diyAudio (
-   Solid State (
-   -   3T HeadPh Amp. JFET-BJT-MOSFET (

lineup 27th May 2009 01:59 AM

Hi :cool:
I am working on a 3 Transistor hifi headphone amplifier.

1. Input uses N-JFET, like 2N4416 or 2N3819, which is more easily found.
2. One bipolar PNP turns the signal to correct phase
and drives the output stage.
3. Output IRF610 N-Channel MOSFET follower powers the load resistor into several 100mA Class A.

Be back with schematics, AC-analyses and Fourier for discussion/evaluation.
All is properly tested in virtual circuit with some of the best instruments we can find.
On the commercial ground my simulator costs way too much.

From this we can forsee, that this topic will not discuss my circuit.
But people will do more postings of the usual debate on using sims, paper or calculator
or design by listening to transistors output through headphones of different qualities
after throwing it quickly together with solder.
And hope you have a lucky day and that your luck will stay with you
as you change this and that capacitor or resistor.

So just another good amplifier idea is completely ruined by these off topic posters.
Those that are not really interested in design,
but only to show how clever they are and how good diyaudio builders
and how golden their audiophile ears are.
Those they were born with and have finetuned to perfection in many a sessions with a bunch of them other audio-jet-setters.


lineup 27th May 2009 02:17 PM

2 Attachment(s)
I wont change the basic concept = 3T
Neither I want any changes to my original idea
like fooking CCS ideas.
If you want to make your own ravishing unique amplifier
start your own topic and start worrying about problems in that circuit :D

But suggestions regarding what transistors are very welcome.
In such very basic and simple designs
we can not just throw in non-fitting devices
and think any 'tons' of feedback will cover up for a bad design idea.


Attached is the first public version.
Running @ the important first milli watt level, 1 mWatt into speaker.
Of course by trial and error there were version before this one.
They simple were nothing much to publish and get my well deserved critic for.
From all you fooking knowledege besserwissers out there :cool:

scott wurcer 27th May 2009 03:00 PM

lineup, How do you adjust the DC bias point? J305's can work here too.

lineup 27th May 2009 04:59 PM

2 Attachment(s)
I am beginning to think we could become good friends
- if ever we meet.

One good friend, not fiend, is enough to miss my dead mother less.
(passed away in a heavy brain stroke at 29 October,
I carried her wooden box to the hole my self on Friday 28 Nov 2008)

Any good man reading a lineup post is worth notice.
Any good man posting a comment, Nelson or Scott, is remembered for ever .. I have a daarn good memory for things!
Dont you mistake me for old and hard-to-remember syndrome
eventhough i just at the 22nd of May
celebrated my 2-years-before-60 birthday totally alone in my 1-room-flat
without a bdaycake or even one candle to light up my day.

You want proof I am not just talking, but I'm alive and living, too?

Here you are (in perfect destroyer X - Carlos M. fashion).

There is always one in a crowd.
One that does not give a daaaamn about not wearing a yellow jacket.
Perhaps he doesnt even have the money to buy suitable dull black tuxedo ..
And he is not very much afraid of crying out loud his tears
.. if and when they want to come down his cheeks :(

jam 27th May 2009 05:28 PM


Sorry to hear about your mother.

Looks like a good about current sourcing the output stage?

Hey, my birthday is also on the same day.



lineup 27th May 2009 08:15 PM

Scott Wurcer suggestion J305 input
2 Attachment(s)
I have, as John Curl usually like to put it (and from a scientific point of view this makes good sense)
maintained 'all things else equal'.

Which means
~1mA input (the default J305 is a 6mA IDDS)
~20mA in second BD140 stage to drive the input capacitance of MOSFET
~360mA in output IRF610

This below is the resulting circuit for Scott Wurcer suggestion to try
= one J305 N-Fet for input.

difference in gain is now almost x10
vs. ~ x5 in Lineup original PN4416A (almost equivalent to 2N4416 JFET)

In next post I will show the compare of Fourier analysis
at output level of First Milliwatt = 1 mW into 32 Headphone (fully resistive!)

lineup 27th May 2009 08:24 PM

2 Attachment(s)

The fourier show almost the same.
Only fractionally higher distortion in Scotts circuit,
with a lower level of even order harmonics especially 4th, 6th, 8th etc.

If we want to have, need as high voltage gain as x10 ( +20dB)
then go for J305 as Scott suggests.
This can be most suitable for HP:s with an impdance at 300-600 Ohms.
Like the very good best ones by Sennheiser.
Among world's finest, but very expensive, HeadPhones.

Otherwise, you get a more decent gain of x5 with 2N4416
as suggested by Lineup.

This is suitable for lower impedance HP:s like 32-100 Ohm.
Think for example Grado classics series of HiFi phones.

/lineup, circuit design - May 27th 2009

lineup 27th May 2009 11:30 PM

2 Attachment(s)
This is a rather interesting analysis.
I think my few readers, pupils if you like, will agree totally :)

AC Analys = Frequncy response + Phase behavior within the audioband.

We are most interested, of course, how one machine that is involved with sounds behave vs. audiable signals.
In this case via one pair of Headphones.

More specifically we are concerned with what we can hear.
We leave other things to the bats and the supersticious :D

Usually most music takes place between 50 - 10.000 Hz for a normal middleage white man.
Black women, according to tests, has got a slightly better hearing = larger spectrum they can hear.
John Curl, syn08 and older white skinned men like Scott and me
.. we just have to be happy with what we still can perceive.

Inside the audioband, we strive for a stright line, 'stright wire' behavior.
The ideal Amplifier is said to be like a very short wire :cool:

Well, friends ---
... using 2N3819 (25VDS) + BC560C(45VCE) for 2nd stage, I think we have come very close.
At least me my self. Maybe later you. too ;)

Comments are welcome.
This one uses 0.5mA in JFET and
~ 5mA in the PNP BC560C low noise high gain device.

Wavebourn 27th May 2009 11:38 PM

What if to remove that Miller cap, but to add a gate stopper?

scott wurcer 28th May 2009 12:26 AM


Originally posted by lineup
I am beginning to think we could become good friends
- if ever we meet.

lineup - Don't take such a simple component substitution so seriously. A lot of FET's would work in that position, just plug them in and see what you get no big deal.

You have yellow jackets and I had one of my father's best friends come to his wake after a whole liter of vodka so we suffer our friends because there is no other way.

All times are GMT. The time now is 09:21 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio