Opinions on the role of the preamp in a modern single source system

[egidius]

berning?

I might actally might be able to shed some more (High rez ) light on it in the coming weeks , i am in the process of completeing a deal on a very interresting tube power amp which i believe has also a passive vol control , ill keep the cat and drive it with the cat also .
Its 2 times 8 watt , i think i might go see a docter

that has to be a sigfried?? FABULOUS!!!
i always wanted one, but David Berning told me to get the ZH230 instead for my speakers..

egidius

 

[andromedaaudio]

Thanks Zanden model 7000 300B with sophia electric tubes at the moment , no longer produced.
A classic , i never read a review anywhere either

 

[ack]

Could someone with more technical expertise than myself explain what would be required of the source component and the amp to make an appropriate match and rule out this potential problem? Something similar to the way Amir just cleared up impedance matching issues would be perfect, if it's that simple.

Tim

With my comment there, I actually partially disagreed with Amir (as I disagreed with him in similar situations with certain amps being able - or not - to drive tough speakers like MLs, for example). In cases like the one we are discussing here, a source component may simply not be able to deliver adequate current to the amp (if we are dealing with an amp requiring unusual input amperage to be driven properly), and the sound is fairly thin and lacking a bit in dynamic contrast. This is the case with Spectral amps, for example, where even the Berkeley Alpha DAC (which was developed driving them directly) isn't as capable as with their preamps in the middle, which put out up to 2A peak (1A RMS) to the amplifier (whose manual claims 100mA min required)- it's only then that the whole system's dynamic capabilities are at full swing. In those cases, the issue can be the DAC's limited output (from said amp's perspective), and certainly the impedance on either end can be a contributing factor. This all goes back to system matching.

 

[Phelonious Ponk]

With my comment there, I actually partially disagreed with Amir (as I disagreed with him in similar situations with certain amps being able - or not - to drive tough speakers like MLs, for example). In cases like the one we are discussing here, a source component may simply not be able to deliver adequate current to the amp (if we are dealing with an amp requiring unusual input amperage to be driven properly), and the sound is fairly thin and lacking a bit in dynamic contrast. This is the case with Spectral amps, for example, where even the Berkeley Alpha DAC (which was developed driving them directly) isn't as capable as with their preamps in the middle, which put out up to 2A peak (1A RMS) to the amplifier (whose manual claims 100mA min required)- it's only then that the whole system's dynamic capabilities are at full swing. In those cases, the issue can be the DAC's limited output (from said amp's perspective), and certainly the impedance on either end can be a contributing factor. This all goes back to system matching.

OK, then, let's assume I'm putting together a system for which my computer will be the only source. How would I go about choosing a DAC/pre and an amplifier that are a good match? What is the optimum output/input ratio, if that is an applicable way to look at it?

Tim

 

[AudioExplorations]

OK, then, let's assume I'm putting together a system for which my computer will be the only source. How would I go about choosing a DAC/pre and an amplifier that are a good match? What is the optimum output/input ratio, if that is an applicable way to look at it? Tim

There are 3 I know of:

1. The ratio of amp input impedance / source output stage impedance must be 10x or higher,
2. The source must be able to output enough voltage to achieve adequate listening levels. In the case of digital attenuation, components need to be carefully selected to avoid too much output voltage and hence the need for large amounts of attenuation - the output voltage must be matched to the sensitivity of the amplification and speakers to achieve desirable listening levels, needing only slight tweaking with the digital volume control (some DAC's made by very clever designers have adjustable output voltages to lock this in perfectly).
3. The current delivery capability of the source output stage must be sufficiently high as ack pointed out.

Let me know if I am missing anything.

 

[ack]

OK, then, let's assume I'm putting together a system for which my computer will be the only source. How would I go about choosing a DAC/pre and an amplifier that are a good match? What is the optimum output/input ratio, if that is an applicable way to look at it?

