DIY Power Amplifier designing by Listening Only

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Here is the best whole world audio amplifier

And this one was made as a kit till 2006..... Hugh Dean is the one produce, Aspen Amplifier from Melbourne - Australia.

No more kits..... he found better to assemble to customers, as reduced his time with follow up, people had not knowledge to adjust, and some could not assemble correctly... then he stopped to sell as kits...only some very special folks have received kits...but this was to personnal friends... and i am not even sure if he has sold kits or if those folks have received boards assembled.

Well.... objectivists have bothered him a lot.... if he have not felt bothered i felt angry to see what they have done with Hugh into this forum.... very bad educated folks, some of them having only College and pride (no experience, not skilled, just pretencious and awfully arrogant) attacked hard Hugh.... and one, the worst one, have made a brand with AKSA letters... that one may be devil's personnal friend.... well... this may have bothered him too... despite he (Hugh) said no problems.... he said... "a pitty they do not know the real thing.. the are wasting time searching for solutions into the wrong places".... well.... a gentleman.... even smashed by some idio-- have keept his education.

This one sounded better...not perfect!.... perfection is what i am searching as an exercise to keep my soul vibrating... to search for paradise, Nirvana, God or the universe sound... i know i will never find that.... but will be pleasant to see how they are different, say, amplifiers are different... this will keep me on.... motivated.... the speaker will kill all the sound...perfection is alive and nothing more can beat...and it is a non perfect perfection.... alive sound is awfull!.... reverberations, echoes, noises, cancelations, absortions...well..... perfect because real...but not really very good....so... it is just a search for something that matches your needs.

This amplifier, the Aksa 55 is the most near to perfection ever listened, and i did it many times, many models and i have them here.... unbeatable without sub circuits and special parts and some clever tricks applied to them..... mine looks alike and havethe advantage of no power on thump..but loose in reproduction... and the difference is clear... advantage clear to the 55.

I do not like the bass... i think it should be bigger ... if not in level...bigger into our perception (transference to speaker)... i have tweaked and no way.... reason why i continue my search.

The day i found a sound alike this Aksa...having a good bass to my taste, them i will be really happy... and i did it with DHR turbo...but treble was not so good.... and the bass advantage was four times more power..ahahahahah.... not fair.

Many folks have supported Aksa 55.... this is something you should think to listen.... to find someone near you to listen to it.

I am deeply have Line up is not a "simulator" anymore... a great guy, deeply intelligent and sensitive... his thougths will be very good to our forum, he will be a strong arm to open some blocked minds to accept there are things in this hobbie that only calculations cannot reach.

Listen to Leach folks...beautifull?.... and measure very well.... be carefull not to kick it angry... it is heavy and may damage your toe...use a double barreled!... or a hammer or ache....all those ones with double simetrical differential sounds the same.... aaagh!

There's a single capacitor into the Aksa that widens the separation, other sub circuit "eats" distortion while cleaping (reduce), other circuits send the sound stage to another dimension.... how simulator and calculations can understand that?

regards,

Carlos
 

Attachments

  • aksa 55.jpg
    aksa 55.jpg
    92.9 KB · Views: 378
The music before everything. I would put the constitutive man in front of it, who formed it, and who performance by it.

I welcome you, Lineup my friend!

I am glad that you exist.

How strange, that this the superb thread I discovered it now only. Very thought-provoking!
Whatever the live music lecture irreplaceable, from which I was lucky yesterday again to make sure. I had been on Siófok one where I was official in the evening yesterday onto my father's memory on an ordered concert. He quite a few choirs his founded and leaded in his life. The reputation of his oeuvre shows the ministerial honours received before his death.
Europe's country was not like that very much, where not his choirs would have acted with a big success.
In april 1999 The Siófok Child and Youth Choir founded by him, acted in the United States, at Walnut Creek.

http://www.ci.walnut-creek.ca.us/header.asp?genericId=1&catId=6&subCatId=1027

On the yesterday's evening his "little girl" Cecília Timár, my half-sister made a step conducting his choirs with a big success into our father's wakes.

The life is beautiful after all!

Gyuri
 
Hello, Carlos!

You are my man! Of course some people others, than for example the individual expressing himself before you! You wrote a lot of things that are noteworhty again, and with which I agree.

