Natural Sound

[DasguteOhr]

If you have monoblocks that becomes really inconvenient.

If you can drive long cables then you can place the preamp with the front end of the system in a location where airborne sound won't affect it as much.

So does its output use a transformer or a coupling capacitor?

Transformers (and autoformers) introduce distortion via hysteresis. This issue causes any tube amplifier using one at its output to have an elliptical load line. This causes the plate Voltage and current to be out of phase which increases distortion of the tube (in addition to that of the transformer itself). If there is a transformer (or autoformer) used to change signal level, it too has hysteresis and so will make distortion. This is unavoidable even if the transformer is properly loaded; its a function of transformer operation. All transformers have cutoff frequencies both low and high which introduce phase shift (which is why a resistive volume control is more transparent and wider bandwidth).

Coupling capacitors don't have the hysteresis problem, but they cause the output impedance of the circuit to vary with frequency as they will have a low frequency cutoff (which introduces phase shift; in the bass region the ear interprets this as a lack of impact despite the cutoff going lower than the ear does). At the output of a preamp, this will cause the 20Hz output impedance to be quite a bit higher than it is at 1kHz. This effects bass impact and varies with the load of the amplifier input impedance its driving. This is one reason amplifier input impedance in home audio is quite high, such as 100kOhms so to try to get around this issue.

One might think these issues are minor (perhaps only 'measurable') but IME they are quite audible. The additional transparency and bandwidth gained by direct-coupling at the output of a preamp isn't trivial.

Balanced cathode follower two output coupling caps per chanel, really good parts inthere( amrg rsistors ,milflex caps.)
They've increased the capacitors somewhat over the course of the model series, judging by the pictures. Power supply: B+ 280VDC. Possible operating point: ~ 150V, 7-10mA. I haven't measured it.
Exsample pic

 

[Atmasphere]

Balanced cathode follower two output coupling caps per chanel, really good parts inthere( amrg rsistors ,milflex caps.)
They've increased the capacitors somewhat over the course of the model series, judging by the pictures. Power supply: B+ 280VDC. Possible operating point: ~ 150V, 7-10mA. I haven't measured it.
Exsample pic
View attachment 159849

That approach yields a balanced circuit, but because the outputs reference ground, AES48 is not supported.
In a proper balanced line system where AES48 (one of the balanced line standards) is supported, ground is ignored. This is one of the techniques used to minimize artifacts from the interconnect cable (the 'sound' of the cable) and is also done to prevent ground loops.

 

[microstrip]

That approach yields a balanced circuit, but because the outputs reference ground, AES48 is not supported.
In a proper balanced line system where AES48 (one of the balanced line standards) is supported, ground is ignored. This is one of the techniques used to minimize artifacts from the interconnect cable (the 'sound' of the cable) and is also done to prevent ground loops.

I can understand the need for floating the inputs, but what is the technical issue with outputs referenced to ground?

 

[Atmasphere]

I can understand the need for floating the inputs, but what is the technical issue with outputs referenced to ground?

The first is more obvious: ground loops. They can color the sound a bit even if buzz doesn't show up since intermodulations can occur in the ground due to non-linearities scattered around the system.

The second issue is the geometry and capacitance of the cable. As you know, capacitors are made of different materials for a reason and some sound better than others (due to measurable performance). If you eliminate or reduce that as a variable the interconnect cable will have less effect on the sound.

The other bit is the ability of whatever is driving the balanced line to be able to make a bit of power into the load which its driving. This has the effect of swamping capacitance and noise from high impedance sources (such as an induced magnetic field). For example my Neumann U67s can drive 150 Ohms which allows them to drive extremely long cables without any coloration at all, regardless of the kind of cable used.

 

[Kingrex]

Why can't manufacturers include this amplifier function in the power amplifier? Is it an industry conspiracy to force us to spend more money?

Isn't that what we call an integrated amp.
I get how technically its not. But if you went that far, just add the resistive ladder or pot.

 

[microstrip]

The first is more obvious: ground loops. They can color the sound a bit even if buzz doesn't show up since intermodulations can occur in the ground due to non-linearities scattered around the system.

