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.
Back when I started building audio amps, the versions of Spice that didn't cost $1000 limited you to 10 transistors and 500 Newton-Raphson iterations. And had to be run on an 8088 with 640k of RAM. Good luck simulating a real amp with that. And models, YEAH RIGHT. Even now when it comes to models you get what you pay for and an incorrect simulation is worse than none at all. I still don't bother using computer simulations for audio amps. Just data sheets and a calculator.

And you don't need an oscilloscope to design an amplifier - just to test it. They're helpful if it needs debugging and to check for oscillations.
 
lineup said:

Those who still believe that good amplifiers effect SOUND more than speakers,
may be could take a little break and figure their basic school math out
.. at least a little tiny bit, please ;)


Hi,

Oh I see.
It is another one of those threads where the poster is asking a silly
question just so he can tell us what the real question and answer is.

I do find this form of discourse somewhat tedious. Not quite as
much as meaningless descriptions of amplifier sound quality ....

;)/sreten.
 
Your conclusions Line up, exists because you are near the end of the big journey

of learning amplifiers.... you are into the step 5... the final level in my point of view.

Some others, many folks, are into level 3 and 4..... and those are near to conclude things.

Others are into first or second level...the bureaucratic level.

Continue the good work Lineup, to search for the ones have realised those things you told..... folks thinks you really want to avoid datasheet, and this is not your idea for sure...you are not slave from data sheet details.... your vision is macro...not micro... a general view....using some informations from data sheet of course.

I will give a break for a while... will wait you select folks...near future i will be here with you.... it is not the correct time... now ... young folks will be scandalized and will trow you stones... also in me... they will make agressive questions... have to tolerate those things (You).... as i have not opened the thread, i will have a good breath of fresh air around.

God bless you.... good feelings... good intentions... trying to bring them your deep "insigths".... conclusions of an experienced man.

Hello dear Dupont.... of course he use datasheet.... i think you know that.

I will unsubscribe for a while Lineup... will wait you to select... will return next month to see if you could "fish" some MORE folks to share those interesting ideas.

regards,

Carlos
 

Attachments

  • flash 073.jpg
    flash 073.jpg
    68.7 KB · Views: 977
Yes... i have my personal opinion... of course i can be wrong

Yes... i have my own subjective, personal idea, about
Folks evolution related amplifiers.

1) To study and to learn the basic electronics
2) To start to design and evaluate using instruments
3) To start to realise instruments are not guarantee
4) To start to study human hearing
5) To use instruments as reference, adjust using ears

I am glad i have passed all those steps.

I think the last step (6) is to give up to build/design amplifiers and start to go to listen alive music.... say... going to the real thing and not more focused into "reproduction" of real alive music.

regards,

Carlos
 
I will have a good cheese, french bread, hot and toasted and red wine

To celebrate dear Lineup, the ammount of folks that thinks this way is increasing...i am feeling happy that many others are entering the club....i was the number 11 or 12... and this point of view i have learned with a very good man.

I have not big know how... i am just learning things about... experienced because a lot of amplifiers made...but the best point was "insigth"... conclusions that made me step another level... to discover things even not beeing good in any of of those steps..but i had tracks...clue to understand things and to jump to another level of research.

You know...i was into simulators...i was reducing distortions and reducing and reducing... and assembling and testing... and nothing special happened... them i perceive this was not the solution... was not because distortion... this is "insigth"...then you go to research other things.... for instance..what distortions we like... and finally the conclusion that distortion is not the problem... the human audibility is the problem... the speaker is the main problem!.... this is jump step up!

Design good sound amplifiers is much more than take decisions, do calculations, do simulations and build amplifiers....there are much more than measurements and numbers into this work.

Hold your helmet and be prepared... a big battle will start.

I never had courage to start such kind of conversation...you are really a man!... i am a coward and will be eating my cheese.

Cheers!

regards,

Carlos
 

Attachments

  • camembert.jpg
    camembert.jpg
    78.8 KB · Views: 960
Re: Re: Yes... speaker are so bad that you really need a horn to help them

snoopy said:


With all of your experience it shouldn't be hard for you to design and build an amplifier which is subjectively and objectively superior to anything else on the market ;) What is stopping you from doing this ??


It is also very sad that these premium human tools are premium while you have the least experience, in your early twenties and deteriorate in time and after 48 or 60 years they become what I refer to as obsolete. Even if you have every right to think you can hear and advise others, this is not the case any longer.

