Addressing John Curl's concerns on low noise designs

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

scott wurcer

R.I.P
Joined 2004
Schmook/Curl MC step up

Here’s another one for your list lineup. I always wanted to try this. 4:1 is almost a perfect ratio for the Schmook/Gilbert multi-tan input. Here is a fully complimentary floating version, an homage to John Curl as it were. This circuit floats on a 9V battery. The linearized transfer function is pictured below, in sim almost -100dB THD at 250mV p-p out. I bet with 2N4401/2N4403’s pulled off of a strip you could get it to work. Noise is sub 1 nano-volt.

Offered up as food for thought, trims and bypassing could be added. You could even get the cartridge current to 0.

The file size limit seems to have dithered away a couple of lines, you should be able to fill in (it's only a couple).
 

Attachments

  • scmoo.jpg
    scmoo.jpg
    92.4 KB · Views: 528
john curl

john curl

Member
Joined 2003
My design was made to operate to a maximum of 100mV rms. It was a pre-preamp, designed to replace a transformer. A normal 30-40 dB gain phono preamp followed this circuit. I became famous with this design, because it could have 0.4nV/rt Hz noise, virtually unheard of, in 1973. but it is flawed, in that it has limited dynamic range.
 
Wavebourn

Wavebourn

Member
Joined 2006
wavebourn.com
john curl said:
My design was made to operate to a maximum of 100mV rms. It was a pre-preamp, designed to replace a transformer. A normal 30-40 dB gain phono preamp followed this circuit. I became famous with this design, because it could have 0.4nV/rt Hz noise, virtually unheard of, in 1973. but it is flawed, in that it has limited dynamic range.
Click to expand...

I see, John. I mean, pair of transistors, collector to collector, with feedback from collectors to bases.

Why you did not increase collector - emitter voltages?
 
lineup

lineup

Member
Joined 2005
john curl said:
Lineup, I regret that I am retarded when it comes to computers. I can't easily create a link, and I can't post anything directly.
However, I can tell you the 3 laws of design: Developed from Clarke's Laws, and put into Murphy's laws over the decades.

1. It won't work.
2. It works, but it's not important.
3. We invented it.
This and its variations have been around for decades AND can be found by anyone who is energetic enough to Google:
Click to expand...

Thanks, John.
In 2 days is my birthday (22 May). Was born by my ,now dead, mother 1951-05-22.
If I read comments to my 2 posted circuits (which are not really my ideas from start)
I see me an Curl has similar understanding.
I have never claimed they are top of designs possiblee.
I have presented them to show other other diyAudio hobby members
how you easily (with less parts or many lowcost) can build fairly good amplifiers.

syn08
If noise in 270 Ohm resistor is major concern .. this does really show Some Potential! of the rest of circuit.
Now 16 transistors are many, but even a small PCB can easily harbour them.
Metalfilm is understood.

After syn08 commemnt I might look into chainging 270 resistor.
Two 470 Ohm in parallell?


Curl
The second 2x LSK389 NE5534 I am not surprised you know about.
And you understand the circuit as you have tried it long before I first read about it.

Someone said it was something wrong with attaching the replacement outside input compoenents this way.
But if we know really abiout Classsic old NE5534 from Signetiz
we can see how it works.
NE5534 makes good use of all 8 pins. As good we can do it without making it all discrete.
And so We can replace input transistors of NE5534
And possibly using modern discrete lownoise JFET devices, like transforming into one JFET Op, in my case.
This will impropve the high input bias current circuit of NE5534.
Bias Current noise is one factor too :)


Thanks for explaining John Curl.
I had trouible with my former Harddrive.
Your compedium is not yet in my new SATA HDD

But you can be sure I will get i ;) :).
Wasnt Steve Eddy offering a link to download .. besides www.DiyHifi.org

Dimitri AND Curl are my searchwords.
 
S

stinius

Disabled Account
Joined 2008
lineup said:




syn08
If noise in 270 Ohm resistor is major concern .. this does really show Some Potential! of the rest of circuit.
Now 16 transistors are many, but even a small PCB can easily harbour them.
Metalfilm is understood.

After syn08 commemnt I might look into chainging 270 resistor.
Two 470 Ohm in parallell?


Lineup

This is very elementary. You should read syn08’s post one more time. He is giving you the formula for resistor noise. You can actually parallel as many transistors you like, but the main noise source will be the resistor.
By paralleling two 470 Ohm resistors you are now getting 235 Ohm instead of 270 Ohm so it will be slightly better, but not much.
Click to expand...
 
lineup

lineup

Member
Joined 2005
john curl said:
Lineup, I regret that I am retarded when it comes to computers. I can't easily create a link, and I can't post anything directly.
However, I can tell you the 3 laws of design: Developed from Clarke's Laws, and put into Murphy's laws over the decades.
1. It won't work.
2. It works, but it's not important.
3. We invented it.
Click to expand...
lineup said:

Thanks, John.
In 2 days is my birthday (22 May). Was born by my, now dead, mother 1951-05-22.
----

Thanks for explaining John Curl.
I had trouible with my former Harddrive.
Your compedium is not yet in my new SATA HDD
But you can be sure I will get it ;)
Click to expand...
:cool: Fine!
I feel like a rich man again. Have John Curl files onboard!
Because now I have found your (dimitri) compendium again.
I like to call it John Curl On Audio.
Most of it, but not necessarily every line, is still your thinking, I guess?
For those still interested ... hehe ... in reading some John Curl quotes

I provide this temporary download link at my personal webserver here:
http://lineup.homeftp.net/pub/

I hope this link will work. And for a couple of weeks, at least.
But better you get those papers now .. while you can :)
Hope you do not mind, John Curl?

Any other papers people would like to have ??
and I can consult my vast, big, gigantic Audio Schmeatics and PDF archives
and put in same place for download :cool:


Regards
lineup
 
syn08

syn08

R.I.P.
Joined 2005
synaesthesia.ca
john curl said:
That is why my MC preamp stages run open loop. It is almost difficult to fix, without using a power amp to drive the feedback resistors.
Click to expand...

And live with the distortions. It seems like today is a day you don't care about.

BTW, can you tell us (before I do, 'cause I got the reverse engineered schematics) what is the Vendetta's headroom aka dynamic range?
 
L

Lumba Ogir

Member
Joined 2005
Edmond,
I'm impressed by the number of input trannies
Click to expand...
I`m not. It won`t sound good. Sound quality decreases with the number of paralleled devices.

One conclusion of the discussion in this thread could be that, due to the alarmingly nonlinear Vbe/Ice characteristic of bipolars and some other annoyances also mentioned, makes them inferior to FETs in transconductance stages. (That´s why the more concordant functional Ibe/Ice relationship often is used for calculations).

stinius,
any objection?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom