John Curl's Blowtorch preamplifier

Status
Not open for further replies.
Jump to Latest
#10,681
john curl said:
For the record, I have and still design products that operate at 0.4nV/rt Hz across most of the audio band for the last 35 years. For SOME reason, they do not pick up hum or significant noise, and they operate single ended. You are making a 'tempest in a teapot', or making a problem where there is little or none. Also for the record, my phono stages have a bass gain of 80dB, which is a lot more than most here have bothered with in their designs.
Click to expand...

For the record, it would be a trivial problem to put a setup together that will make your equipment hum. Or pick up noise.

Hum issues plague now, or have in the past, 100% of the people on this planet who have used single ended equipment that is connected to the AC supply.

I do not believe for a moment that 100% of you customers have never had a ground hum with any piece of your equipment. So don't bother saying "it aint so..", cause it is..

Face it, you have little if any control of how your equipment is used out in the field...you cannot design for all possibilities, even silly ones. (seems to me I recall you saying that once)

Look at your blowtorch design..individual supplies for every gain stage... Wouldn't be necessary if you could eliminate all the current path issues..

You have done what you felt necessary to achieve your goals, there is nothing wrong with that..you have worked around a problem. Simple, and you should be lauded for that..which I have done now, as well as in the past..

I also know that if all ground/shield/coupling issues were resolved by techniques which were common knowledge, your advantage as a designer would be reduced to just the circuits..so I understand why you'd balk at any discussion which levels the playing field..

Nice try John, but I'm wise to ya..😀

Ya gotta get up earlier in the morning dude........

Cheers, John

ps...the alternative is of course, that you are condemning a discussion of low noise grounding/shielding techniques without examining them....who was it that said that was "prejudice?? Oh yah, your tagline...)😉
 
#10,683
john curl said:
This is nonsense, Jneutron. Please quit trying to 'bait' me into a confrontation.
Click to expand...
It's not baiting..and it's not confrontational.

Never has been dude..

It's about low noise techniques.

I speak of them, you call them nonsense. Perhaps you should speak to Tom Van Doren, explain why it's nonsense to him.

He garners buckets of money teaching this stuff.

He teaches the biggies....people and companies that wouldn't give you the time of day or even think twice about your skills (mine either).

So discuss, John...I'm ready..

Cheers, John
 
#10,684
You should be thanking me dude.

In one fell swoop, I have shown with classical circuit analysis, that an IC can affect the sound, the PC can affect the sound, the connection integrity for IC's and PC's can affect the sound...EVERYTHING WHICH YOU HAVE BEEN CASTIGATED FOR AS BEING IN YOUR MIND.

And yet, you try to quash a discussion on a topic which SUPPORTS your assertions..

All this because you don't like me..:bawling:

Ya gotta get over it John..

Cheers, John
 
#10,688
jneutron said:


Thanks. I don't know when I'd get a chance to watch the whole kittncaboodle..Bill's a great guy. From what I've seen of his work, he's definitely ahead of the curve..

I loved his adaptation using a 12 volt transformer/rectifier to drive ground current to find sensitivities.. I didn't get the sense that he was looking for inductive coupling however, just IR sensitivity of the chassis/ground star geometry.

But he's definitely worth it for anybody interested...I recall he had some papers available online, wasn't it Jensen?

Cheers, John
Click to expand...

I believe Bill Whitlock was writing a book on the complete subject
and it should be available by now.

T
 
#10,690
Terry Demol said:


I believe Bill Whitlock was writing a book on the complete subject
and it should be available by now.

T
Click to expand...
That would be excellent. There's about 30 or 40 high energy physicists and a coupla hundred engineers here that I'd absolutely tell to buy a book like that. I'll get a copy, then perhaps convince the powers that be to arrange a bulk purchase..

When Tom Van Doren came here, they could only have 35 or 40 guys in the class.. Many of the guys who needed the larnin dinna get it...