I am probably missing something - even if an output is referenced to ground we are just sending the two phases to the receiver. How does it create a signal ground loop?

The second issue is the geometry and capacitance of the cable. As you know, capacitors are made of different materials for a reason and some sound better than others (due to measurable performance).

Ok, but unfortunately there is still a lot we can't understand about cables ...

If you eliminate or reduce that as a variable the interconnect cable will have less effect on the sound.

If the source is active it is never floating in the sense of being isolated. IMO the variable stays.

The other bit is the ability of whatever is driving the balanced line to be able to make a bit of power into the load which its driving. This has the effect of swamping capacitance and noise from high impedance sources (such as an induced magnetic field). For example my Neumann U67s can drive 150 Ohms which allows them to drive extremely long cables without any coloration at all, regardless of the kind of cable used.

Ok, I was just considering low output impedance active sources.

 

[Atmasphere]

I am probably missing something - even if an output is referenced to ground we are just sending the two phases to the receiver. How does it create a signal ground loop?

When grounds of two different circuits are connected together the signals can mingle. Quite often in bad cases this results in a buzz. The issue in a grounded balanced system is now you have audio ground currents in the shield of the interconnect, something you don't get if the device supports AES48.

Ok, but unfortunately there is still a lot we can't understand about cables ...

I'm not sure that's really true. I do get the feeling a lot of cable manufacturers like to promote that idea though.

Nor do I think it matters; if the balanced system is set up properly you can't hear the difference between cables of various types.

Anecdotally, about 10 years ago I did a demo of our preamp compared to another balanced preamp which was using interconnect cables that at that time cost about $1000/foot (and he had 24 feet of this stuff). He mentioned that the cables might be a difference and he didn't want to be swapping them since he was 70 and didn't want to be crawling about. I told him don't worry about it- if our preamp didn't do the job don't buy it. He called the next day and bought the preamp and the Mogami cables. IIRC he sold his old interconnect for more than he paid us for the preamp.

If the source is active it is never floating in the sense of being isolated. IMO the variable stays.

Not sure what you're trying to say here. If I do understand then it may help to know my Neumanns are active and floating. There's no connection at their output to ground; the ground connection is the case of the microphone. This is accomplished by a transformer, which in the old days when tubes were king was how it was done when there was a balanced line.

Ok, I was just considering low output impedance active sources.

A lot of high end audio preamps, even tube ones are fairly low output impedance at 1KHz. But usually their output impedance is much higher at 20Hz so while they can play a certain level at 1KHz into 600 or 1000 Ohms, they can't do anywhere near that at 20Hz. This is especially true of tube preamps and is why transformers were usually used between mics, tape machines, mixers, amps and so on during the tube era.

 

[microstrip]

When grounds of two different circuits are connected together the signals can mingle. Quite often in bad cases this results in a buzz. The issue in a grounded balanced system is now you have audio ground currents in the shield of the interconnect, something you don't get if the device supports AES48.


I'm not sure that's really true. I do get the feeling a lot of cable manufacturers like to promote that idea though.

Nor do I think it matters; if the balanced system is set up properly you can't hear the difference between cables of various types.

Anecdotally, about 10 years ago I did a demo of our preamp compared to another balanced preamp which was using interconnect cables that at that time cost about $1000/foot (and he had 24 feet of this stuff). He mentioned that the cables might be a difference and he didn't want to be swapping them since he was 70 and didn't want to be crawling about. I told him don't worry about it- if our preamp didn't do the job don't buy it. He called the next day and bought the preamp and the Mogami cables. IIRC he sold his old interconnect for more than he paid us for the preamp.

Not sure what you're trying to say here. If I do understand then it may help to know my Neumanns are active and floating. There's no connection at their output to ground; the ground connection is the case of the microphone. This is accomplished by a transformer, which in the old days when tubes were king was how it was done when there was a balanced line.


A lot of high end audio preamps, even tube ones are fairly low output impedance at 1KHz. But usually their output impedance is much higher at 20Hz so while they can play a certain level at 1KHz into 600 or 1000 Ohms, they can't do anywhere near that at 20Hz. This is especially true of tube preamps and is why transformers were usually used between mics, tape machines, mixers, amps and so on during the tube era.

