| DarkHorse |
hi:
have a look with this circuit diagram,can it works well?
I draw this circuit diagram last night...
RP1 control the currency of output stage,set it to 50K at first;
RP2 control the voltage of ouput point(0.22V+1/2Ucc);
SK170 can replace with sk389 or NPD5564(twins);
;) |
|
|
| Upupa Epops |
| 4 W amp with 1.1 bias current ? What are you waiting from this " miracle " ? |
|
|
| triode_al |
| quote: | Originally posted by DarkHorse
can it works well?
;) |
I'd take the gate resistor R8 of the top IRF510 higher --> 510k too, and add a gate resitor >> 10 ohms.
The input/driver will have reduced distortion as you take a higher Vb there, say 50V. This might be a considerable reduction. It is also something that the feedback in the way you design it, cannot correct. This means that RP1 has to increase along the way.
It does have a Pass flavour to it that is nice. Its simple. What do we want more? Well, Q4 is not really needed, and by having a resister it should work out well too methinks. |
|
|
| padamiecki |
| quote: | Originally posted by DarkHorse
hi:
have a look with this circuit diagram,can it works well?
I draw this circuit diagram last night...
RP1 control the currency of output stage,set it to 50K at first;
RP2 control the voltage of ouput point(0.22V+1/2Ucc);
SK170 can replace with sk389 or NPD5564(twins);
;) |
it is fet/mosfet version of Aikido Amplifier by John Broskie, TCJ |
|
|
| DarkHorse |
| quote: | Originally posted by Upupa Epops
4 W amp with 1.1 bias current ? What are you waiting from this " miracle " ? |
hi,Upupa Epops :
4W(RMS) is 8W(peak)in sine-wave ,so load is 8ohm,and P=I^2*R,then I=1.0A,so I=1.1A is suitable!
driver stage using SE instead of using bootstrip or using hight-Voltage supply,and it can win more low distortion...
cheers
WINCO |
|
|
| DarkHorse |
| quote: | Originally posted by padamiecki
it is fet/mosfet version of Aikido Amplifier by John Broskie, TCJ |
I don't know where this diagram from!but it's topology so simply that make me so surprise...
this is original schematic which I referenced:
;) |
|
|
| Upupa Epops |
| If you take this connection as excercise, so go on. But with four transistors don't wait miracles, although they are relatively modern ones, all is too simply ;) . |
|
|
| DarkHorse |
| quote: | Originally posted by Upupa Epops
If you take this connection as excercise, so go on. But with four transistors don't wait miracles, although they are relatively modern ones, all is too simply ;) . |
haha:
yes,I understand your meaning.
but as a small "kid"(only 4W output power),it's capability was enough strong,THD=0.15%,phase angle is +0.2 degree(20Hz) and -0.7 degree(20KHz).I think it is so great...
all for it's simply.I like it,and hope it's sound can sweet than ZEN2(the class-a amp I have alread build).. |
|
|
| Fuling |
| Note that the original circuit is a transconductance amp (current feedback). |
|
|
| DarkHorse |
| quote: | Originally posted by Fuling
Note that the original circuit is a transconductance amp (current feedback). |
hi:
yes,I have note this problem.
current FB is can cut-off the speaker's reverse voltage...
but distortion is too XXX(4%).
see you tomorrow. |
|
|
| Upupa Epops |
| Better sound than Zen ? Your references are too low :D . |
|
|
| padamiecki |
| quote: | Originally posted by DarkHorse
I don't know where this diagram from!but it's topology so simply that make me so surprise...
this is original schematic which I referenced:
;) |
you are right
my look was too quick
I have mismatched it with mentioned TCJ |
|
|
| DarkHorse |
I will build a sample to examine the sound class of this circuit,wait for my report please!about the day after tomorrow...I show you.
enjoy!
:smash: |
|
|
| tlf9999 |
There are substantial differences between yours and the original design but I like yours better.
Yours has the gain in the first and 2nd stages: both are drain-loaded-with-CCS and with global feedback. The original is a source-follower, followed by a drain-loaded with CCS for gain, without any feedback.
It is my experience that it is easier to have gain with small signal. So I was wondering if you can improve upon your design by using a source-follower in the 2nd stage (source follower + CCS).
