John Curl's Blowtorch preamplifier part II

[Ken Newton]

Is anyone ready for a real schematic? This is the Para JC-2. Want to build it? Be my guest:

John,

I suppose, the first thing that draws ones attention in this schematic is the use of an IC op-amp based buffered feedback loop in conjunction with an 16-JFET + 4-MOSFET discrete gain stage. I can imagine a few reasons why this might have been done, but I wonder if you might provide your reasons?

 

[aeronautical]

It doesn't hurt to leaven the audio talk with a bit of humor from time to time, even if it means another post to read.

I agree, I wasn't being cruel.
I like funny too.

 

[john curl]

Ken, the IC's are SERVOS and work below the audio bandwidth.

 

[aeronautical]

That's not specs, that's a marketing speak. A spec is like: I want my preamp to have a S/N of -110dB referred to 2V signal.
You shouldn't care less whether the designer does that with internal compartments or a shoot of Broccoli.

jan didden

Broccoli...I haven't tried that...but if might add coloration to the sound...green. da dum ching, Ok, that wasn't funny.

Just kidding you.

Seriously though, I will get a set of specs together but I believe they shouldn't necessarily be steaks in the ground. This is because with that discipline, one could freeze themself out of potential "good things" by being unwilling to bend on a set of parameters that might be at least slightly arbitrary.

So, I want to remain flexible.

Personally, I believe the internal compartments deserve at least some consideration for noise considerations. Let me try to explain this to a fellow aviator...when Mooney Aircraft built their plane they didn't have the high power engineering departments that did say Beechcraft and Cessna. (civillian market stuff). This is their explanation by the way. So, when they built their wing they built is so strong that there was no way it would fail under any concieveable use....except maybe as a bridge truss. They were using what I call the "valley girl" approach..."fer sure fer sure". I appologize if you aren't familiar with the reference.

Anyway, it is intended just as an ounce of prevention, this internal shielding. I was actually somewhat fishing to see what others thought.

I am here to learn so if I act oblivious to something I should know, well, it is because I don't.

 

[Wavebourn]

Anatoliy, I am a bit surprised. The fact is that there are clearly audible differences in sound of preamplifiers that have distortion even below 0.001%. The difference of preamp's sound is similar to the difference between power amplifiers. Regardless the fact there were 20 opamps in the recording chain during recording session. Please try to make thorough listening tests and you should come to the similar conclusion.

Pavel;

you know I meant properly designed preamp. The one that measures "thd below 0.001%" on 1 kHz with source 50 Ohm and 1 mOhm load not necessary is properly designed. The real crap can have such low THD on steady measurements in such laboratory conditions.

Speaking of opamps and consoles, there is big difference between unity gain buffer and an amp on a single opamp that requires at least 40 dB of voltage gain. For 40 dB of gain I may prefer to use at least couple of opamps than one, to get better results, if I was forced to stick with opamps. And you know that voltage gain is not what matters; what matters is power gain.

 

[aeronautical]

What preamp and who are "we"?

we...why you and the rest of us of course....so cowboy up buddy.

 

[Pano]

This isn't a preamp build thread. That should have been obvious twenty thousand posts ago. It's a discussion of audio design, more or less.

Build threads are great, this just is not one of them.

 

[Chris Hornbeck]

The Parasound JC-2 looks way too complex for most homebrewers, but the details are still very interesting. Ferinstance, it looks like the first stage idles at about 1/3 the current of the second stage. I wonder if different color Idss groups were needed. Stuff like that, like the best of this thread, gets at the thought behind the piece itself.

Thanks,
Chris

 

[aeronautical]

One topic that you bring up is partition of the various functions and channels. This is VERY IMPORTANT and not as easy as many engineers think.
My first experience was when I visited a competitor who made a much cheaper preamp than the JC-2. Playing a familiar record, and the same speakers I was then using, this competitor's preamp put CAT STEVENS right between the two speakers. MY set-up could NOT do that!. So, what is the problem? I found it to be asymmetrical X-talk between the left and the right channels. Now, was I alone in finding this?
No, at a hi fi show, I heard a glorious violin sound coming from one of the rooms. I said to myself, that is a REAL violin, you can't fool me, I married a violinist!
Well, I was fooled, by my own JC-2 preamp and Dick Sequerra's ribbon tweeters. However, taking to Dick about the JC-2, he independently stated that it was the best MONO preamp available, denoting that it had 'imaging' problems. Trust me, the JC-80 was deliberately designed not to have this problem, and the CTC Blowtorch, as well.
SO, partitioning is GOOD! Separation by air is good too. It must be remembered that most SPECS. are derived from IDEAL conditions, Zero drive impedance, input pot all the way down, etc. NOT real world working conditions.

I kinda got beat up for even mentioning it. But when I was about 22 I was building a project and a guy at the local hi fi shop said that even if I went in and did some just indescriminate shielding it would make it quieter. I didn't do it but it got me to thinking about all the magnetic fields interacting in our boxes. If you ever saw an aviation radar transciever, everything is in its own tiny "box"

 

[john curl]

Gee, Chris, maybe that is why I didn't show it before!

 

[john curl]

Chris, for the record, different jfet selections are really not necessary, IN THIS SCHEMATIC. However, in my Constellation schematics, I use different MANUFACTURERS of the jfets.

 

[john curl]

Be careful of mumetal, however, or even steel, as a shield.

 

[john curl]

Aluminum only, or non-magnetic stainless steel.

 

[morinix]

This isn't a preamp build thread. That should have been obvious twenty thousand posts ago. It's a discussion of audio design, more or less.

Build threads are great, this just is not one of them.

That is what it is now but it started as a simple request for a schematic to add to a library.

 

[Ken Newton]

Ken, the IC's are SERVOS and work below the audio bandwidth.

Duh, stupid me. I somehow overlooked the obvious integrating caps. in those op-amp's feedback loop.

 

[Joshua_G]

Aluminum only, or non-magnetic stainless steel.

How about copper, or silver plated copper?

 

[bcarso]

How about copper, or silver plated copper?

Cu is great but difficult to machine and very expensive thus.

 

[aeronautical]

Aluminum only, or non-magnetic stainless steel.

Mr. Curl, is that because steel or ferric metal can be excited into being a magnet in the presence of a magnetic field...thereby fixing one problem and spawning others?

 

[fas42]

That's a nice marketing brief, but my question was "specify exactly in electrical terms what it has to do and not do." You need an electrical target for designing electrical equipment. If you're more interested in the marketing gloss, that's beyond my abilities.

I'll give you a nice one: a complete system, from source to speakers is recorded via top notch mic and ADC playing a variety of musical, highly testing excerpts at a realistic sound level. Then that system is subject to a variety of severe mains interference, RF interference stimula, in various combinations and strengths, vibration of some of the components, including cables, by controlled engineering devices, you get the idea. All variations are recorded, and then Diffmaker or whatever is run against the results. We'll be looking for variations of no more than 80dB below peak under all circumstances ...

Frank

 

[SY]

Aeronautical and I were talking about preamp design, which is a simple matter (again, assuming evaluation with ears alone) once the electrical target is specified.

I'll leave the complex and speculative experiments covering other parts of the signal chain to others. I'll be interested in the results if they can show a correlation with ears-only listening tests.