phase reversal techniques used in amplifiers

Although we still have the Bottles versus Transistors promotional teams out there one thing which must be said for the tube amplifier is it’s simplicity in design.  Most tube amplifiers of today are a copycat of what the engineers knew a long time back, possibly more than 60 years ago in fact.  Very high output amplifiers, 1kW and above with possibly only 6 tubes.  The supply rails are  lethal and what didn’t kill you broke the bank balance, the cost of the output transformer of course.

Easy to read, not to understand

MOSFET amplifiers are also schematic friendly but like tubes, they have very high input impedances and gate voltage control can be a very delicate operation.

SET Amplifiers 

Push-pull tube design is not the latest fad, SET amplifiers (single ended triodes) are. Strangely enough purists believe that the lower the power the more realistic the sound.  And of course we must not forget that the so called SET amplifier was the first ever design for an audio amplifier.  What’s the tagline of this website any way?

Bridging an amplifier

In a previous article we wrote about how to bridge an amplifier to get more power. Well of course a push-pull amplifier follows the same train of thought – driving two output stages, one in phase and the other out of phase by 180 degrees.

Basic phase splitter circuit
Figure one – Basic phase splitter circuit with a) triode and b) NPN transistor

In the circuits above, the triode anode output is phase reversed and amplified, the NPN collector output is phase reversed and amplified. The general rule in such “concertina” style splitters is to have emitter and collector load resistors the same value and the emitter is not decoupled to allow for 100% feedback and unity gain (1).

Transformer drive

The transformer coupled splitter shown below used to be very common in the older Germanium radio receivers. Predominantly PNP and working off 9V. Current lack of bias to the output pair will result in crossover distortion.

Transformer coupled splitter
Figure two – Transformer coupled splitter

Not having any bias to get the transistors to partially conduct is not of course what one wants but this is for representation only.  There would be severe crossover distortion.

Long Tailed Pair

Figure 3 - Long tailed pair - audio amplifier phase splitting
Figure 3 – Long tailed pair – emitter coupled  audio amplifier phase splitting

The more common emitter coupled long tailed pair is shown in figure three.  This is one of the most popular means to generate split phase signals. The two transistors should be well matched.

Maplin Bridging

Maplin Bridging Adapter for the 75W - 100W amplifier
Figure 4 – Maplin Bridging Adapter for the 75W – 100W amplifier

Figure four and five is that of the famous bridging adapter used in the Maplin MOSFET amplifier design, supposedly good for about 350 to 400W.

Maplin MOSFET Bridging circuit
Maplin MOSFET Bridging circuit – click to enlarge

The Maplin bridging and speaker protector was a very popular circuit in the 80s.  In the DIY Audio forums note is often made of the circuit, the power restraints and the designed loading.  Due to the significant on resistance at saturation of MOSFET devices there will be significant power dissipation. For rated output when bridged it is highly advised to parallel extra transistors at the output stage.  The recommendation is an extra N and P channel for each channel.

Phase reversal

The LF347 J-FET opamp is where the signal phases are split to get 180 degree shift between pins 14 and 15.


The output signals from the two amplifiers are monitored by the 741 op amp.  As the signal are 180 degrees out of phase with each other the resulting output will always be 0V.  However if one amplifier had to fail causing the loudspeaker output to swing either positive or negative the diodes connected to the bases of the 557 or 548 will conduct.  The 4 x 2-input NOR gate acts as a three second timer in normal operation or if there is a fault will latch, triggering BFY51 which will conduct and pull in the relay.

The circuit

The component  values have been deliberately left out of the circuit. It would be cheaper to purchase this kit at R230.00 from Yebo Electronics in any event, taking the time of making the PCB, relay pricing and other components into account. Note the supply is not included in the price of the kit. Also check your schematic, mine was hard to read. The circuit is mostly advertised as a bridging adaptor/adapter but it of course includes speaker protection.

Again, that warning about bridging

When bridging an amplifier which is designed for 8 Ohms the output circuit must be capable of driving half that impedance i.e. 4 Ohms. Modern high voltage consumer amplifiers are designed for 8 Ohm loads but many can drive 4 Ohm load but not 2 Ohms. Most professional series amplifiers are designed to drive 2 Ohm loads (until you see the insides).  I prefer to play it safe and not bridge an amplifier unless it has specifically been designed for this purpose.

Although the Maplin MOSFET is designed to crank out 100W or so, it will not safely drive a four Ohm speaker in bridged mode and will fail.

Next:  Long tailed pairs in audio amplifiers – we take a look at MOSFET amplifiers



Power Amplifiers and Output Powers

Why the amplifier sounds more like 4.5W than 20W

A common circuit for most beginners to build as their first project will be something to use inside their or their friends car.  Amplifiers of this ilk come in various flavours but the most common will be the 6W or 4.5W RMS version.

Here’s why?

Your power supply is 12V to 16V (I am being optimistic here, for you and not your battery).

complimentary pair configuration - audio amplifier
Basic complimentary pair configuration – audio amplifier

A quick calculation to get DC power

The amplifier above is in it’s barest detail, a complimentary pair out with diode bias.  The maximum peak to peak voltage as can be seen by the adjacent diagram is going to be 12V, the battery supply of the vehicle.  Of course there are some parameters we need to take into account, one being that a transistor when conducting hard will have a small saturation voltage between the collector and emitter.  I have set this to a not too unreasonable 500mV.  So in fact the p-p voltage will be closer to 11V. The RMS voltage which one uses for measuring power if the circuitry was wholly DC is 0.707 * 5.5V (half the p-p) equals 3.8885V.

Now if you have a load resistance of 8 Ohms and we use Ohm’s Law to find current (current = volts / resistance) then we have 3.8885 /8 = 0.486A.  Power can be calculated by power = Volts * current or 3.8885 * 0.486 or 1.89W. Using the same means we can determine that a 4 Ohm load will dissipate 3.78W.

Of course if we leave out the saturation voltage (which is quite high) we get to a figure of 4.498W (4.5W) for 4 Ohms and 2.25W for 8 Ohms.  Not very significant isn’t it.

Of course if you have robust enough transistors you can load the amplifier with a 1 Ohm load to dissipate 15.12W which is a little more reasonable.

Increase the Supply rail voltage with a switching supply.

But the captains of the International Brain’s Trust came across another remedy, a much better plan.  Why not increase the supply rail voltage to say +35V.  This can be done of course and now we can get 38W per channel out of a car audio system.

I recall the old ETI (Electronics Today International) abbreviated the entire power calculations to

quick formula to get power output in RMS
Power RMS = Vcc squared / 8 RL

Test it out with the above circuit, say 4 Ohms and 5.5*2 or 11V supply.  132/32 = 3.78.  Remember you will need to take off the saturation voltage off both transistors or 0.5 * 2 = 1V.

What about bridge power?

Not my most favourite subject but for car audio works pretty well until the installer earths a speaker wire (even though there warnings in all 10256 languages).

Bridged complimentary output stage
Bridged complimentary output stage – the audio signal is fed into the two inputs, in phase and 180 degrees out of phase.

The 2 x the voltage equals to 4 times the power formula

The first bridged amplifiers I came across was the pre- switched mode amplifiers for automobile auto use in the early 80s. These were almost always advertised as 20W * 4, 25W * 4 and even 50W * 4.  The power just kept on going up but the supply rails and load impedance stayed the same. Nothing like a good solid piece of marketing to keep the ZARs moving.

Using our tried and trusted formula above still holds water of course but except now we no longer have a maximum voltage swing of 12V, it is now 24V.

If an amplifier is designed to dissipate 4.5W into a 4 Ohm load, in bridged mode it will be 4 times that.  In other words your supply rail has effectively doubled.  Using the Vcc squared formula we get approximately 20W into the 4 Ohm load.  If you have 2 x 150W amplifier, the total power (possibility) is 600W. In a case such as this though the impedance is never the lowest as advertised by manufacturer e.g. 4 Ohms, but now 8 Ohms.  This gives double the power output or thereabouts.

E.g a home system deliver 100W into a 8 Ohm load. (This is equal to about 28V across the load). This is also equal to about 200W into a 4 Ohm load. By bridging this amplifier the load sees a maximum amplitude now of 56V, a power dissipation of about 400W. Therefore by doubling the output rail voltage we quadruple the power dissipation, a very daunting task in design for engineers.

Some problems engineers sit with:

The ordinary consumer is not usually concerned about thermal related problems in amplifier design, neither matching impedances. Output mismatching for a tube amplifier can cause arcing inside the output transformer (never use one without a load), semiconductor amplifiers can usually run without a load but never a short.

Safe Operating Area

All transistor amplifiers are designed around a Safe Operating Area of all the transistors, the big issue usually being driver and output transistor circuits. Early semiconductors were not forgiving, especially the Germanium types. Transistorised amplifiers, or rather the outputs can suffer from what is termed secondary breakdown.


When using your amplifier in bridged mode every possible law in electronics can be broken at the flick of a switch, all because of wrong load impedance.

Temperature and Current Limiting

Much has been done to add current and temperature limiting into design but there will come a time when parts suffer from thermal fatigue and will fail. Engineers try to use the best possible (or the most VI resilient) devices in their designs.

