The JC-2 preamp schematic on the website is wrong...Mr.John C?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The JC-2 preamp Sch is on the famouse unoffical website of Mark Levinson

http://marklev.com/

After altering by another man from China,I just found the wrong in the ouput module sch.

http://www.hifidiy.com/dispbbs.asp?boardID=13&RootID=68247&ID=68247&skin=0

An externally hosted image should be here but it was not working when we last tested it.


In the ouput module(=line stage),the phase polarity of - input is same as the output.In the words,the feedback of this module is postive not negative,the output module is an OSC not a NFB amp.

To correct it,IMO,the base pin of 2N4401/2N4402 should be exchanged with the emitter pin.I wonder if anyone had rebuilded the JC-2 according to the marklev.com schematic.

Mr.John C and the other member,do you agree with me?

Thanks your any input.
 
diyAudio Retiree
Joined 2002
and the other member

Two inverting stages give a non inverted output. Looks like positive feedback to me.

"To correct it,IMO,the base pin of 2N4401/2N4402 should be exchanged with the emitter"

No, swap the feedback and input gates of the jfets to get a noninverting stage for to jfet pairs. I believe most all of the voltage gain is in the bipolar second stage to reduce the Miller capacitance at the amplifier inputs. You will need jfets with Idss of at least 8 mA to get this circuit to work with 100 ohm drain resistors for the fets. Since most of the Toshiba jfets ideal for this circuit that are easy to find are the GR bias group (Idss from 2.6 to 6.5 mA), you may want to increase the drain resistor value and cascode the jfet. I would use 20V rails if cascoding the input stage and bias jfets at least 10 Volts drain to source to get the devices capacitance down.
 

Attachments

  • k389capacitance.gif
    k389capacitance.gif
    8 KB · Views: 3,837
X.G. said:
The JC-2 preamp

In the ouput module(=line stage),the phase polarity of - input is same as the output.In the words,the feedback of this module is postive not negative,the output module is an OSC not a NFB amp.


Mr.John C and the other member,do you agree with me?

Thanks your any input.



Hi XG!!!

You are right!!! The feedback is positif!!!

The overall feedback must be connected to the + input making a inverting amp with shunt feedback!

Here ,you have from the some author, another project ,that is essencialy the same in the voltage gain stages but where the feedback is correctely connected.



http://marklev.com/JC3/jc3schematics.jpg

Cheers ;)
 
Disabled Account
Joined 2003
X.G. said:
To correct it,IMO,the base pin of 2N4401/2N4402 should be exchanged with the emitter pin.I wonder if anyone had rebuilded the JC-2 according to the marklev.com schematic.

Thanks your any input.


I believe in the JC3 discussion, John said that this is one of those rare amps with inverting input. so all you need to do is to tie the feedback to the inverting input end of the amp (+ input in your schematic), like in the JC3 schematic.
 
Disabled Account
Joined 2003
I found another way.

you can swap the drains of the Jfets and leave the rest unchanged. for example, the left E212 should have its drain connected to the positive rail via a 100ohm resistor. and the right E212 should have its drail connected to the emitter of the 2n4402. the schematic looks like the attached chart. I got the ideal from the "is 150w enough" discussion of yesterday.


It essentially swap'd the two inputs and turning the original inverting amp (in the JC3) into a non-inverting design. I remember John talking about why he wanted an inverting design back then.

edit: the emitter resistor in my previous chart had the wrong value.

edit 2: hm, the new chart didn't get uploaded. Anyway, R2 should be 20ohm to be consistent with the original design.
 

Attachments

  • jc2 modified.jpg
    jc2 modified.jpg
    41.6 KB · Views: 4,330
Disabled Account
Joined 2003
Mr. Curl, if you have time, I would like to get yours thoughts on how the current on the VAS and input stage is determined in the JC2 design.

It looks almost identical to the complimentory VAS we talked about in the 150w thread (here: http://www.diyaudio.com/forums/showthread.php?s=&threadid=28266) and we were struggling in a mighty way to understand it.

if you could share your thoughts on this with us, it would be greatly appreciated.

thanks in advance.
 
Can't a guy get a night's sleep without everyone wondering how this schematic could be possible? ;-) Well, the schematic is wrong. It was caught by me after this schematic was first published, in 1977, and was corrected by me in a LTE of 'The Audio Amateur' in the next issue or so.
Yes the polarity is inverted in the schematic.
You must reverse the connections to the emitter and base on both sides.
 
diyAudio Retiree
Joined 2002
For the love of ..............

Swap the gate leads or alternately the drain leads for both jfet pairs.
I don't what all this nonsense about swapping base and emitter is about. THIS IS A VERY SIMPLE CIRCUIT. There are only four transistors in it for crying out loud. It is not that difficult to understand. I don't understand what all the confusion is about.

From: http://marklev.com/JC2/index.html

"John Curl's correction to the Line module was that as it is in the drawing you got the signal from the E212 to the 2N4403 should have been taken from the OTHER E212. The same applies for the E175 and the 2N4401."

I am starting to wonder about some of you.......:whazzat:
 
Disabled Account
Joined 2003
Re: For the love of ..............

Fred Dieckmann said:
I don't what all this nonsense about swapping base and emitter is about. THIS IS A VERY SIMPLE CIRCUIT. There are only four transistors in it for crying out loud. It is not that difficult to understand. I don't understand what all the confusion is about.


