Incredible quality amplifier by Graham, prepare your ears for it

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all ,
hmm, I thought of takes an oath PCB of this circuit, but I'm lost in Your problems of c-mil,bias, oscilations etc....
I have read that Graham has chosen of "heats" even the VAS, then I deduce that he has done a diff-pair +current mirror on a card and all the rest fixed to the headsink. I have seen the Carlos trace, but ( without offence ) you never won't convince me to connect the power bjts and the VAS with the wires ( rather don't do it ).
The motive is that here connections introduce much more inductance and cap. of that hypothesize from the theoretical model, for which to the end "listens" something of entirely different of the simulator. It is ineffectual (shareable) theories behaviour on the nH among IN and out and introduce uncontrolled variables of hundreds of nH and pF !
Person am even discordant with the separation of the diff- pair from the VAS, on account of the enormous sensibility that that connection has in the comparisons of the EMI.
It is possible to take an oath PCB, to understand, similar to the Krell KSA50 clone?

The 600 ohms of input; I have thought automatically to putting the 10K (and ~12K//390+470µF etc... on NFB), to avoid to using on preamp of " super buffered ". Some problem?

Obviously I use the 2SC5200 in place of 2SC3281 , but I don't believe that this is a problem...

Known breve to "defence" of d. Self;
( Graham to depart ) the technical difference among the "fantastic" circuits that see on this forum and his is very, very little. On the other hand he says clearly to working with a "classical" circuit.
I find "annoying" this aversion toward the D.Self circuits ( are not my absolute references, but I have feel of much worse!). If it doesn't like you, they throw out of the Your great analytical ability and modify them you.
I don't exclude to think to apply the (good) Graham ideas to my "Load Invariant", but for the prejudices have not simulators...

Ciao

Mauro
 
Hi Mauro,

You ask if 10k input/NFB-sense would be a problem.

I have not tried this, so I cannot answer.
I go for lower impedance at the differential bases to minimise voltage field pick-up.
(This reduces the risk of oscillation when longer wires and a star earth are used.)
I can touch the base of the diff NFB transistor with wetted finger on screwdriver and the amp remains unaffected. On other '10k' amplifiers this causes oscillation !

For all I know (and if it is stable) 10k input with 10k/15pF/390ohm/470uF NFB might actually 'sound sweeter' in the treble region.

However, it is also the case that less accurate amplifiers can sound sweeter because they muddy the hf distortions/noise generated by an audio source.

My aim is accuracy - not a 'sweeter sounding' treble.

If necessary I would rather make the treble 'sound sweeter' before it leaves the source or pre-amp, than have the amplifier modifying all input.

I note that Carlos has already fitted an additional capacitor across the input of his amplifier in this regard, and there is nothing wrong with his personal choice.

I don't see any problem with having the VAS base driven by a short wire either.
Inductance is not a problem because the base current variation is too low, and the predominant voltage field is out of phase, thus any wire capacitance acts like an extremely low value Miller C.dom.

I do however feel that the VAS, Zobel components and all output stage wires need to be kept away from the input diff/mirror transistors and input cabling.

Also, sitting back here now and thinking about the additional signal wire lengths that Carlos uses between his pcb, his drivers and the outputs - these are all in phase and at the same voltage field potential, so maybe I was wrong to criticise his method of construction (apart from twisting them - which can introduce inter-electrode coupling plus signal path capacitance to supply rails).

Maybe Carlos did have something to to teach us after-all !
He has been building amplifiers this way for a long time, where you and I have suggested that it is wrong.
I said he should wire everything after the input stage on the heatsink, but he has not, and he does not have oscillation !
You say you would not separate VAS from diff, but I do, and I do not have oscillation !
The only wire that is actually out of phase with the class-AB output stage is the constant class-A 'base' feed, and this is immune to voltage field induced feedback due to its low source impedance.

I suppose that until different constructions are actually auditioned side by side, then none of our suppositions can be authoratative, even though we might all have our own construction 'ideals'.


Super buffered at 600 ohm ?
Top flight commercial pre-amp/source equipments are 50ohm drive/headphone capable.


Cheers ........ Graham
 
Yes Graham, i feel afraid of VAS connected with wires too, but when VAS works hot, i

Use to put them 4 to 5 centimeters away from the board.

They normally do not oscilate, in the reality, i do not remember the last time i had oscilations because of that distance.

Of course i cannot say that this is a good practice, as it is not, but certainly, as you could find theorical bases to explain the non oscilations i had, something more than be lucky is happening.

As i used to construct Radio Frequency Linear amplifiers to HAM radio, and already made a good quantity, i had some idea of the care we need to avoid pick up and oscilations....but really, i never worried too much related audio frequencies.