Ok, thanks. I see you are mainly addressing the potential problems with improperly designed gear or old gear in complex systems - not my typical concern.

I usually check for ground loops or noise issues with an audio spectrometer, much more sensitive than ears. Never had any detectable issue with the usual balanced outputs referred to ground, as used by most modern balanced gear, as long as the XLR cables were properly wired.

 

[Atmasphere]

Ok, thanks. I see you are mainly addressing the potential problems with improperly designed gear or old gear in complex systems - not my typical concern.

I usually check for ground loops or noise issues with an audio spectrometer, much more sensitive than ears. Never had any detectable issue with the usual balanced outputs referred to ground, as used by most modern balanced gear, as long as the XLR cables were properly wired.

IME you can get lower noise if the ground is ignored. In some systems its not a big deal and in others it can be enormous.

 

[microstrip]

That approach yields a balanced circuit, but because the outputs reference ground, AES48 is not supported.

It seems this sentence is the source of confusion. What is mean't exactly by "outputs reference ground" ? Are you addressing the non existence of galvanic insulation?

 

[DasguteOhr]

That approach yields a balanced circuit, but because the outputs reference ground, AES48 is not supported.
In a proper balanced line system where AES48 (one of the balanced line standards) is supported, ground is ignored. This is one of the techniques used to minimize artifacts from the interconnect cable (the 'sound' of the cable) and is also done to prevent ground loops.

I can only say that the passive preamp + buffer performed very well for me. There's little hum/noise .there's always room for improvement.
I used ecc82 telefunken smooth plate for me the best tube overall, totally harmonious sound.

 

[Atmasphere]

It seems this sentence is the source of confusion. What is mean't exactly by "outputs reference ground" ? Are you addressing the non existence of galvanic insulation?

No. A balanced output should not reference ground. A transformer is not needed to do that. When a ground is involved in the situation, it opens up the circuit or system to ground loops. Imagine in a recording studio you might have 20 or 30 devices that are used: mic preamps, mixer, amplifier, compressors, EQ units and other special effects. They will all have their grounds tied together. You can imagine it might take weeks to find that one device that is generating a ground loop buzz because the manufacturer referenced the signal outputs to ground. That is part of why AES48 exists.

The inverting and non-inverting outputs are not supposed to generate their output Voltages with respect to ground; instead generating that Voltage with respect to each other.

An output transformer used in this application will have simple secondary output with only two wires. One wire is tied to the non-inverting output (usually pin 2 in the US) and the other wire to pin 3 (inverting in the US). There is no center tap as that would degrade the Common Mode Rejection Ratio; instead pin 1 of the XLR is simply tied to chassis or the audio ground.

There are other circuits that allow this connection but they don't involve coupling capacitors. One of those is a solid state circuit that allows for a floating output that can sink current into either output without damage. There are chip manufacturers that off such products, for example That Corporation. The link shows their balanced line drive IC.

There is a third technique you've seen before that involves direct-coupling which can be executed with tubes or transistors. As far as I know of, these three are the only ways to generate a balanced output that is AES48 compliant.

It is notable that nearly all phono cartridges (such as Peter's recent acquisition) are balanced sources. In a good tone arm their outputs are not tied to the tone arm ground. That is why when a cartridge is run 'single-ended' there will be a ground wire that other single-ended sources don't seem to need. You can see that the cartridge is floating with respect to ground. The ground wire is thus the shield that is the tone arm tube itself.

I can only say that the passive preamp + buffer performed very well for me. There's little hum/noise .there's always room for improvement.
I used ecc82 telefunken smooth plate for me the best tube overall, totally harmonious sound.

The smooth plate is a nice tube! So is that an option over the 6SN7?

 

[microstrip]

No. A balanced output should not reference ground.

Again, common active devices generate a voltage signal that references ground.

A transformer is not needed to do that. When a ground is involved in the situation, it opens up the circuit or system to ground loops. Imagine in a recording studio you might have 20 or 30 devices that are used: mic preamps, mixer, amplifier, compressors, EQ units and other special effects. They will all have their grounds tied together. You can imagine it might take weeks to find that one device that is generating a ground loop buzz because the manufacturer referenced the signal outputs to ground. That is part of why AES48 exists.