I am Gyuri, one for fatherless-motherless (boyless) orphans.
 
thanks carlos and gyuri
i never comfortable with people tell good things about me:
when people will put you high, you just wait, they will put you down again
but your friendly words :) are noticed and remembered in the darkest, most sad period of my life
we have a saying in my country of sweden: ~ what is in your heart, you mouth will tell
(can not hide what is in your heart, because will come out, truth will be seen)

i think we have to realize,
that all sensible, known and unknown designers, builders of amplifiers use their ears :cool:
this goes for Douglas Self, Hugh Dean, John Curl, Nelson Pass, M. Leach and Quasi, MikeB, PMA etc.
to call anyone of them only objective or only subjective, i think is making a big mistake
they probably listen more or at least as much to good music, as anyone does

electronics follows strict physical laws
where one electron enters a circuit - it will come out somewhere else
there is no subjective part in this
but the endresult, the goal of all this is to get a subjective impression
one that you, personally, can be pleased with

thanks, lineup
 
Lineup, I believe you create confusion in this thread because you talk of designing and tweaking as if it is the same.

I believe that the ears / brain are the most important measuring instrument but to rely on them only is a mistake also. When designing a new amplifier, good measuring equipment is essential, only after measuring you can start with listening tests.
 
I do not think I create much confusion :)
And I can agree with you that many times design and tweaking are two separate things.
But also design and listening evaluation can interact.
What is found during listening while trimming/tweaking can make you change design details.

Regards
 
lineup said:
I do not think I create much confusion :)
And I can agree with you that many times design and tweaking are two separate things.
But also design and listening evaluation can interact.
What is found during listening while trimming/tweaking can make you change design details.

Regards

On the fly tweaking and listening often results in a different experience, not necessarily better nor worse but different. So much so that you may decide to keep it that way. Measuring afterwards may show that your choice was incorrect, but the outcome of a system is to listen, not measure.

Image inviting friends to an audition with instruments hooked on your amp and saying, hey guys just look how this sounds. And the replay WHAA, it is good! What is that we are looking ? Can't you see it is the oscillogram of Diane Krall, it should be obvious, don't you also look at it at home. NAH, I don't have that CD, Lets rather look at Walter Beasly, I like the peaky parts.
 
jomatecc said:
The only use I know for an audio power amp is to listen to, I don't buy an amp to measure it, whats the point if it does anything else good, if it has no dc offset(the only thing I test for most of the time) and sounds fantastic Im good with any specs it has.
regards

Some years ago, (probably 20 maybe more) it was trendy to have spectrum analyzers and scopes displayes attached to the front panel of some high-end amplifiers.

Whether your speakers could actually make a sound at 50 Hz it does not matter you can see that it does from the running LEDs.
 
Some may argue that the speaker inductance changes as it is moved to maximum excursions, that might be true, however speaker manufacturers try to keep the coil in the magnetic field else it would burn out rapidly as the coil resistance will be the dominant part in dissipating the power.
 
Of course not possible to design only listening

To design you need, at least a brain... this is the first tool..them some basic understanding..ohms law, some frequency calculations of 3dB points... a pencil or pen to write, piece of paper, a simulator to help you and latter a scope may be interesting.

During testings you will need simulator or tone generator plus scope.. distortion analizer... will need tolls alike speaker, soldering iron...all those can be classified as instruments.... a table, a chair, a multimeter, probe points.

No one is able to do nothing without tools or instrument...of course fellows...have to engage brain and force some blood to circulate to help connections internally where rusted things and stucked things may exist..... Lineup, for sure had not intention to say that.... he means only tweaking to tune.... to develop, to increase, to upgrade, to update, to adjust... but this is post project..post design...cannot design tweaking ... will explode into your face...but this is so obvious that i never could believe someone could misunderstand that.

After design, when checked wave forms, measured distortions, them you gonna listen...during that you can tweak and may result a better sound...usually you can increase a lot..there are a lot of details experienced tweakers knows... and also the one that intend to do will learn too.

Cannot twist words this way...he talked in a generic way.... of course, sitted into the sofa, listening cannot do design.... or the guy love to redesign and recreate the same amplifier playing or will design another one..because while listening you are listening not creating the amplifier playing...the amplifier playing was already designed of course.

So... it is not good not to engage brain and go saying low intelligence things.... first think... second talk will be a good idea.

regards,

Carlos
 
I think I agree with Jan.

You need datasheets, CROs, signal generators, DVMs, to do the basic engineering. This to ensure that the amp clips symmetrically, delivers a flat frequency response, has no obvious distortions across the audio range, exhibits perfect stability in a wide variety of loads. A distortion meter is also useful, though not, in my view, mandatory.

Once all this is done, the tweaking for best possible sound begins. This involves component choice, operating point to a minor degree, and detail topology changes - with the engineering done over to ensure stability, flat FR, etc.

One of the biggest tasks of all is the layout of the pcb. It takes many days to come up with a good layout, and once achieved, it too needs to be thoroughly tested wrt stability. Most amps use global feedback, and such amps need careful attention to poles to avoid oscillation. Add to this the variability of the load - a diversity of speakers - and you have a very tough design problem.

