John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
Let me try again:
 

Attachments

  • 4558.jpg
    4558.jpg
    349.6 KB · Views: 162
  • barrie41.gif
    barrie41.gif
    6 KB · Views: 159
Here's the pic.
No ground at all is better, isn't ?

What is have done: replace my preamp AC wire with a big AC one. I use-it as the unique ac ground connection to the outlet. Tried the two senses of AC wires to get the less voltage betwwen ground wire and ground outlet before connecting grounds. Then connect my amp with no ground connection, find the best sens between the two chassis for minimum leakage voltage too. My signal wires does not connect ground, i have a big wire between the 2 chassis, following my signal cables. Then all parts connected to the preamp are tested for the sens in the same manner, and no ground connected to outlets.
All my signal wires shields have their ground connected only at the source part, with a ferrite near the plugs. The wires are on pair, and the cold wire connect the grounds.
 
Mostly, it's the recording and mastering; the playback gear has much less to do with it. With the few really good recordings out there, you get exactly what you're supposed to and the sound can be terrific. With other ones... well... you get the sound the producers intended with no low resolution media to plaster over the cracks. That may not be to your liking- it's certainly not to mine.

It is definitely not only the recording and mastering.
Indeed, the majority of the recordings from the last few decades are crappy, however, good recording and mastering sound crappy on most digital gear. The digital gear that sound really good (with good recording and mastering) is rare.
 
Sounds that the designers of the AMR have learnt a thing or two.

Indeed.
The designer of AMR is Thorsten Loesch.

My suspicion is that the power supply of the Oppo is what's causing problems here ...

My assumption that it isn't only the power supply. The clock for the transport matters, as well as interference isolation, or lack of, between various circuits.
 
4558, well that restrained the discussion, didn't it? Scott, are you not prepared to DEFEND the use of the 4558, even with serious audiophiles? '-)
Unfortunately, even Ron Quan's tests still have a fairly high residual, that keeps us from evaluating the 'better' op amps completely, as their FM distortion (if any) lies below the residual of Ron's test equipment. Don't worry, improvements will be made. Maybe, a BIG company will address the issue and give us serious test equipment to measure this distortion. I sure would hope so.

4558 can't use 'em ;), we don't make them anyway. Ron's tests would probably yield to a digital solution with a high quality A/D and enough raw data.

Now address the point that highly regarded class A amplifiers have lots of PIM well above Ron's noise floor, and then go back and say again his test needs more resolution. This is just making no sense.
 
No ground at all is better, isn't ?

What is have done: replace my preamp AC wire with a big AC one. I use-it as the unique ac ground connection to the outlet. Tried the two senses of AC wires to get the less voltage betwwen ground wire and ground outlet before connecting grounds. Then connect my amp with no ground connection, find the best sens between the two chassis for minimum leakage voltage too. My signal wires does not connect ground, i have a big wire between the 2 chassis, following my signal cables. Then all parts connected to the preamp are tested for the sens in the same manner, and no ground connected to outlets.
All my signal wires shields have their ground connected only at the source part, with a ferrite near the plugs. The wires are on pair, and the cold wire connect the grounds.

The intent is to eliminate any bouncing of the reference node the amplifier sees. Your technique is certainly viable, but my concern is that safety.

If your amp line-in hot hits chassis, will the path through the pre be sufficient to clear a 20 amp breaker? If not, the possibility of the entire system going hot becomes non-zero.

I note you put the grounding conductor with the IC's, that is very good. If you put the IC's within the grounding conductor, that would be even better. A cylindrical shell of current has no internal magnetic field. (see pic, it is a cross section of the magnetic field of a cylindrical conductor formed by hundreds of individual wires..I limited it to roughly 1/4 of the cylinder due to processor speed, but it shows what happens quiet cleary). If the IC's were routed within a copper pipe or a length of pex with a braid over it connecting the amp and pre, the IC's would have a path which will not see any ground loop magnetic field, there would be little RF intrusion.

The issue would be one of making sure the ground is reliable for life...both equipment's and humans.

jn
 

Attachments

  • wirearray06axial.jpg
    wirearray06axial.jpg
    28.1 KB · Views: 151
Last edited:
The digital gear that sound really good (with good recording and mastering) is rare.
As a sound engineer, bringing my tapes to the mastering room, i can't find too much differences between my recordings and their digital copies, whatever my original tape, analog or digital. Nothing to be upset off.
Vinyl grooving was just a matter to kill my day. All the dynamic and little details gone away, and all those noises, you don't need any blind test, just you cry !!!
In my last studio, with a lot of digital equipment, mixing desks, recorders, effects, computers etc... as long we stayed in the digital domain, never something to worry about. But we had a VERY good master clock.

In my home system, my unique secret is "RE-CLOCK" for no jitter.
But my system does not play nicely bad recordings, it plays the sources as they are.
 
Last edited:
The intent is to eliminate any bouncing of the reference node the amplifier sees. Your technique is certainly viable, but my concern is that safety.The issue would be one of making sure the ground is reliable for life...both equipment's and humans.
You are right. I'm divorced, my sun prefer to listen his headphones, and with so few very good records to buy each year, i don't worry to die :)

Well, just a question, serious for a while. Any interest to use a loudspeaker's line adapted in impedance , while amp source is near 0 Ohm ?
 