Yes, we know the aim of the standard.

The inverting and non-inverting outputs are not supposed to generate their output Voltages with respect to ground; instead generating that Voltage with respect to each other.

Are not supposed but are not forbidden, as far as I know. The signal is the difference between them.

An output transformer used in this application will have simple secondary output with only two wires. One wire is tied to the non-inverting output (usually pin 2 in the US) and the other wire to pin 3 (inverting in the US). There is no center tap as that would degrade the Common Mode Rejection Ratio; instead pin 1 of the XLR is simply tied to chassis or the audio ground.

Ok, known since long - the pin 1 rule.

There are other circuits that allow this connection but they don't involve coupling capacitors. One of those is a solid state circuit that allows for a floating output that can sink current into either output without damage. There are chip manufacturers that off such products, for example That Corporation. The link shows their balanced line drive IC.

Ok, a complex IC to generate balanced floating outputs.

There is a third technique you've seen before that involves direct-coupling which can be executed with tubes or transistors. As far as I know of, these three are the only ways to generate a balanced output that is AES48 compliant.

As far as I know the AES48 standard does not address floating outputs. Can you point me the section you are referring?

AES48 refers that output must not use pin 1 for its reference or return and warns against using transformers having a central pin connected to ground. If the input stage is a transformer properly connected the rule is obeyed.

BTW IMO we must consider that allowed grounding safety rules are different for studios and domestic installations - we can't have a technical earth bus at home!

 

[PeterA]

I thought I would bring the thread back on the topic of my system.

I remember well when I had the Micro Seiko SX8000 II turntable, the Lamm LP2.1 Deluxe phono stage, and the vdH Colibri cartridge. It was a great combination, and I was quite happy with the sound of the system. However, after hearing some of my own LPs on David Karmeli's system for a second time and meeting Tima in Utah, I concluded that for me, David's American Sound AS2000, his Lamm LP1 Signature phono stage, and his Neumann DST cartridge were the "Holy Trinity" of vinyl playback. It was the best I had heard, and it became my new target. Now, about two years later, I have completed, step by step, my quest to get those three components in my system. Alas, I could not find an original Neumann DST, but this replica was released, and from what I remember, it sounds pretty close to the original. Perhaps this winter I will be able to do a direct comparison with David and Tim.

Some of my previous recent videos were with my Lamm L1 preamp as I wanted to really learn how it differs from the LL1. This is with the LL1.

Here is a new video:

 

[tima]

Nice to have a post tied to the thread.

 

[DasguteOhr]

The smooth plate is a nice tube! So is that an option over the 6SN7?

Of course ,other types of tfk ribbed or 802s don't have the rich tone or the magical midrange. They are precise cleaner more neutral.it plays out finer details than 6sn7 that I had here.
1. Smooth plate ecc82 Tfk
2. Siemens & Halske ecc82 early years 1955
3. Ecc 802s Tfk
4. Valvo ecc82/
My personal ranking

 

[PeterA]

Nice to have a post tied to the thread.

I agree Tim. Here is another short video: female voice and double bass. For @bonzo75 , the video is representative of what I hear in the room, especially the sense of clarity, energy, and presence as exemplified by this new DST Replica cartridge.

 

[pcosta]

Peter
Are you using the internal step-up transformers the Lamm has?
Have you tried any others? Such as the BV33 that ing. J.P. (Hans) van Vliet has?

 

[PeterA]

Peter
Are you using the internal step-up transformers the Lamm has?
Have you tried any others? Such as the BV33 that ing. J.P. (Hans) van Vliet has?

Yes, I’m using the internal SUTs. I have not tried others. I have one moving magnet input, which I am using for my vintage Ortofon with its own SUT.

 

[tima]

Here is another short video: female voice and double bass. For @bonzo75 , the video is representative of what I hear in the room, especially the sense of clarity, energy, and presence as exemplified by this new DST Replica cartridge.

I'll say here what I told you last night:

"Great Zot Peter, that video sounds fantastic! Best I've heard from your system, particularly the bass and sense of context. All your work with the DST is paying off."