Fred, maybe this is too simple for you but XG's solution of swapping B and E fixes the problem in a very elegent and simple way.

Of course, some of us are very gifted at finding an overly complicated solution instead, :)
 
This is a pretty good circuit, but don't try to use the original parts. It is better to use the Toshiba 2sk389 and 2sj109 parts for the input. The second stage can be any darn good 1/2A-2A rated transistor complement. This design comes from 1973, more than 30 years ago, and we used what was available then. The mistake in the schematic came from Mark Levinson's file that he gave to me, and I had transcribed by a tech. I should have seen the problem before it was published, but I didn't. This circuit is defined as a transconductance amplifier, rather than an op amp. It has a slew rate of about 100V/us and a fairly high open loop bandwidth. It is the basis of all my designs for later preamps and power amps. PS Millwood, you HAVE to use a smaller resistor in series with the output emitters or it won't work. 20 ohms is on the original schematic.
 
diyAudio Retiree
Joined 2002
no insult intended but.......

'Fred, maybe this is too simple for you but XG's solution of swapping B and E fixes the problem in a very elegent and simple way."

No it does not. Most of this circuit's votage gain is in the second stage. Swap the base and emitter and you have a totally different topology that does not have the open loop gain required. This is a pretty classic circuit after a quarter of a century and I've plenty of variations on it. I'm sorry if this ruffle anybody's feathers, but I think people would like to know how the circuit really works.
 
To Fred D ------the other member;)

Thanks your reply.
You posted an very good idea to advance the performance of the JC-2 output module,but...:smash: I just want to try achiece/keep the correct original sch of JC-2 and not to redesign/modify it,cause it was the great work by our master designer Mr. John Curl.;)

To Tube_Dude

thanks your encourage!!!:)

To millwood,Jan Didden

my meaning is...nearly same as millwood's 'another way' SCH.Sorry for my poor English:bawling: .Here it is the corrected SCH which redrew by my friend wongkm33 according to my opinion.My name is С¹íÍ· on here.
http://www.hifidiy.com/dispbbs.asp?boardID=13&RootID=68247&ID=73528&skin=0
An externally hosted image should be here but it was not working when we last tested it.


To John Curl

Thanks you honestly tell me the truth.I had tranported it to other Chinese DIYers,including your advise.If have enough time(I wish),I will rebuild the JC-2 line AMP using 2SK389/2SJ109,2SD669/2SB649(Hitachi).

In my mind,the JC-2 design still is perfect today. :cool:
 
AX tech editor
Joined 2002
Paid Member
Re: no insult intended but.......

Fred Dieckmann said:
'Fred, maybe this is too simple for you but XG's solution of swapping B and E fixes the problem in a very elegent and simple way."

No it does not. Most of this circuit's votage gain is in the second stage. Swap the base and emitter and you have a totally different topology that does not have the open loop gain required. This is a pretty classic circuit after a quarter of a century and I've plenty of variations on it. I'm sorry if this ruffle anybody's feathers, but I think people would like to know how the circuit really works.

Fred,

Don't want to upset you, but you either don't get it or you are to stubborn. The idea is to swap the B and E connections TO THE INPUT PAIR. It seems that everyone including JC agrees that that fixes it, except you. I know, if the majority selects a wrong it is still wrong, but in this case I think you are out in left field.

Jan Didden
 
Fred, think it through. Even with the leads switched, the second stage is still essentially a common emitter drive. I just added the alternate input lead to the emitter of the second stage because it seemed to be an OK thing to do. It does have some subtle advantages.
What you are thinking is that it becomes a common base second stage, and sometimes I do design circuits that way, but in this case the common emitter overwhelms the common base connection. However, if you put a large bypass cap at the base of each output transistor to ground, it would become common base driven and the circuit could practically only operate open loop. Both Chas Hansen and I design some of our latest circuits with a common base drive for the second stage. It is then called a folded cascode connection.
 
diyAudio Retiree
Joined 2002
Functionally equivalent?

"It is then called a folded cascode" Yes John....... I know what a folded cascode is and designed several in the last 15 years or so.

The 64 dollar question is was your JC2 correction changing the phase
of the feedback and input pins? I would also contend changing the second stage to fold cascode is much different circuit than the circuit with the above change. I am not trying to split hairs. I thought there was actually confusion on this point of the changes to the bipolar
stage being equilent to input pin polarity in performance. Surely it isn't.
Right?
 
Disabled Account
Joined 2003
Re: Functionally equivalent?

Fred Dieckmann said:
[BThe 64 dollar question is was your JC2 correction changing the phase
of the feedback and input pins? Right? [/B]


Fred, for most of us, that wold have been a 2 cent question (because it is so easy).

Just follow XG's redrawn schematic and re-arrange it a little and you will hopefully see the light of the day on this.

I cannot believe someone as famed as you are can struggle this much on such a simple circuitry.

a wise soul once said "when Curl belabor the point, it is definitely worth listening." and he has done so a couple times, in addition to others laboring on this point.
 
Disabled Account
Joined 2003
Re: no insult intended but.......

Fred Dieckmann said:
This is a pretty classic circuit after a quarter of a century and I've plenty of variations on it. I'm sorry if this ruffle anybody's feathers, but I think people would like to know how the circuit really works.

years of working or numbers of variations do not assure correctness on any matter, even something as simple as thing circuitry, Fred.

Yes, you are wrong, Fred. Sorry to tell you that, but no insult intended either. Just factual.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.