Very high powered amplifiers often have more critical protection circuits in the the final design than in the voltage gain and power output chain.  In the real world we need the least possible components.


Mosfets have a much higher saturation voltage than BJTs. They are also not known to suffer from thermal runaway as in conventional bipolar transistors although there are different trains of thought in this matter. Otherwise they will fail just as easily as any other component if not cared for. Design criteria for MOSFET amplifiers, although seen to be simple according to the schematics is often related to the very high input impedances and PCB layout.





Protecting your loudspeakers

Velleman Kits - K4700 dual channel speaker protection

Modifying the STR-DG500/600 and Protecting the speakers

The one issue I have picked up with the Sony amplifiers in this range is their ability not to switch on when there is high humidity. Many owners are not aware it’s a humidity problem and with all the functions these amps have one would have thought an LED indicating thus would have been relevant.

High humidity causing no sound

Actually there are quite a few amplifiers on the market which exhibit loudspeaker protection mode so it can be fairly common problem, I don’t like the idea of switched mode supplies and humidity though.   Use a hair dryer to get the engine running again.

The reason why one has protection to the speakers is usually due to one or both power stages going faulty and allowing DC through to the load.  Plus or minus 50V across a voice coil is going to cause a serious disruption to service – in some cases it’s after the damage to the speaker that the fuse blows.

Current limiting, fusing your speaker rails

Some ways to limit the probability of this happening is to put a current limiter in series with the speaker, a pos temp coefficient resistor or poly-switch. A light bulb in series with the speaker is also a solution. In both cases purists don’t like any resistance in the signal path, especially one which is not pure and can be non-linear.

passive speaker protection
Figure 1: Passive speaker protection

Active speaker protection

The audio amplifier when switched on usually holds the speaker protection relay open, tests for DC, stabilises and then connects output signal to speaker load.

sense and trigger circuit for speaker protection
Figure 2 & 3 – sense and trigger circuit for speaker protection

Assume the circuit Figure 3 is in active mode i.e. protected or has just been powered up.

How does it work:  Looking at Fig 2, R1, C1 and C2 makes a potential divider chain, the audio signal is passed through the rectifier D1~D4.  When the DC voltage starts climbing towards the switch on voltage of T1, the transistor conducts pulling down the base of T2 in figure 3, switching the Darlington T2 and T3 off, switching off the relay (or opening the circuit to the loudspeaker).

C3 and R3 have a time constant and acts as a switch on mute circuit.  When the circuit is powered up the capacitor keeps the base of T2 low, charging up through R2 until the devices T2 and T3 conduct, switching on the relay.

One problem which can rear it’s ugly head is that the relay contacts burn closed which may not then protect the speaker.  Figure three shows a circuit with a SPST relay which for most smaller amplifiers should be adequate but the proper way is to use a single pole dual throw relay and earth the NC contacts.  See the article on

Proper wiring for speaker protector – SPDT with NC to ground

Proper way to connect speaker relay
Figure 4: Proper way to connect speaker relay – NC closed contact to ground.
Velleman Kits - K4700 dual channel speaker protection
Velleman Kits – K4700 dual channel speaker protection (note the Halfgaar comment about earthing NC relay contacts. The Velleman does have a to ground resistive path though).


The Sony STR-DG500/600

Sony Speaker and Headphones protector
Sony Speaker and Headphones protector

Active Speaker Protector - Protecc Electronics

Active Speaker Protector – Protecc Electronics

In conclusion

  • I purchased the Velleman K4700 (pdf) from the US a few years back and shipping to SA took about three weeks, customs and excise clearance done prior in the USA. The kit is well made and simple to follow. Highly recommended.
  • Due to the higher powers that we encounter in audio amplifiers one needs to also be cautious as to whether the relay contacts are adequate.
  • A general rule of thumb is to allow for 50 to 100mS switch-over in the event of a fault. 1/10 of a second is pretty fast. 1/20 of a second is better but here we are nearing the extreme where erratic switching may take place at peaks.
  • In a new build it may be best to test with a coupling capacitor. Check rating and polarity.  Coupling capacitors to loudspeakers pass oodles of current, check current capacity as well.  (ripple).





Sanken MN2488 and MP1620

Sony - section of original schematic STR-DG500/600

The Power of Darlington – Sony amplifiers and receivers

Another DIY project

I have had numerous Sony amplifiers which have been stripped down for spares, mostly power supply, power transistor driver chips, the NEC UPC2581 and the then popular Sanken MN2488 and MP1620 power Darlingtons. Now Sanken has been around for some time and when one refers to audio technology we think of TI, On Semi, National Semiconductors. But Sanken has made some great chips as well, the STK series like the 435. We had the 10W, SI-1010G, 20W SI-1020G, 30W SI-1030G and the whopping 50 Watt the SI-1050G. How about their transistors?