You can then limit the feedback to the first stage as well (you can look into the Zen to see how to implement feedback in the first stage).
Let me know what you think. |
|
|
| DJNUBZ |
| I bet this would sound great with some high effeciency speakers. There is a lot you can do with 4 wrms a chan. |
|
|
| DarkHorse |
| quote: | Originally posted by tlf9999
There are substantial differences between yours and the original design but I like yours better.
Yours has the gain in the first and 2nd stages: both are drain-loaded-with-CCS and with global feedback. The original is a source-follower, followed by a drain-loaded with CCS for gain, without any feedback.
It is my experience that it is easier to have gain with small signal. So I was wondering if you can improve upon your design by using a source-follower in the 2nd stage (source follower + CCS).
You can then limit the feedback to the first stage as well (you can look into the Zen to see how to implement feedback in the first stage).
Let me know what you think. |
hi,tlf9999:
At first I really don't know would it sounds better than the original version, because global feesback can cause other unexpectable problem...maybe it just decrease distortion in parameter..low distortion not mean it can sounds well..and I am so afraid the speaker's reverse voltage loop back and effect the first gain stage..
I don't want to use source-follower,I think IRF510's C-input is 180pF,driver current in 10mA is suitable.too strong driver stage perhaps let sound comes too "hard"(not sweet),I think..
but if you want use IRF530 or IRF540..this source-follower is needed to high frequency(10KHz or more)..
| quote: | Originally posted by DJNUBZ
I bet this would sound great with some high effeciency speakers. There is a lot you can do with 4 wrms a chan. |
hi,DJNUBZ:
in fact,this circuit can output 4.5W(RMS) with 1.1A bias current.
have fun!
WINCO
:D |
|
|
| DarkHorse |
Hello,UltimateX86:
thanks your schematics,the topology is OK:
;) |
|
|
| Ipanema |
Hi Ultimate86,
Your circuit looks like a Le' Monstre variation. Have you build those?
Regards,
Ipanema |
|
|
| DarkHorse |
| look my works,I soldering this last night...but too late,didn't finish yet! |
|
|
| DarkHorse |
main section closer view,I didn't find R11(47ohm),so I replace it to 68ohm,and R14 is 1.5Kohm,gain about 23:
:smash: |
|
|
| DarkHorse |
the PCB,I cut the line wire using a knife,is a hard work!!!
:bawling: |
|
|
| DarkHorse |
and using SEC's IRF510(second hand).
I will test it tonight,god bless me!
:cool: |
|
|
| DarkHorse |
hi,all:
I test it last night.
it works abnormality ,that is to say---it doesn't works well,
result show it has about 0.5Hz surge,because first stage's low frequency gain is too strong,R3 should not connect to feedback loop,it's positive feedback...
so I decide to use other bias way(SK170's)...
look it's ac analysis graph: |
|
|
| DarkHorse |
Take it easy!I will go on tonight,it has more suitable topology,and lower distortion...I believe it can sounds perfect!
the test in last night,I relocate R3(down end) connect to ground.
the surge problem removed,but the output point of first stage upto about 14V,no matter what,it sounds,and sounds pretty good,I can't hear any noise from the speaker,I only use one 10000uF/50V cap and shunted with one 10uF/100V MKP cap,
ouput stage's bias current is about 1A..
step by step,I walk to success...haha
this is updated version of schematic:
:D |
|
|
| padamiecki |
cheers!
could you unconnect nfb and describe the sound after deconnecting? |
|
|
| DarkHorse |
| quote: | Originally posted by padamiecki
cheers!
could you unconnect nfb and describe the sound after deconnecting? |
hello, padamiecki :
nfb???
it's hard to say that now,but I will tell you tomorrow!
regards
WINCO
:smash: |
|
|
| DarkHorse |
:bawling:
fail again!!!
the super low surge(0.5Hz) still existing!
I suppose it's open loop gain rate is too high,and the feedback is too deep,so it occur...isn't it???
last night,after the failed testing..I had to remove the Q4 and RP3,
replace with one resistor,the value is 1Kohm...it sounds:
the sound is so soften and sweet,put my ear close to the loud speaker,about 0.3 meter(1'),I only hear the heat noise of semiconductor device..no 50Hz low frequency noise---good!
sound quality is more than my ZEN2,I hear Michael Jackson's "billie jeans"...you can lisend it for a long time,didn't feel any uncomfortable and any noisily...