With the exception of the input stages, the hi gain stages including the VAS, i use to wire... and there are moments that i need 20 centimeter twisted wires, and absolutelly no problems with Output and drivers.

I also, alike the forum majority thinks, prefer to install all transistors around the board, with the more short leads as possible, avoiding output to input coupling.

No, i do not want to create any "revoluctionary" new "mess assemble"...but i am forced, by circunstances, to stand by in the position that those wiring are not the terrible thing we use to imagine.

regards,

Carlos
 
Hi Carlos, you did a very sincere description of your layout, and your experience with RF circuits you allows to understand my perplexity.

Hi Graham, thanks of your seem. When I find the time, will look for to draw a prototype to the height of the circuit.

I have worked much with video circuits and digital ( plan and develop complete ), and then you will understand that are accustomed to " hold for the horns " the problems of joining among the stage.
In truth in audio environment ( professional, because the diys perform miracles in comparison with the informations that they have!) I am much a few the developers of competent PCB.
I would be able say much on does takes an oath mediocre layout to a amp, in departs has thought D. Self in his book, but it is only the peak of an iceberg...

Super buffer? he , he, a video circuit to 1000 V/uS 100MHz 50 ohms out to 20 Euro /Pz in accordance with my parameters is a " super buffer " .
If had to over dimensionl a video circuit with this criterions, have to use the micro-waves ! .
I don't amaze me if then it is have to use of the super-PSU, but then they serve the super- trafo, but then they serve the super-main power... opppsss but this is the " audio reality ", excuse, I thought to speak I owe you a lot of money reality...:D :D

thanks and ciao

Mauro
 
To Mr. Carlos

Dear Sir, a little bit of the topic :)
It is great to have such a funny and full with creative energy guy on the forum.....:smash:
A question..... on one or two replies on this thread You have said that You have made a hood amp (I think an 1969 version), and also You have said that mj15003 are a little bit dull, dark or something - - I have made hood amp 1996 version with an ccs at the input and also at the driver stage. It is with mj15003 at the output.
Did You try that version? Also, I would like to maybe convert to TO247 devices.... have You any experience with some of those devices and hood amp? ....... on other thread Geoff has recomended me to use mjl21194..... what do You think- any recomendations?

also - i was impressed how hood sounds - it is good amp - wonder how mr. grahams amp wpuld sound since You have said so much good about it......:bigeyes: :smash: :)
best regards
 
Mauropenasa and Sunrise, thank you about comments.

Mauro, vá molto benne signore, it is all rigth man.

I am your brother in profession Mauro, all life i worked in Broadcasting Television Stations, starting in the Engineering area, responsable of Operations and also Maintenance, having some studies made in Atsugy Plant, Sony Japan.

The biggest fun with audio, our lovely toy... is that we can do all foolishes that we cannot do with video and radio frequency...it is very relaxing, and easy, when you are not too old and damaged.

Of course i understand your point of view, and we do not think too much different, but here, beeing a diy, i can do whatever i want, construct over wooden pieces, cross and parallelling wires, unmatching impedances, i can let output without terminations...in other words...life is great and free here...my professional obligations stay there, in the television tower, the power Transmitters.... here i use to put transformer in series, in parallell and all king of strange things....and the better.... IT WORKS!

Sunrise....hummm, perfect name for a sensitive guy, that can capture feelings hidden behind words....yeah!>.. i am deeply alive and happy as a hell..... having my bad moments also...but more happy than sad...thank you...many thanks your kindness to "see me"...with your heart eyes.

I made some JLH, as i do not keep constructions made...i use to dismount, not only to avoid too much objects inside my home, but also to not stop too much money in parts...so...the same old TIP3055 and some RCA3055 are running here, having runned in many circuits.

JLH was assembled with 2N3055, with TIP3055, with TIP41 and With 2SC2922.... this last one a high speed device that reproduce a little bit more brigth and more agressive Saxophone, and high frequency details you can have when you scratch a guitar with your fingers.... the "quinch" sound...scrinch if you prefer. If this noise is heard, more details you will have in sound reproduction

The JLH, as beeing class A, is very tollerant to use other transistors, respecting current and voltage limits.

I prefer not to suggest a transistor.

In place of that, i prefer to suggest you to make a prototype, ugly and fast to hear, and them test the output units you have, as they are more critical than the other stages, related frequency response, more difficult they are, to reproduce nice trebles than a lower power units (this is practical result, not theorical).

Better than that, try the DOZ, as it is simple and very easy, and a JLH design very well adapted to a little more modern components than the 1969 version.... and the more important, take it as a "reference" amplifier, to compare with all you made and will make in future...this way, you will have some "language", that more than 70 percent of forum guys will understand.

