What's made a good amp.. ?

[gaetan8888]

Hi

I did read somes of the messages about relation between distortions and sound quality, it was very interesting.

Since some years, after reading J. Hiraga and somes other ones articles and doing tests using my old HP333A on power amps that I did have since years (my last one was a Crimson 60w amp), I'm not a power amp geek but here's what I think a nice sounding power amp should have;

Good open loop gain at high frequency

No need of a hot-rod slew rate, 10 to 20 v us are sufficient

Low rise of distortion at high frequency, so it have quite same distortion from 20 hz to 20K hz

Low rise of distortion at very low power output (.1 w)

Odd harmonics distortions are not good

Low phase distortion, about phase distortion; to much of phase distortion and the negative feedback will not work good.


Tell me where I'm right and where I'm wrong.


Thank

Gaetan

[unclejed613]

generally, your assumptions are correct.
however even with well designed amps, it is extremely difficult to escape the fact that there is transit time (and so some high frequency phase shift) in any power amp design. output devices have the largest input capacitances, and so introduce the most transit time.

[fotios]

Hi Gaetan
Right your opinions. The subject consequently is not if you have error, it is if you omit something. Already uncleJed pointed out something right. I receive very seriously under aspect the Transient Response of an amplifier because under today conditions it is very difficult you will meet a circuit with unacceptable THD+Noise or bad frequency response. The examination of an amplifier by injecting square waves of different frequencies in his input and the observation of his output with or without load by oscilloscope it provides optical information beforehand audition tests. Four very important measurements are the following: 1) Under a square wave of 10KHz we observe the output if it exists bounce or oscillations in the rising edge of pulse. If it is so, then the Miller capacitance – which it accompanies each common emitter arrangement - of the Voltage Amplifier stage it is not completely compensated. The Miller pole compensation it is a main nested feedback loop into the whole closed feedback loop. Its complete compensation it is against the response in higher frequencies. 2) Again under a square wave of 10KHz we measure the Rise Time of the pulse in output. An amplifier with good response in mid - high frequencies should not have a Rise Time more than 2ìsec. There is an indirect calculation method of the frequency response in KHz by dividing the number 300 with the measured Rise Time. Thus for 2ìsec Rise Time we have: 300 / 2 = 150 KHz. 3) Again under a square wave of 10KHz we observe the increase of Rise Time by connecting a 8Ù dummy load in output. A good amplifier with good power supply should not show increase of Rise Time more than 5 percent. An acceptable increase is for example from 2ìsec without load to 2,1ìsec with load. 4) Under a square wave of 100Hz we observe the output. The tilt of the duty cycle of pulse it gives very useful information of response in low frequencies, and with other words the force of amplifier to manage the bass assault during audition tests. As long as smaller it is the tilt (a few milliVolts difference between the leading edge and the trailing edge of pulse) so much better control has the amplifier in low frequencies. If indeed the amplifier bears in square wave of 20Hz without it burns fuses or activate the protection circuit, we check how many is increased the tilt of duty cycle from 100Hz to 20Hz. As long as smaller it is the increase so much higher Dumping Factor has the amplifier. Furthermore one useful information which given by the shape of tilt if it is curved and not straight, means that the feedback loop has no good coupling in low frequencies. Summarising, this type of measurements can substitute the measurements with an RTA under the condition that we know to interpret what we see in oscilloscope. Also it should they become even if we have a RTA as this gives an average graph, while with them we examine in concrete and critical frequencies the reaction of an amplifier. Consequently they are more near with audition tests in practice.
Regards
Fotios

[Pocoyo]

What's made a good amp.. ?

M O N E Y

you can prove it

[auriches]

Originally posted by jeffry_widjaja
What's made a good amp.. ?

M O N E Y

you can prove it

hi,

how much money is needed to make a good amp? what's the average cost of the amplifiers you guys make here?

also, do those really expensive transformers, resistors, caps and switches really make a big difference to the quality of the sound?

i'm a newbie, so thanks for any input and your patience.

[Pocoyo]

The best think you must consider is "BUDGET"

After that you can decide what is the next step ?

like :
- make your own pcb or buy somebody else ?
- What amp character you will build ?
- What grade component you will use ?

You can make any amplifier, if you use different
component maybe you got a little different sound

But This Little Differences Will Annoying You
And You Start To Do Some Mods And ...
Welcome To The Club

[MaxS]

Hi,

Love is an other way to get a good amp. If you love it, you will match the transistors, check the solders. And then, enjoy its sound !

Max

[fotios]

Originally posted by auriches


hi,

how much money is needed to make a good amp? what's the average cost of the amplifiers you guys make here?

also, do those really expensive transformers, resistors, caps and switches really make a big difference to the quality of the sound?

i'm a newbie, so thanks for any input and your patience.

For example, the cost of this 2X170Wrms/8Ù class B amplifier, as it is presented in the photo, it is about 350Euro.
Fotios

[The golden mean]

As I own a Dynamic Precision A1 power amp (Norwegian design) and like it,I have to agree with most of it´s constructors claims

http://www.dynamicprecision.no/Pages..._index_UK.html please click on "Philosophy".

I recently read an interview(in HiFi+) with some leading persons on the Swiss Goldmund company, and one claim was that our brain is very sensitive to time delays (down to 20 picoseconds), therefor the latest Goldmund amps are internally superfast,with some custom built transistors unavailable to us the amateurs...But filters on input and output of the amps,makes me a bit doubtful.Internally the amps had a delay in the low nanosecond area.My own belief is that transient information, is a very substantial part of the music.Thus I like Thiel&Partner or as the also are named: Accutone ceramic drivers.And Eton ER4 Air motion transformers...I have a superfast amp for the treble,btw.
Naturally , any fast amplifier has to be stable when driving real world loads!

Their(the Goldmund people) views on keeping the signal in the digital domain for as long as possible ,with the possibility to compensate for time delays in ordinary crossovers in loudspeaker,(through DSP),made sense to me.

[Pocoyo]

Theres a lot of range about amplifier value

- in power : from headphone to 1000 w
more power of course more budget

- in quality : theres general to special component
Build small amp with cheapest and other the expensive
If you use ( Ex ) Ni VX better Than BGN, tell it to me

- in fashion : good look casing, better heatsink etc
More smiling you have more money you pay

Im active to read several topics like solid state, digital
part and component, vendoor bazar & trading post
I read day & night, sometimes my wife unpatient

I build my amp too, you can look at
AussieAmplifiers NX150 Spyder

You can calculate how much i spend the money
http://www.aussieamplifiers.com/index3.htm

Not include heatsink, casing, transformer, wiring and
connector. At the first time i just want buy creek amp
5350 but why dont we move the budget to nx150