Vintage Sony

Older Sony equipment, especially the Trinitron and audio amplifiers can make one nostalgic, never known for great audio possibly, but bringing out the Sony TA-N77ES in the mid 80s made the audio world sit up and notice.  This was a juggernaut, 200W per channel and absolutely gorgeous.  Possibly one of the best looking amplifiers ever (next to McIntosh of course).

Sony Modification - 200W audio amplifier
Figure 1: Sony Modification – 150W / 8 Ohms audio amplifier – TinyCad

Puff the Magic Dragon – Bang goes Aiwa

The Sony takeover of Aiwa I thought was a f****up of note. Aiwa made some really great stuff and now they are non-existent. I even wonder whether Sony gleaned anything of value from  this once great company. At one stage there wasn’t a household in South Africa which didn’t have at least one Aiwa product, or so it was perceived. Oh yes, Aiwa also made better component systems. But I am not here to ‘diss Sony, we still have many die-hards out there and I really would like to see them make a comeback. What, you didn’t know that they had pulled out of South Africa?

The popular STR-DG600 and STR-DG500  (7.1 and 5.1)

The Sony STR-DG600 7.1 was brought out to some rather dismal and really vague reviews. Most of this I shoved to the back of my reptilian brain and did my own testing, with cheap Sony speakers. Note that this amplifier can be purchased from the USA for less than R 2 000.00 new. In SA, second hand about R 1 500.00 or less. This is not break the bank stuff and gauging by some of the reviews one would have thought they had purchased a Levinson series amp for  3000 quid.

It’s clean, very clean. It cannot power more than two speakers (the front) at full throttle without moderate to severe distortion. At 2/3 volume with SACD or CD input it’s remarkable, for the price.  What makes it tick? Here we stick to the analogue parts of importance – the output stage.

Darlingtons, high gain circuits and 8 Ohm loads

The amplifier uses a very non-unique amplification stage to their entry level range, utilising a NEC chip-amp (IC) driver stage and a Sanken Darlington pair in the output.  Because of the +50V/-50V supply rails, the SOA of these transistors and configuration lends itself to 8 Ohm loading and driving two 8 Ohm speakers in parallel is inviting trouble – just don’t do it. By removing the other channels from the supply and adding an extra pair of darlingtons to the output stage of the L + R channels I do believe will allow for 4 Ohm loading. The transistors, with a current gain of over 5 000, paralleled, should not provide undue loading to the driver chip. Remember the 0.22 Ohm emitter resistors.

So why the focus on this configuration? With the hype on gainclones and the like here is an opportunity for those home tinkerers to build a 200W amplifier with a minimum of components. These amplifiers are very popular. Their failures almost always came down to possible “micro-controller” problems, blown output transistors, mains transformers and the ever popular  “not switching on” after a lengthy period switched off at the mains to allow for condensation.

Building your own stereo amp minus the DSP

To build your own home stereo amplifier with components from these amplifiers (the STR-DG500 is similar BTW), I would recommend one transformer per channel or better still, a toroidal rated at about 300VA, secondary windings 42V-0-42V. If you are going to stick to 100W per channel the transformer may do it.  Don’t hedge any bets though.

Sony - section of original schematic STR-DG500/600
Figure 2. Sony – section of original schematic STR-DG500/600

There are some things one needs to remember if you do get your hands on some parts to build the modified circuit:

  • I have not built this. It’s on my to do list 🙂  which includes many, many things.
  • The article is based on the presumption that you can get your hands on an old Sony 500/600.  You should be able to pick one up for about R100.00 if it is faulty. Because of the pre-driver/driver NEC chip it greatly simplifies the build. The transistors are Darlington which simplifies it further.
  • Use the components in Fig 2. for Fig 1.  Paralleled transistors need the 33 Ohm resistors at the base as well as 0.22 Ohm emitter resistors.
  • This is a very high gain circuit and oscillation can and will occur if the board layout is not properly thought out.
  • Power output should theoretically be about 150W / 8 Ohms.
  • You can parallel two transistors on the existing board but remove power to the non-used amplifiers.
  • You will need a cooling fan on the existing heatsink.
  • You will need to add a speaker protection module.  (Velleman is just one such supplier).
  • Please let me know if the publication of the partial schematic is inappropriate.
  • The original schematic can be found on the internet incl. Hi-Fi Engine.

There are near equivalents for these now obsolete transistors, possibly 2SD and 2SB types. Because this is a from scratch build (possibly) make sure you have current protection across the  (both) fuse holders on test.

Article: Loudspeaker Protection