Nico
 
As I saw earlier a response to my ".00000000001%"
amplifier this was truely hypothetical and only meant to describe
the outrageous claims of commercial sellers and to
highlight that they are not as important as the "interface',
speakers and human hearing.

In listening tests there IS an audible difference in amps,even a
bass amp. I just replaced a generic plate amp (MCM)100w
with "brother of quasi" and at the same 1v average input
there is a BIG difference in the sound.

If this was not the case DIY would be a useless endeavor
not worth the added $$$ and effort.Overdesign based on
sound electrical principals will produce amps you can hand
down to your kids not end up in landfills after 2 years.
 
lineup said:
.
01. take one idea for power amplifier AMP
02. build one power supply for AMP
03. build one prototype of AMP
04. attach dummy load
05. power on AMP
06. if no smoke, power off AMP
07. :att'n: attach your favourite speakers :att'n:

08. power on AMP
09. play some music
10. :cool: listen carefully :cool:
11. power off AMP
12. make eventual changes in AMP according to listening result
13. repeat from 08. until satisfied
.
--------------------------------------------------------------------------
Lineup audio regards
Lineup Audio Projects forum, members only
hello.
today i did use this method described here
with an one older amplifier class a
i tried different values of input cap (polypropylen)
uF = 4.7 2.2 1.0 and even 470 nF
also I did increase compensation cap (ceramic) in steps
pF = 22 47 100 220 470

The best sounding combination I could get, was with
1.0 uF input cap + 220 pF comp cap

I guess this will make the lower & upper bandwidth rolloff frequencies
fit my ears & used music the most

Note:
Only this one CD Album was used:
Dolly Parton - Little Sparrow (2001)
> dedicated to her daddy, Lee Parton

Especially used this classical song the most
8. I GET A KICK OUT OF YOU
Suberb music! Wonderful singer!

Lineup
 
Re: Re: DIY Power Amplifier designing by Listening Only

lineup said:
today i did use this method described here
with an one older amplifier class a
i tried different values of input cap (polypropylene)
uF = 4.7 2.2 1.0 and even 470 nF
also I did increase compensation cap (ceramic) in steps
pF = 22 47 100 220 470

The best sounding combination I could get, was with
1.0 uF input cap + 220 pF comp cap

I guess this will make the lower & upper bandwidth rolloff frequencies
this is not amplifier designing.
Refer back to your thread title.
 
It is very hard to obtain some "consistence" when you create amplifiers

I have made some...but each one of them, despite flat into frequency response (measurements) sounds better into some part of the audio spectrum... one is great for bass...other for treble...other for mids... another one is great for bass and trebles and have strange mids.... i could not...had not competence to produce something can be great into all audio bandwidth.

Also many other things.... presence, crispness, realism, deepness,spatial dimension, dinamics....and other qualities...not only frequency response.

When you tune the amplifier you bring to life the circuit tophology better qualities... and to be excelent into all audio band is not a quality i could obtain.

I perceive that i can have something unbeatable using 3 amplifiers i have made... one for treble, other for mids and other for bass or sub bass.... also that i can obtain this result using other amplifiers but mine.

It is interesting... even beeing flat into LEVEL.... the audio frequency response over loads, speakers or complex loads, each amplifier show his own "character"..... also speaker behaves this way...so... you have a never ending possibilities, combinations, that made our audiophile life so interesting... every day something new to learn, to try, every day a new result.

The best possible amplifier may be our human listening system... and listening alive music... i had not the information that someone could make something perfect.

Also it is not possible... music, the way we listen is a "reproduction"... and reproduction is not a reality... copy is a copy... acoustic is acoustic... electrical is electrical that will suffer multiple conversions.... acoustic and electrical are different.

We can have a lot of fun tweaking, adjusting and combining circuits, subcircuits and use a lot of tricks....i could not make something perfect...just very good units...but anyone of them is perfect.

Carlos
 
Speakers are the main problem

This is what i think and i am glad to share this with you...we can go tweaking amplifiers, they are already very good....many times better than speakers to reproduce music..... i will go fooling myself tweaking amplifier.... and the speaker will go producing a terrible "sabotage" into my efforts.

I think we are into the illusion that we can turn the speaker... the air pump...with ridiculous piston (diafragm) something better... it is already the same compared to 100 years before...difference is the permanent magnet and bigger capacity to hold power...even the material (paper) remains the same...we have tried a lot and we finished into the better one...the old and good paper cone..that sounds good when you scratch your nail over it...try another one to see how sounds bad... the material sound changes the audio tone as the diafragm is vibrating while pumping air....not dumped to scratch of nails...also will not be dumped to mid highs and high frequencies too.

