John Curl's Blowtorch preamplifier

Status
Not open for further replies.
Jump to Latest
#12,642
syn08 said:

Now, from a practical perspective: paralleling a reasonable number of low end/quality DACs may help linearizing the devices.
Click to expand...

May.
Even that would depend on the nature of the DAC error.
It would work only for noise-like errors.

If the DAC had a systematic missing code at digital value of +4096,
even paralleling 1000 DACs would not make that go away.

Noise is probably the least problem with DACs. _Much_ more noise
does not make a master tape sound bad.
IMD and idle tones seem much worse.

Gerhard
 
#12,643
gerhard said:


May.
Even that would depend on the nature of the DAC error.
It would work only for noise-like errors.

If the DAC had a systematic missing code at digital value of +4096,
even paralleling 1000 DACs would not make that go away.


Gerhard
Click to expand...

I gather the assumption is just that, reduction in random errors. Usually characterization finds most systematic errors. I wonder if anyone has recently done any number density measurements on CD data. Maybe pointless but a curiosity.
 
#12,644
Tubes!

"LAMM sounded best", but John! it is tube equipment, have you leer at tubes, I have ordered a lot of JFET Toshiba devices making my own preamp.
What shall I do? I thought that JFET is the only active component that's worth dealing with.

I do like tubes a lot but I'm weak in my belief, I once build a Trevor Lee design I was very happy with this RIAA it was together with my early JC-2 clone line pre.

Have any seen the Trevor Lee design I have lost the schematics.

Kamskoma
 
#12,645
CES 2009

CES 2009 Show report: Positive Feedback

John Curl was the Most Legendary Electronics Designer in the show. John designed the first Mark Levinson preamplifier and amplifier in the 1970s, which are widely recognized as the first solid-state designs that could give tubes a run for their money. John's most recent designs have been for Parasound, and those products have received rave reviews. John is resurrecting Vendetta Research, to allow him to produce no-holds-barred products (many consider his SCP-2 to still be the best phono preamplifier ever made). John is pictured here with Jim Bongiorno.

JC at the left
 

Attachments

  • 28_%20john%20curl%20&%20jim%20bongiorno.jpg
    28_%20john%20curl%20&%20jim%20bongiorno.jpg
    37.8 KB · Views: 481
#12,647
Originally posted by SY
The Buffalo schematic shows a 1u cap on the output. The DAC data sheet shows a 47u. Big difference. Neither is good with a unity gain 797 (if I am to believe the 797's designers), but oscillation is far more likely with 1u than 47u; the latter value will merely bog down the opamp's output and pretty much negate the advantage of using a high performance opamp in that position. You might be able to get away with it sometimes, but it can't be relied on and is poor engineering practice.
Click to expand...


What you call "poor engineering practice" is a circuit that works very well.


Originally posted by SY

I'm not sure why you think this is an exemplary design. There are far better ways of getting a stable, low noise supply.
Click to expand...


Such as?
 
#12,649
See the work I cited to you previously by Walt Jung and Jan Didden for examples of reliable circuits with superb noise performance and much better regulation.

It is not good engineering practice to violate datasheet guidelines and hope that unspecified parameters will be constant from device and batch to batch.
 
#12,652
SY said:
See the work I cited to you previously by Walt Jung and Jan Didden for examples of reliable circuits with superb noise performance and much better regulation.

It is not good engineering practice to violate datasheet guidelines and hope that unspecified parameters will be constant from device and batch to batch.
Click to expand...


What you call "not good engineering practice" is a circuit that works very well.
Your view is based on partial knowledge, while this circuit was designed with a knowledge which is more comprehensive than only reading the data sheet. Read again ESS application notes, they relate to this issue.

In 2 high-tech companies I worked for, often we encountered phenomena that at first didn't make sense, they looked like opposing common knowledge in electronics (to qualified and experienced electronics engineers). Of course, electronics isn't mysticism – what turned out eventually was that some points (concerning electronics, or PCB payout, or EMI shielding) were overlooked at first glance. From experience I know that though the information in data sheets is essential, not always it is enough. Data sheets alone aren't enough in designing electronic circuits.
 
#12,653
janneman said:
Please forgive the OT post, but I guess several here would be interested to read this article: Top-10 analog designers. Lots of well known names, and a few obscure (for me).

http://www.planetanalog.com/showArticle.jhtml?articleID=212701512&cid=NL_planet

Highly subjective, of course 😉

(Scott, Barrie is at # 4...)

Jan Didden
Click to expand...


Thank you, Jan.
Obviously, not all engineers have the same capabilities, like not all doctors, all shoemakers or car mechanics.
 
#12,658
Status
Not open for further replies.
Top Bottom