Cheers, John
 
#10,692
jneutron said:
I loved his adaptation using a 12 volt transformer/rectifier to drive ground current to find sensitivities.. I didn't get the sense that he was looking for inductive coupling however, just IR sensitivity of the chassis/ground star geometry.
Click to expand...

This is more important than is realized. I don't think inductive coupling is an issue in a preamp or MC amp design. This was the thought behind my question a few days ago.

Regards
 
#10,696
john curl said:
Grounding is NOT why we have multiple supplies. Xtalk, and line isolation are the reasons.
Click to expand...

Um, John?

I'll let ya in on a little secret..

What is being discussed is the primary reason for crosstalk in low impedance circuitry...that is why line isolation is important.

Once you understand conceptually what I (and Whitlock and Van Doren) are speaking about, you'll see why it's important to understand the fundamentals.

You have to control where the currents are going. As I stated previously, you have used multiple supplies to control the problem....so you have simply used an alternative solution...good for you.

I attack the fundamental reasons. Those fundamentals will not change because you say they are "unimportant"..



Cheers, John
 
#10,697
As far as the CTC Blowtorch is concerned: We use up to four separate power supplies, but for a line amp only product, we use only two. This is for left and right channel isolation. I found when developing the JC80, over 25 years ago, that total isolation of the power supplies between the left and right channels was important. I tried to use a single remote AC-DC source with separate wiring to individual channels, but when it was measured in detail by a famous Japanese Audio magazine, I had 'some' residual low frequency x-talk. This embarrassed both me and my Japanese importer. Therefore, since then, I have used separate power transformers for separate channels in my best designs. You never know when you might get evaluated by a techno-tweek. The reason for four total power supplies only came up a few times, including my own personal unit. The two other, almost identical, power supplies were added in order to power the separate, AND higher current, Vendetta phono stage. The original power supplies, even using the largest high isolation transformer that I could get conveniently, could not survive being loaded by the added phono stage, if I would have tried to put them in parallel with the line amps.
Now this entails significant complexity:
The number of wires that I initially 'might' have used would be 3, from the power supply box to the signal box. This was done by the Mark Levinson JC-2, the Threshold --10 preamp, etc.
In the case of the CTC Blowtorch, the minimum was 6, and with phono stage, at least 12. This is a lot of extra work, that we would rather prefer to avoid, IF we were not concerned about a shared ground through the connecting cord.
I first noted this problem with the Levinson JC-2. There, we just let a single +/- power supply do everything, including phono. Real problems with X-talk, one of the fundamental weaknesses in the Levinson JC-2.
I came to think of this as if you had two antagonists, and you wanted to separate them from each other. It is easier to use separate rooms, rather than a referee in the same room to keep them apart. This is why I use separate power supplies.
 
#10,698
MikeBettinger said:
This is more important than is realized. I don't think inductive coupling is an issue in a preamp or MC amp design. This was the thought behind my question a few days ago.

Regards
Click to expand...

My experience has been different.

Coming out of my ttables is a pair of rca's, a ground wire, and a 2 prong line cord. One day I decided to put the table pair into a wood coffin, built the pre/mixer/headphone amp into a 2 by 3 by 7 case, and spaced the ttables 4 inches apart, mixer in middle. (btw, it one repeats this, never, ever, wear a cuffed shirt and spin without rolling up the sleeves...the cuffs can grap the tonearm..it sounds like lightning hit the club).

This design forced me to stuff the line cords in the leftover space in the coffin, so the shielded cables were not just intimate to the line cords, they were in the same "ball". Of course, hum was present.

Back in the day, I thought to put a braid shield over the phono's, but it did nothing as the hum was not E-field generated. It was magfield induced....inductive coupling.

By tightly wrapping the phono ground around the rca pair, I eliminated the hum. Inductive hum.

Within a phono amp or mike pre, effort must be taken to reduce induction. Our tools are: eliminate the stray currents, distance them to reduce proximity effects, or use strategies to make the circuits less susceptible to them.