We have to worry, during a storm, to create some ligthining array, a good ground and a ligthning rod.... not to pray for God help or to make a mess inside our house not to make our home atractive to receive a ligthning visit... we still behave alike indians i think...we go tweaking amplifiers when the problems remains into the speakers.

Another kind of speaker is needed for treble...maybe a high frequency spark of high voltage between two points.... the spark modulated with audible frequencies...amplitude modulation for instance... this will be great in sound.....ahahahah.. the problem will be the high voltage (dangerous)... the efficiency (very low)... the magnetic interferences (very high)...but it is omnidirectional and very good to high frequencies.... but so expensive that folks already have invented and give up to the more simple "air pump"

This is something interesting and very stupid too... also automobiles have high technology on board....but the machine...the power machine, the motor, continue to be the stupid alternating piston movement, burning fuel, poluting and producing noise...the same problem...we go researching automobile anti skidding, special break pad and electronic assistance, aero dinamics...and the stupid old motor remains there.

Sorry...maybe a little out from the topic...but i found the thread good.... Lineup has a great mind and i feel he is opened to new ideas too.

regards,

Carlos
 
Greetings to all who are sailing here in lineup's boat, especially the captain.

Let us imagine we have "the perfect amp". This amp is imaginary of course. It is called perfect because it is. It has no distortion of any kind, it has infinite input impedance and zero output impedance with unlimited output current. By this definition it has no sound of its own, what you hear is totally due to the source and the loudspeakers.

If we sent this amp around so everyone could have a listen and compare it to other amps, do you think that everyone would think that their system sounded better when using the perfect amp?
 
We can act on two principles only:

One: at random

Two: by applying the sum of our knowledge through experience and/ or intuition.

It seems to me that to act on principle one is bound to lead to vast amounts of time wasted and, at best, fairly long odds on actually acheiving anything at all.

Acting on principle two implies that the more knowledge and experience you have, the more likely you are to achieve your goal. If you allow yourself not to be distracted and diverted.

Given you can only aquire experience through your five senses only, and you can't see, feel or touch much of what goes on in a circuit, you must be better off augmenting your senses in anyway possible?
 
....


Hello my friends :)
And all others :xeye:

I have a couple of my amplifiers that I prefer more than other.
Ii is just that I like to listen more with these ones :cool:

When I think about it, my favourite power amplifiers
... is the ones designing by Listening !!!!!

Of course I did some basic measurments of voltages, to see that was everything setup correct.

But most of the fine tuning was done by method
change-listen-change-listen again .....


--------------------------------------------------------------------------
01. take one idea for power amplifier AMP
02. build one power supply for AMP
03. build one prototype of AMP
04. attach dummy load
05. power on AMP
06. if no smoke, power off AMP
07. :att'n: attach your favourite speakers :att'n:

08. power on AMP
09. play some music
10. :cool: listen carefully :cool:
11. power off AMP
12. make eventual changes in AMP according to listening result
13. repeat from 08. until satisfied

14. design complete
15. post images & schematic at www.diyaudio.com
--------------------------------------------------------------------------
This way you need only one multimeter to check out voltages
and eventually one thermometer to check out temperature of heatsinks.
And you need not much more than your ears and good portion of patience.
No simulation software.
No analog or digital oscillopscope.
No datasheets and curves of one million things.
 
i think that you have missed out some vital steps in your design process unless of course you are a very lucky man and or use witchcraft.

no way would i connect a speaker to an amp until i have checked its basic performance with test equipment, this is the only way to show up some faults instantly. (such as instability) other wise odds on that somthings gonna fry. especially with a new design.

when satisfied connect speaker and listen.

checking the amp in this way each time after any modifications reduces the likelyhood of frying parts.

also when you test the thing a few times the process gets easier and quicker, until the connecting of the scope ETC becomes second nature, you get to know what to look for after any modification.

of course when you transend to the higher levels of amplifier consiesness dispensing with test equipment is good for the second hand test equipment market
:D

but then again,i could be incompetant and know nothing about amplifiers.
 
step 7,8,... etc... would go so much smoother with an o'scope...... ah, gee i connect the speakers and now the amp runs very hot and blows fuses...... my wife complains i'm interfering with the radio upstairs..... if i use an o'scope i can see the low level 1.58Mhz oscillation, but never will i hear it..... yes there are uses for decent test equipment. the "totally sound based design" might be an interesting excercise, but i wouldn't want to do this with every amp i design, or repair..... like a President once said "trust but verify" and proper test equipment verifies quickly and accurately.