Hugh
 
Im not sure if Lineup is talking about building from other schematic or designing a power amp from scratch, my views are coming from real world experience, designing to me is combining different topologies in each stage + components choices + amount of bias for each stage, ALMOST every amps schematic I've seen so far are variations/and or combination of these different topolies
singled ended or complementary, single ended or differential input
cascode or non cascode bjt or jugfet differential input



This is how I design an amp

1) I choose what topology to use for the input differential
simple, cascoded or CFB
type of transistors: fet or Bipolar
Bias current to use for input stage
Type of bias CCS, resistor, Zener + Resistor

2) Vas stage
Type of transistors bipolar or mosfet
Single , darlington, cascode, topology

type of CCS
Then the amount of bias current for the stage.

3) Output type
Either CFB or Emitter follower
amount of output transistors depending on the required
power output


4) Zobel network

I should have stated that I take layout and grounding very important and from past experiences learn to follow certain proven but not bulletproof procedures.

THINGS THAT I EXPERIMENT WITH ARE
Output ballast/emitter resistors values and type.
Type of input or feedback caps
value of input filter RC
Vas or Miller capacitor type and value(usaually polystyrene or silver mica)
Output Bias/queiscent current value
I will spend hours tweaking and changing till I get the sound I want.

TESTING I DO

Test output for DC offset value
test some operating points for correct voltage.
test output for oscillation < 300khz

I cannot afford and have never had a distortion meter or a CRO all I use is my multimeter and a transistor gain measuring circuit. plus a high frequency voltage meter to test for oscillation.


They usually come out Ok, Ive lost a lot of transistors experimenting but Im much better at it now, since Ive been doing this from in the 80s since the days of Graham Nalty and practical electronics, Edward Cherry, John lindsay Hood, Douglas self, I know my designs may not be perfect but they work and most of the time and I really enjoy listening to them on my system.
It may not the best way but my choice or limited.

thats my two cents.

PS please excuse the grammer
 
AKSA said:

You need datasheets, CROs, signal generators, DVMs, to do the basic engineering.
---
Once all this is done, the tweaking for best possible sound begins
---
One of the biggest tasks of all is the layout of the pcb. It takes many days to come up with a good layout, and once achieved, it too needs to be thoroughly tested wrt stability.
---
Add to this the variability of the load - a diversity of speakers - and you have a very tough design problem.

Hugh
jomatecc said:


This is how I design an amp


1) I choose what topology to use for the input differential
simple, cascoded or CFB
type of transistors: fet or Bipolar
Bias current to use for input stage
Type of bias CCS, resistor, Zener + Resistor

2) Vas stage
Type of transistors bipolar or mosfet
Single , darlington, cascode, topology

type of CCS
Then the amount of bias current for the stage.

3) Output type
Either CFB or Emitter follower
amount of output transistors depending on the required
power output


4) Zobel network

I should have stated that I take layout and grounding very important and from past experiences learn to follow certain proven but not bulletproof procedures.

THINGS THAT I EXPERIMENT WITH ARE
Output ballast/emitter resistors values and type.
Type of input or feedback caps
value of input filter RC
Vas or Miller capacitor type and value(usaually polystyrene or silver mica)
Output Bias/queiscent current value
I will spend hours tweaking and changing till I get the sound I want.

TESTING I DO

Test output for DC offset value
test some operating points for correct voltage.
test output for oscillation < 300khz

I cannot afford and have never had a distortion meter or a CRO all I use is my multimeter and a transistor gain measuring circuit. plus a high frequency voltage meter to test for oscillation.

They usually come out Ok, Ive lost a lot of transistors experimenting
but Im much better at it now,
since Ive been doing this from in the 80s
since the days of
Graham Nalty and practical electronics,
Edward Cherry,
John lindsay Hood,
Douglas Self


Thanks Hugh.
You points (I shorted your post a bit) are good and reasonably good practice.
I could very well be using your way at times.


jomatecc
A good post where you try do describe one example method of designing.
Like you have have leaned by readning forum and litterature + internet
how the more experienced do it.

Commonly there are very few, if any, that starts designing by ear from scratch.
It is simply not an effective method, leaving alot to just chance to be lucky to find something good.


1. Have a good idea
2. Write down a) what you want = goal, quality level how much money to spend etc.
3. Simulate or draw thercircuit and calculate some currents and possible devices.
4. Use simulator, if you wish
5. When you think circuit will work: Buy, order real stuff
6. Build one protype and setup same circuit in simulator.
7. Run the real, adjust run simlator with adjutments, try the real
and so on, until it gets good


... something like this.
Could this be a recommnded method.
Or some other way maybe :)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.