Bill Whitlock's 12 v tranny/rectifier is a tool to find sensitivity. It is deemed an IR measure, but with low impedance circuits, inductance becomes a larger issue. I had the prob with a 250 microohm resistor, it gave almost a volt error at 20 Khz and 1 ampere..Just find a 4 ohm load resistor that has no current slew rate error...you won't find any..

Cheers, John
 
#10,699
john curl said:
As far as the CTC Blowtorch is concerned: We use up to four separate power supplies, but for a line amp only product, we use only two. This is for left and right channel isolation. I found when developing the JC80, over 25 years ago, that total isolation of the power supplies between the left and right channels was important. I tried to use a single remote AC-DC source with separate wiring to individual channels, but when it was measured in detail by a famous Japanese Audio magazine, I had 'some' residual low frequency x-talk. This embarrassed both me and my Japanese importer. Therefore, since then, I have used separate power transformers for separate channels in my best designs. You never know when you might get evaluated by a techno-tweek. The reason for four total power supplies only came up a few times, including my own personal unit. The two other, almost identical, power supplies were added in order to power the separate, AND higher current, Vendetta phono stage. The original power supplies, even using the largest high isolation transformer that I could get conveniently, could not survive being loaded by the added phono stage, if I would have tried to put them in parallel with the line amps.
Now this entails significant complexity:
The number of wires that I initially 'might' have used would be 3, from the power supply box to the signal box. This was done by the Mark Levinson JC-2, the Threshold --10 preamp, etc.
In the case of the CTC Blowtorch, the minimum was 6, and with phono stage, at least 12. This is a lot of extra work, that we would rather prefer to avoid, IF we were not concerned about a shared ground through the connecting cord.
I first noted this problem with the Levinson JC-2. There, we just let a single +/- power supply do everything, including phono. Real problems with X-talk, one of the fundamental weaknesses in the Levinson JC-2.
I came to think of this as if you had two antagonists, and you wanted to separate them from each other. It is easier to use separate rooms, rather than a referee in the same room to keep them apart. This is why I use separate power supplies.
Click to expand...

Excellent. Thanks for the description. As I said, you solved a problem using the tools at hand..nothing wrong with that.

Me, I'd work on the transmission path..you worked on the receiver. Same difference actually...

Look at a power amp for example..most designers just run the rails willy nilly. I do not.. I will run triax or stripline from the supply to the output stage, and make all efforts to have the output current completely cancel fieldwise within the chassis.. It gets difficult at the pass devices, as it requires either a coaxial ground return path, or running the rail strip directly over the output devices.

I don't need to explain how a 10 amp 5 Khz sine current on a #12 wire that is not near the return current will generate magfield, nor that the amp topology on the input side of the feedback divider is very sensitive to the inductive coupling. Most people don't even understand the merits of a feedthrough capacitor over a standard one..

My preference is to create a chassis which has no stray e/m fields produced. That way, the extraordinary lengths you use are not required. (or at least become less important).

I also deal with the chassis-external parts in the same fashion.

Cheers, John
 
#10,700
john curl said:
As far as the CTC Blowtorch is concerned: We use up to four separate power supplies, but for a line amp only product, we use only two. This is for left and right channel isolation.
Click to expand...

John,

Thanks for disclosing this. While reverse engineering your CTC BT I've noticed this topology and I was really dissapointed; you could get the same results (and even better!) by using a single +/- very low noise (and output impedance) power supply (like the Jung super regulator) and make sure you use a drive/sense topology, shielding the sense wiring and a symmetrical layout of the gain stages.

This would still require extra wiring but would certainly lower the overall complexity. Of course, this would mean you should use opamps in the power supply 😀

Using such a topology I got in my latest phono stage 75dB unweighted S/N @ 0.5mV, 1KHz input, 64dB gain, and almost -100dB crosstalk. The power supply is in a separate housing and has by itself about 7nV/rtHz noise up to 10MHz. This is of course further killed by cascoding the 4 x JFET input stage, locally filtering the supply, and the opamps PSRR.
 
Status
Not open for further replies.
Top Bottom