Mclaren Audio design Philosophy
To try and generalize the design philosophy is quite difficult but a very brief statement of approach to the problems can be useful. Our aim is to find and eliminate the 20% of things that cause 80% of the loss of enjoyment in listening and using the product. This relies on the setting of priorities. I will try and outline those areas which I consider of importance to me at the present time in the following pages.
- Sonic Performance
- Reliability, Serviceability, extendable operational life.
- Aesthetics, Ergonomic
1. Sonic Performance
To reproduce an exact replica of the input (with consideration being given to the ear and brain combination sensitivity to certain distortion types), into any sensible load excepting screwdrivers or No8 wire
THD has become probably the most popular and most easily understood type of amplifier distortion. It is due to this that it has become an extremely overrated measure of an amplifiers quality and performance. THD is I believe of little importance. THD is in fact of exactly the same from as the music we are wishing to reproduce. This makes it totally indistinguishable from the music signal so while effecting the technical accuracy has no bearing on the qualitative musical experience unless it is at extraordinarily high levels.
Other distortions deserve our attention before turning back to the THD from the amplifier. Finally speakers often have THD figures up to two orders of magnitude higher that may mask that of the amplifier and perhaps not other forms of distortion.
Once more we have an area of amplifier design that has had a lot of time given to it, with highly commendable results. However we now have amplifiers that are extremely unlikely to run into speed problems. Able to switch at 75 to 300v/uS i.e. with -+50v rail 750khz to 3meg. Admittedly they are distorting severely but you must agree very fast for audio which will only require very small outputs above 10Khz. 20kHz requires 20v/us at +-50V, within reach of a moderately fast amplifier.
This is a form of distortion that is probably of much greater importance because it has less direct relationship to the harmonic structure of music. The relationships are also not linear giving large magnitudes of distortion from small imperfections in the design. A lot more can be gained focusing on what is occurring in an amplifier circuit to minimise the IMD products.
Energy and power supply requirements
I feel that we do not need as much energy stored as is often thought.
The energy stored in the power supply need not be all that great as the RMS output of the amplifier is indeed a true indication of its power output capability. Measuring speaker characteristics while reproducing music shows that they DO NOT in general draw enormous currents or require HUGE amounts of energy.
There is no reason to believe that you need large amounts of energy because it is NOT POSSIBLE to extract it from the power supply at any greater rate than indicated by the RMS ratings of the amplifier. So why store three times the amount of energy required when you can’t extract it? The simplest explanation is it makes the amplifier circuits job easier by being more constant circuit operating parameters don’t “move” with the supply created by doing this. You will not find many designers of audio products that do not have strong power management at the core of the design philosophy as a result.
The success in this area alone can make or break a design. As a result this is in an unconventional way a corner stone of my philosphy.
Other energy considerations
Current required from an amplifier using an 82db/w @ 1meter speaker
|db||watts||4ohm current||2ohm current||1ohm current|
|109 extremely loud||512||11.3Arms||16Arms||22.6Arms|
So the worst likely load is 1 ohm requiring 23 amps. This is within the capabilities of amplifiers with modest power supplies and output stages and the amplifier must have a voltage supply able to deliver 500w into 1ohm. This is s a large amplifier.
NOTE 1 ohm speaker impedances are usually in the mid or high frequency ranges. Where both the amount of energy and the power handling of the speaker are small. Try to imagine 109dB in the mid or high frequency range.
Another area of energy consideration is that of Class A. This yield an output efficiency of around 25% maximum for this reason Class AB operation should be used in all cases. The benefits of Class A are debatable at best.
The conclusion drawn from this is that you only require a power supply of a size needed to obtain the desired output levels and no more. This implies 10-15,000uF on each rail. Also implies care should be taken to ensure crossover distortion is at an acceptable levels.
In a real amplifier the ability of a standard capacitor to regulate reduces significantly at low and high frequencies and the amount of energy required increases with decreasing frequency. As if that wasn’t enough the noise on the supply is not linear. Finally we have the ripple pulses from the rectifier. Some of this noise in many cases finds its way through to parts of a circuit. This can cause major problems in the stability of operating conditions and place greater demands on feeback systems. These have lead me to be believe that amplifiers that are worth listening to will have a regulated power supply on driver stages if not all stages.
This is another form of distortion often ignored in the past amplifier designers and reviewers (the first may result from the second?). My feeling is that it has considerable importance. As with power supply the distribution of energy in music must be taken into consideration. To optimise this area of performance layout and earthing become critical. This relies heavily on experience of the designer.
It is also notable that a relationship exists between power supply and cross-talk as the power supply is often the only electrical link between the two channels. The other major aspect of cross-talk being design layout.
Once again I am going to emphasise the need to look at the energy distribution in music and use that as one of two arguments for extending low-frequency bandwidth. Since most of the energy is at low frequencies any loss of low frequency information may contribute significantly to an overall change of character.
The other aspect of low-frequency bandwidth relate to something often talked about in relation to the high frequency end of the spectrum. Phase shift! If we put a 20Hz high pass filter into an amplifier we immediately must have a 45 degree phase shift (single pole, multi pole filters are worse). At 20Hz this represent a time shift of 6.3msec or a distance of 2.06meters. Surely a little more significant than 45 dregrees at 20khz representing 6.3usec or 2.06mm
This argument does not however exclude the requirement for an accurate 20-20khz bandwith with as small as possible variation from an absolutely flat response.
Capacitors in the signal path
This is a known and debated area of audio design which can give quite dramatic changes in sonic quality when capacitors quality is improved, or a capacitor is deleted.
Probably one of the major contributors can be demonstrated via an electrolytic capacitors behaviour in a feedback loop where at low frequencies which are also the high amplitude signals the capacitor will be reverse polarised and act as if it were a diode. This will cause all sorts of Chaos for the feedback loop and scatter distortion products over the lower bandwidth (Electrolytics also change capacitance depending on imposed voltage again spelling disaster at low frequencies). This phenomena will be much less pronounced with film caps but I would still expect some non-linear behaviour.
What it is precisely, that makes a good capacitor is debatable as with speaker cable or interconnect cable but it has still been accepted capacitors make a difference and a reverse polorised electrolytic of tantulum is a major problem. This is a topic that I have looked at in detail and have found that the best capacitor of all is no capacitor at all. This is considerably more important than many other forms of audio problem as the distortion generated is once again not related to the output signal harmonically
Once again many theories and solutions abound. For the most part it is a matter of looking at the justification for its use and trying the wire in question, and seeing if it works. I believe the majority of change experienced is related to the overall frequency balance of the system.
Silver and Aluminium conductors produce very good results when matched to McLaren Amplifers.
2. Reliability and Serviceability
This is of Absolute importance. A piece of equipment returned for service does considerable damage to reputation and irritation for the unfortunate customer.
Device ratings should not be exceeded under any operating conditions. High and low mains, High and low temperatures, High and low outputs, reactive loads, short circuits.
*Glass pcbs giving high strength for the longest possible service life.
*High quality plastic conductive pots high quality switches. These are the only moving parts.
* Integrated components wherever possible. This allows you to produce a product that is much cheaper per performance.
Full check listing
Each and every unit is put through a full check list on a semi-automated check list to ensure that it meets all the required specifications.
High standard of customer service
If a unit is returned for service it should be back in the customer’s hands as soon as possible. Most repairs should be turned around within a few days of arrival.
Easy access to all components Top and bottom covers should be easily removed and allow easy access to all serviceable components. This allows quick cost effective servicing if required.
PCB layout easily followed The PCB should be laid out in such a way as to make it easy to tell which part of the PCB is causing any fault described by a customer. The PCB also play a large part in the internal cosmetics of an amplifier.
Not totally no compromise, we aim for best within a budget. The equipment should provide the best possible sonic, ergonomic, and aesthetic qualities available within its price bracket.
This may mean some sacrifices may be required in terms of power output, THD, power supply size , speed. (see notes on sonic performance )
Try and cater to widest range of customers shopping in price range. This can be achieved I believe by producing a tonally “neutral” amplifier one that adds or subtracts nothing. This allows the listener to hear what the musicians intended in detail.
Internal layouts should be tidy and efficient this will allow faster more accurate construction and servicing.
4. Aesthetics, Ergonomic
The equipment should not intrude visibly allowing musical images and impressions to be formed without providing visual distraction. Having said this, all aspects of the amplifiers should be tidy and well finished.
Internal layouts should be functional and ascetically pleasing were possible.
Assemblers, designers can clearly be seen to have taken pride in the work they are doing.
This means controls should be sensibly placed and easy to use. Controls and inputs should be well marked. Back panel marking should be the correct way up as well as upside down to allow easy installation when leaning over the amplifier on a shelf. Control knob position should be well marked making the status of the amplifier easily visible.
The volume control should be such that a high degree of control is available at low levels.
Input Selectors: These should allow maximum flexibility in use.
Tone Controls: This is an option that is little used in modern systems as the general standards of speakers and recordings is very high. Tone controls should be usable as a fine tonal adjustment. Generally not giving more than + & – eight db of variation, I have chosen to no longer provide tone controls for our products.
Digital interfaces. I have chosen to NOT include any “smart” devices inside the amplifier box. This is due to the volatility of these functions. I would be doubtful whether any of the current crop of digital protocols will be in vogue in 10years. Since our Equipment has a design life in excess of 30years. (I have recently serviced and upgraded 30+yr old Mclaren amplifers and sent them back with the expectation they will run for another 20years)
Heat sinks should not reach temperatures greater than 55 degrees C even in high ambient conditions. This ensures that burns are not inflicted by the unit. Particularly important for those with children in the house.
All mains wiring should be secure and completely inaccessible where ever possible. This ensures the safety of the user if fuse replacement is required or internal adjustments made.
The amplifiers should not be able to produce damaging transients under any conditions. This is in regard to the protection of speakers switch on and off muting is now mandatory. DC protection and low-frequency transient protection is desirable.
Sharp edges should be avoided. This will reduce the chances of physical injury to anyone handling the amplifier or to those precious pieces of furniture on which it may be placed.
This material is for information only @2018