as for using ultra simple equipment to analyze the output of an amplifier, one could to to the extreme and use a wave analyzer, which is simply a voltmeter with a tuneable filter. the process of analyzing distortion and noise using such a device is a tedious process. there are still manufacturers who make these extremely simple but expensive devices. i would rather take a glance at an oscope and a spectrum analyzer for a few seconds to get the information i need than to spend a much longer time tuning a filter and recording voltages and then doing number crunching for a while, just to find out an amplifier is clipping. i'm sure that a wave analyzer is a very accurate (and prices for these beasts reflect that) method of measuring distortion and calculating waveshapes, but an oscop tells me at a glance what shape a waveform is, and a spectrum analyzer tells me at a glance what the distortion is and what the noise floor is.
 
Hello

Few month ago, will I was working on my amp schematic, I was mostly thinking about having a better scope, a better sound card for FFT spectrum analysis of thd, a better distortion analyser than my old HP-333A, etc... to do a ton of tests and measurements. For me those equipment was a must to have. And I was also working a lot with my Tina simulator.

But...

After dumping my slightly unstable cfp output for my amp and going back to a Self type II output, and trying many variations of the input and the vas, I decided to ask Hugh Dean for some advices, having myself a little bit of knowledge about phase lead, charge suckup, imput CCS, etc.. I become to agree more and more with Hugh, will chating often with him, that measurements was not the only solution.

Few week ago, I finished my first prototype of my last version of the amp, and after fixing somes problems caused by bad parts, and chating again with Hugh, I decide to do a listening test of my amp.

Wow !! It was magic, with my best cd, and with some old 50's recording, the sound was magic with a kind of "presence" and a living sound stage with more details in the sound than my Radford ever gived. I did use a Denon DCM-420 cd player (not the best of Dennon), and Dynaudio speakers.

After doing tests after test with my scope, my new M-Audio sound card as a FFT spectrum analyser of the distortion spectrum, and others equipments, and doing same tests with my Radford and other amps to compare the result with my new diy amp. All those tests and measurements did not show any significants thd, thd spectrum differences between my new amp and the others amps under tests.

Ear are THE final instrument and experiences (in my case it is more Hugh Dean experiences), and intuitions, are require for musical sounding amp design, instruments are there for basic verification of thd, oscillations, etc..., I did ear the result and now I will not bother to have more tests equipments.

Bye

Gaetan
 
burbeck said:

no way would i connect a speaker to an amp until i have checked its basic performance with test equipment, this is the only way to show up some faults instantly. (such as instability) other wise odds on that somthings gonna fry. especially with a new design.

when satisfied connect speaker and listen.

checking the amp in this way each time after any modifications reduces the likelyhood of frying parts.
unclejed613 said:
if i use an o'scope i can see the low level 1.58Mhz oscillation, but never will i hear it..... yes there are uses for decent test equipment. the "totally sound based design" might be an interesting excercise, but i wouldn't want to do this with every amp i design, or repair..... like a President once said "trust but verify" and proper test equipment verifies quickly and accurately.
gaetan8888 said:
... I become to agree more and more with Hugh, will chating often with him, that measurements was not the only solution.

Few week ago, I finished my first prototype of my last version of the amp, and after fixing somes problems caused by bad parts, and chating again with Hugh, I decide to do a listening test of my amp.
....
Ear are THE final instrument and experiences (in my case it is more Hugh Dean experiences), and intuitions, are require for musical sounding amp design,
instruments are there for basic verification of thd, oscillations, etc...

I have to agree with these opinions, which are not too different to eachother.
Measure and verification is most of the time a first necessary step.
That everything is connected/attached properly and 'working'.

And for making new own (experimental) designs an oscilloscope + signal generator is almost a must.
As well as a few old and cheap 'test-speakers' and/or dummy-loads. That can burn without any tears.

I can see only one possible exception. This is when you buy a seriously proven KIT with instructions.
In this case you can make it with your multimeter.

When everything works and you have got a reasonable sound
then you can try with smaller changes of values.
Such capacitors that effect frequency behavior.
Maybe increase or decrease some current level, by changing resistor value.
Try if one different transistor makes better than another.

This should always be done in such a way that you know what you are doing.
Changing values by chance and hope for luck is not the right way.
Your aim should have some working logic.


If you are the least in doubt, you should try to get some advice from some experienced.
Or consult our forum panel.
Better to ask one question too much, than get your work destroyed.

Regards ;)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.