diyAudio (
-   Solid State (
-   -   The King is dead, God save the King! (

destroyer X 27th September 2004 09:25 PM

The King is dead, God save the King!
My sincere congratulations to Mr Hugh Dean, Mr. Graham Maynard that could make better audio than My prefered John Lindsey Hood, model 1969.

My friends meeting evaluate again the AKSA 55 that was the winner few monthes ago, loosing only to JLH 1969.

Today i call the friends to a judgment, all conditions perfectly controled, no wind, temperature 24 degrees centigrades, no noise , no air conditionair, no one talking, sea level and many speakers used and 19 people judging, 16 audiophiles that belongs to my club and i invited my Maid my Wife and the younger daugther, those last ones do not have special interest in audio, only are normal humans.

1) AKSA 55, special for mids and brigths hi frequencies
2) Graham Maynard amplifier, the class A, because of bass and sonic image
3) JLH by the reasons you all know very well.

Despite i do not agree entirelly, my prefered still the JLH, this is the result of our meeting.

Since i told some lack of bass, Mr Hugh was in close contact with me suggesting modifications, now the bass is OK!

Related the Maynard amplifier, was not used the special transistor he insist that make a big difference...i could not find the unit and used another hi frequency one.

Related to JLH, the transistors are the normal 2N3055 and BD139 units, it may be better with some hi speed units.

Now i want to find one to beat AKSA!

hehe, the never ending novel



destroyer X 27th September 2004 11:44 PM

More details of old AKSA 55 test you can find in
"listening AKSA 55, had problems listening"



Graham Maynard 28th September 2004 06:48 AM

Hi Carlos,

Very interesting and most welcome findings.

The original JLH lets the loudspeaker behave more naturally, this is okay with loudspeakers that do not 'need' to be controlled, or those which have more reactive crossovers.

Cheers ............ Graham.

Graham Maynard 28th September 2004 06:51 AM

Sorry, got my words muddled there, because the original JLH can be less satisfactory when the crossover is significantly reactive.

Cheers ......... Graham

destroyer X 28th September 2004 07:13 AM

that is a fact, i could see that your unit reproduced well
I could see your unit had no problems with the many speaker models we use.... always reproducing good sound.

But the old JLH 1969 had problems with old speakers full of coils and capacitors inside.

In my home i do not use crossover... only series condenser.... 22 to midrange and from the i capture with 1 uf the highs to the twetter, mid range and tweeter are in parallell related the signal, but only the tweeter receive hi frequencies.

I use also one 22 ohms resistor in series with the mid and high drivers.

Speaker is ported and tuned to 35 Hertz.

Your amplifier is very tollerant to strange loads i could see, also Aksa is very good in that aspect, but Aksa not happy with 1.6 ohms loads... it was made to use 8 ohms i think.

Please Graham, take a look in "testing AKSA 55, had problems testing", there you will see the test condition...only ears, no instrument except AC voltimeter to match power.



AKSA 28th September 2004 07:29 AM


Many thanks for publishing your findings. I am pleased that the AKSA did well.

Graham, my congratulations to you - your extraordinary depth of research and math ability, together with sheer hard work, has delivered a brilliant amplifier!

Carlos, the AKSA was designed for 8R loads, that is true. I use it routinely on a 4R D'Appolito speaker - that is, the impedance drops near resonance to 3.6 ohms - without problem but I agree anything below this it does not like. A 1.6R load is verbotten, and recommended only with a four output pair AKSA a couple of my customers have built. This 'special amp' delivers more than 550W into 1.62R loads with complete reliability, using only 43V rails.

My thanks again,



Graham Maynard 28th September 2004 09:42 PM

Hi Hugh,

I often use the virtual Ariel 'loudspeaker' to test simulate amplifiers because it too has similarly awkward impedance dips.

550W available from an AKSA for the 'dips' - not quite a class-A capability !

Actually pure class-A can be found wanting through these impedance dips, but by then the amp would be running at its limits anyway.

Hi Carlos,

My amp is deliberately designed without any kind of input, C.dom or output filtering. Some individual listeners prefer gentle supersonic filtering, and this can be tried with my circuit, however I refuse to fit it as standard for this could limit the performance for those who prefer sharper detail. Feel free to try input filtering, also you could try a 0.22 ohm wirewound resistor in series with the output to reduce amplifier-loudspeaker damping. The reproduced sound will become subtly changed, though the amplifier would still measure similarly at audio frequencies.

One of my all-time favourite test sources is the 1986 CD by Luther Vandross. Give Me The Reason. EPC 450134 2. Apart from Luthers freshly mature voice, good backing and a decent instrumental stage without treble pre-emphasis there is magical recording of a glockenspiel. This will test any player-amplifier-loudspeaker combination, and especially show up amplifier-loudspeaker interaction and/or crossover-driver discontinuities on a multi-driver capacitor-inductor crossovered loudspeaker system.

Cheers ............... Graham.

destroyer X 28th September 2004 11:38 PM

Yes, some monthes ago i find AKSA too much brigth on my own speakers
As this brigthness is wonderfull, i was trying some changes to reduce a little the brigth without erase it completely, without destroy all hi harmonics, that cannot be heard but can be perceived as "natural" sound, giving to the bells the characteristic sound that diferentiate it from other metals as cymbals and others "hit on metals" sounds.... this way i used your top hi frequency "Maynard modified Zobel filter"...the 2R2 resistor used in Zobel filter, and also i used one inductor with some resistance...i use thin wire and a lot of turns, one experience, but could not perceive problems related the damping.... despite of the resistance introduced together the inductor, the bass came alive, and the reduce of "brigther appearance" was not so easy to perceive.

This way, some theories that using hi emitter resistor or series resistance between amplifier and speaker, do not create problems in bass reproduction as i expected, i could have a better bass, when i was waiting loss of bass.

I turn myself confused.... i am already confused.

Resistance in series with the speaker, will that reduce bass or not, as a effect of modified amplifier damping?

regards you and all guys


AKSA 28th September 2004 11:50 PM


Let's consider this output inductor thang.

You wind it with 0.5mm wire, 53 turns, to give around 0.1 ohms. The output impedance of an AKSA at 1KHz 20W is around 20 milliohms; this is a nominal damping factor of 400. You could argue that for many speakers this is overdamped, and the bass will sound very tight and rather dry, with very limited decay. With the 0.1R introduced outside the feedback loop, the Zout is now 120 milliohms, giving a damping factor of 67, six times less. This should have the effect of diminishing control, particularly decay, and the bass will wetten out somewhat, sounding fuller and arguably more musical.

Your modification is probably a good idea!



destroyer X 29th September 2004 01:10 AM

I feel ashamed sometimes in our forum, i have not this knowledge
I am one "changeneer"...i go changing till the sound turns agreable to my ears.

Sometimes i change so much, that i did not realise "what was the main change" that produce this or that result.

I need that capacitor that put things "on focus", the one you told me.... i have not good focus.

I changed so much your amplifier that i am really do not know if increase the quality or decrease..... i will make one more, 100 percent "standard" AKSA to compare to my "clone".

Your circuit board is wonderfull...hehe, i have one that never receive solder...not to destruct the "purity"... wonderfull board.



All times are GMT. The time now is 12:16 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2