-Better than JLH in this or that
- Worst than JLH in this or that
- Dinamics equal JLH
- Power much better than

Using this same reference, Graham amplifier is better in power, in dinamic and in bass reproduction related JLH

Do not know about sound stage, as i used not to construct more than one channel.

regards Sunrise...wonderfull name!

Carlos
 

Attachments

  • cartão radio amador 2.jpg
    cartão radio amador 2.jpg
    67.8 KB · Views: 769
:cannotbe: mr. carlos - are U trying to break a sound barrier.... I think that this was the fastest reply I ever get. An very information full..... thanks for that.....
I have allready made hood amp stereo with 21-0-21V and 1.5Amp bias and with mj15003 from onsemi and with 2n3055 from st. mj***** are better, but ...... :D
they seem a little bit dull in high freq's.......
maybe I will make an graham amp for myself and finish this hood amp for my friend so that we could compare them..... will see ..... anyway thanks for Your ultrafast reply
:D
best regards
 
:)
BTW.... would I ask to much of You mr. Carlos if I would ask to put the part manufacturers - what You have used here in Your amp..... if this sounds good it would be very interesting for me to know what have You put in it...... what snake oil :clown:
really - i am between three sides - at home I have an Onix power amp OA601, powered by line amp made arround 12B4A, home made dac- and it works nice - well worked nice until hood amp for my friend came to be tested if everything is working properly..... My Onix is a very good amp but hood - that little devil - just beets him with no respect.....:bawling:
And now I was searching for some other parts for the hood amp since it's output transistors are a little bit hot because of the heatsink setup and I have stumbled on this amplifier - :bawling: :censored:
oh boy - nice hobby.......:D
 
Here is the "snake oil" Sunrise.

-For low level aplications, differential amplifiers, input, where
high gain and low noise are needed

2N5401 and 2N5551 - Fairchild

-For CCS and similar applications, and sometimes to input;

BC546 and BC556 - Philips Semiconductors

-For Higher current, to Constant Current Supply (CCS);

BC639 and BC640

-For driver applications, or Voltage amplifier:

2SA968 and 2SC2238 - Toshiba

-Driver for limited voltage, 40 volts simetrical and Big current CCS;

BD139 and BD 140 - Philips Semiconductors

- Output applications;

2SC2922 and 2SA1216- Sanken

In the Graham amplifier i used some borrowed transistors from another circuit, some excelent plastic transistors (published in that thread) that are hard to find here...so, the 2SC2238 are more used as driver.
There are many good transistors, the ones you use are wonderfull and will not loose from the ones i have...in special those Motorola, those MJs you have.

Sound quality is a result of many things, and some of them already waiting discovery.... thopologie, good design, care in parts selection, adequated transistors to voltage and current application are the basic things.

Simple circuits use to be good too.

regards,

Carlos
 
JLH designs switch on our emotions.

If the music is sad, you may cry.

If the music is happy, you may dance.

You can hear for long time without turn tired off.

Those quality amplifiers, produce two strong emotions, but those emotions are adequated in the order they appear to you;

The JLH amplifier produce happyness when you switch it on, and produce sadness when you switch it off, because will not hear more that good sound that made you dream.

Others amplifiers, the non quality ones, produce also two emotions:

the first is when you switch on, a enormous deceptive feeling, and enormous happyness when you switch off that hell thing!..beeing free of the torture to hear it.

Let's say that the emotion polarity is inverted, one each other.

Have to listen to understand.... i do not think someone can inform emotions with words....i can describe a kiss in thousand words, but a love kiss, with deep love and passion, no word can explain.

Sun, despite i love Sunrise, i cannot continue with our "ping pong" mails, because the thread is talking about Graham amplifier, i am inviting myself and you, dear Sunrise, to go with me into the rails again... (because forum rules and good sense needs, to make it work fine)

Despite i like that kind of more personal contact messages, directed to a target person, we already made 10 repeated messages one to the other, let's open space to other guys enter without the feeling that are interrupting something.

regards,

Carlos
 
Oh! Sunny, i was the one that felt very sorry, as i had to say that to you, to avoid

moderators interference, as this is a big shame to us, when adverted or even punished.

They are very patient, i could see that many times, and all them very good guys, well selected, and when their actitudes are not reasonable, they disappear from the forum.

To talk directly, and personally, exactly the way i like to do...as doing directly to you now, is not adequate, as many people have to read our dialogue, and this must be done to all communitty, always talking with everybody.

As i am a little bit older than you,not only related the forum age, but maybe also related real age (54), really, intentions are the best.

I am sorry...really sorry...but did that to avoid more disconfort to us.

But if you like to talk with me, i feel very proud to produce a new friend, and my personnal mail is open to you everytime you need.

nanabrother@yahoo.com

Yours,

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