I will go on!:mad: |
|
|
| padamiecki |
I have used he same circuit,
but without nfbes,
even local
my posts should be somewhere here in solid state forum
but for some person the sound was too weak.
With nfb connected the sound becamed nervous. |
|
|
| DarkHorse |
| quote: | Originally posted by padamiecki
I have used he same circuit,
but without nfbes,
even local
my posts should be somewhere here in solid state forum
but for some person the sound was too weak.
With nfb connected the sound becamed nervous. |
hi:
your "nfb" is means "globe feedback"?
I think "nfb"is "none globe feedback".
but not all "globe feedback" connect would make sound became nervous,only occur on which circuit using high power transistor(with big input capacitor,as if IRF044,IRF140..),and amp has high gain rate,and deep feedback loop,this C-input will effect the sounds,phase compensator almost using a cap(10pf---47pf) shunts with feedback resistor,lead to "TIM" distortion...
I think if the output transistor(IRF510) has small C-input,it's effect will reduce more,and it can sounds sweet too...
cheers!
WINCO
;) |
|
|
| padamiecki |
| quote: | Originally posted by DarkHorse
I think if the output transistor(IRF510) has small C-input,it's effect will reduce more,and it can sounds sweet too...
cheers!
WINCO
;) |
maybe, who knows? :spin: |
|
|
| UltimateX86 |
| quote: | Originally posted by padamiecki
I have used he same circuit,
but without nfbes,
even local
my posts should be somewhere here in solid state forum
but for some person the sound was too weak.
With nfb connected the sound becamed nervous. |
what is certain, it is that French likes it |
|
|
| padamiecki |
| quote: | Originally posted by UltimateX86
what is certain, it is that French likes it |
so maybe is there in the world a definition "The French sound?" |
|
|
| tlf9999 |
Here is my variation.
It is essentially the same front end as DarkHorse's design. Except that in the 2nd stage, rather than using a drain-loaded output stage with gain, mine is a source follower.
Simulated quite well, and harmonics are mostly 2nd, 3rd and 4th, with the 2nd harmonics dominating.
Warning: conceptual design, never meant to be built, not optimized. |
|
|
| tlf9999 |
| One quick comment: the feedback pick-up point (right side of R3) can be at either the MOSFET's source or on the load. |
|
|
| bocka |
Hi DarkHouse
| quote: | | but the output point of first stage upto about 14V,no matter what |
You MUST NOT load a collector or drain follower with a current source when there is no way that global feedback sets your DC operating point. This is a no no and only works in simulation. .:whazzat: Both FETs are not equal. Use a resistor instead. Maybe this is also the problem for your 0.5 Hz surge |
|
|
| DarkHorse |
| quote: | Originally posted by bocka
Hi DarkHouse
You MUST NOT load a collector or drain follower with a current source when there is no way that global feedback sets your DC operating point. This is a no no and only works in simulation. .:whazzat: Both FETs are not equal. Use a resistor instead. Maybe this is also the problem for your 0.5 Hz surge |
hi,bocka:
yes, you are right!I remove R14 away,the circuit can works well,and I have try 10Kohm /7.5Kohm /5.1Kohm of R14,I find R14 must large than 5.1Kohm that the circuit can works!
let's see the last passed test circuit: |
|
|
| DarkHorse |
as a class-A amp with 1.1A current bias,let it's input end shunts to the ground,put your ear to the speaker,nothing you can hear!
the filter cap just only 10000uF/50V and 10uF MKP cap...
it sounds so softly and sweet...I like it! |
|
|
| lineup |
| quote: | Originally posted by DarkHorse
as a class-A amp with 1.1A current bias,
let it's input end shunts to the ground,
put your ear to the speaker, nothing you can hear!
the filter cap just only 10000uF/50V and 10uF MKP cap...
it sounds so softly and sweet...I like it! |
can such a simple class a mosfet amplifier really perform
????
would be interesting to know, really ;)
Is my attachment a correct schematic ??
of The Final Version of this All FET Class A amplifier |
|
|
|