Last edited:
Wavebourn, I agree that in mid fi applications, the 4558 is 'adequate' I have been listening through 4558's on an SACD player that I have had for the last 10 years or more. Without DIRECT COMPARISON with better equipment, it is difficult to find any problem. YET, Barrie Gilbert's analysis PREDICTS FM distortion, so we should note it, and perhaps avoid it with our 'best' designs.

John, please see my disclaimer: for higher power gain I would not use them. But for buffers, and stages with gain about 20 dB they are adequate. Of course, I mean biased by 6K8 to the + rail, because as-is they are not acceptable at all for any high end duty.
 
Can someone explain me this sudden interest in this poor poor 4558, that I would not use more than any TL0.72 for audio or even NE5532 ?
Very high slew rates and current feedback make all the difference between a dump system and an other, fast like light, transparent and light like air, on my point of view.

I'm not sure about 072 but 4558 sounds MUCH better than
LF356 (similar to 072).
 
As a sound engineer, bringing my tapes to the mastering room, i can't find too much differences between my recordings and their digital copies, whatever my original tape, analog or digital. Nothing to be upset off.

Most of the people are quite happy with most of the digital gear available today.
Some people, me included, find that most of the digital gear available today sound crappy.

As long as it sounds good to you, all is well. You don't have to spent a lot of money for a really good sounding digital gear.

Vinyl grooving was just a matter to kill my day. All the dynamic and little details gone away, and all those noises, you don't need any blind test, just you cry !!!

My experience is different from yours.
Indeed, of all the vinyl records I have, only few are in a good condition. However, those records that are in good condition sound to me better than the best CDP I've ever heard.

To my view, based on my auditory experience, this is the perfect example of the lack of correlation between how audio gear is measured (by conventional measurements, that is, THD and SNR) and how it sounds.

Indeed, on most of the vinyl records I have there is some surface noise, at times also some cracks. Yet, those surface noise and cracks have lesser impact on the sense of realism of the music than most digital gear I've heard.

Possibly, different people pay greater attention to different aspects of the sound. Possibly, some people mind more how quiet it sounds, while others mind more the sense of realism the music have.

In an ideal world, we would have both – quietness and sense of realism. In our present world, I'm still looking for something that will have both (something which I can afford to purchase). However, when it comes to CDs, my suspicion is that, due to the Red Book Protocol limitations, even the best CDP possible will not have the sense of music realism that vinyl records have.

To my experience, Hi-Rez audio can sound as good as vinyl records. So far, I didn't find any Hi-Rez recording which is good enough for me to pay for it. To my view, the poor quality of recent recordings and masterings can be explained by the fact that so many sound engineers and music producers are happy with most of the CDPs available today. When they aren't aware of the sense of music realism, they will not produce recordings that have such a sense of realism. Another reason is that the music industry, including the recording and mastering processes, is aimed for the masses. It seems that the masses aren't looking for a sense of music realism.
 
You are right. I'm divorced, my sun prefer to listen his headphones, and with so few very good records to buy each year, i don't worry to die :)

Well, just a question, serious for a while. Any interest to use a loudspeaker's line adapted in impedance , while amp source is near 0 Ohm ?

Using a matched impedance speaker line is the best in terms of getting the information to the load at the same time. The debate is and always has been, does it make a noticeable difference?

All transmission lines have a propagation speed, v = c/sqr(LC), c being lightspeed. It is erroneously assumed that the information makes it to the speaker at that prop speed. That is incorrect. The propagation speed of any cable is the speed at which a signal with the proper current to voltage ratio will propagate. If the cable is 100 ohm, only a signal with a voltage 100 times the current (100 volt, 1 amp) will travel at the cable prop velocity. If the load is 10 ohms, the first transit will not produce 10 amperes at the load. It will take much longer for the load current to reach 10 amperes. What confuses everybody is the fact that audio is so much slower than t line propagation that the leading edge delay is lost in the transition speed.

The current rate of rise at the transducer at the far end will have a frequency and load impedance dependence rate. If there is a huge mismatch between line and load, there will be a slower response to a change in the signal. If the line and load match, the response will be exactly delayed by the prop velocity and line length. If horribly mismatched, the delays can be in the sub 10 microsecond domain, localization territory.

If you use load which matches the line, the amp will NOT see any capacitance, no matter what the cable C is, no matter how long it is.

If the load unloads at hf, the amp will see the capacitance. If the amp is hot where the load goes high z, it can oscillate.


This also occurs with IC's as well.
jn
 
If horribly mismatched, the delays can be in the sub 10 microsecond domain, localization territory.

um, 10 us = ~1/8”, 3mm sound prop distance in air

I don't have my chair, speakers that closely aligned – do you? multiple drivers lots further apart on the cabinet – even the tweeter cone/dome has modes on that scale...

only barely possibly relevant if you have > ~10 us mismatch between R,L channels with speakers and rooms - the sub 10 us numbers come from headphone listening
don't we expect speaker identical models speaker terminal impedance to match even a little? , maybe even better than 10%?
 
Status
Not open for further replies.