Tim

Optimal ratio? There are plenty more technical folks here than me to answer this question, but usually we want high input amp impedance (of which said Spectrals are not) and low output preamp/DAC. I don't go reading amp's manuals all the time, but I assume most list theirs plus required input voltage (if not also current) for rated output power (aka sensitivity); it then becomes a question of whether the DAC can deliver both. As an example, I just looked up the top Krell amp (100kOhms balanced, but 50ohms! CAST - their current-connection thing), the Mark Levinson 532H (60kOhms balanced) and the Spectrals (10kOhms unbalanced). I am sure we have all discussed/read impedance being important in lots of other places: cartridge loading, amp's output; why would other inter-component be fundamentally different, even if potentially smaller in effect.

 

[Phelonious Ponk]

There are 3 I know of:

1. The ratio of amp input impedance / source output stage impedance must be 10x or higher,
2. The source must be able to output enough voltage to achieve adequate listening levels. In the case of digital attenuation, components need to be carefully selected to avoid too much output voltage and hence the need for large amounts of attenuation - the output voltage must be matched to the sensitivity of the amplification and speakers to achieve desirable listening levels, needing only slight tweaking with the digital volume control (some DAC's made by very clever designers have adjustable output voltages to lock this in perfectly).
3. The current delivery capability of the source output stage must be sufficiently high as ack pointed out.

Let me know if I am missing anything.

Ok. The impedance is easy and there appears to be a pretty wide zone of tolerance there. Still not sure I get the output voltage part. It only needs to be sufficient to drive the amp to high enough listening levels? If it's loud enough will have plenty of control within the range? Really that simple?

Tim

 

[Phelonious Ponk]

usually we want high input amp impedance (of which said Spectrals are not) and low output preamp/DAC. I don't go reading amp's manuals all the time, but I assume most list theirs plus required input voltage (if not also current) for rated output power (aka sensitivity); it then becomes a question of whether the DAC can deliver both.

So if I wanted to look up the manuals online, I should be able to determine if I've got a good match of DAC/pre output/Amp input by the numbers? I don't know what typical voltage ouputs are for DAC/pre units, but there are plenty out there with very low impedance output numbers.

Tim

 

[ack]

So if I wanted to look up the manuals online, I should be able to determine if I've got a good match of DAC/pre output/Amp input by the numbers? I don't know what typical voltage ouputs are for DAC/pre units, but there are plenty out there with very low impedance output numbers.

Tim

Frankly, this is one of those cases where a knowledgeable dealer is of high value for proper system matching. If you take the Krell as an example, you would expect their current/CAST connection - it having a 50ohm input impedance - to demand high output current from the source; yet they rate the CAST-based sensitivity at just 3.08mA RMS for rated output! I don't know how they do it, but it feels awfully low to me, unless the input voltage is very high. Contrast this with the 100mA I mentioned for the Spectrals for rated output. On the other hand, the Krell's balanced-input sensitivity rating is 3.08 V RMS; some DACs don't go above 2V RMS at their output; in fact, I would venture a guess most don't.

 

[AudioExplorations]

Ok. The impedance is easy and there appears to be a pretty wide zone of tolerance there. Still not sure I get the output voltage part. It only needs to be sufficient to drive the amp to high enough listening levels? If it's loud enough will have plenty of control within the range? Really that simple?

Tim

This is also where my technical knowledge ends, but this must be the case as higher output voltage = higher volume. In the case of my DAC the output voltage can be set anywhere between -10dBu (0.245V) and +27dBu (17.35V) using a rotary potentiometer at the back of the unit. I have it set at 0.49V which is the most volume I ever need (for Friday and Saturday nights ) and otherwise fine tune with digital volume control. This is how Daniel Weiss himself has recommended me to set it up in power amp direct configuration.

 

[JackD201]

One can also over drive an amplifier's input section and cause the amp to go into protection. It's a rare thing but was symptomatic for some Nagra amplifiers that were shipped out with 1v sensitivity where owners used predominantly digital sources. Inside the amps (the amps that looked like pyramids) was a switch that could set input at 2v. Problem solved.

I'm just pointing out that the mismatch could be at either end. In fact I suspect that the amplifier end is more of the problem rather than source or pre because this is where we typically see greater variance (input sensitivities). The fix isn't too difficult usually requiring a few resistors soldered on the ICs or attenuators like those by Scott Endler if the problem is an overly excitable amplifier. When the case is the opposite, that's when a preamp or distribution amp comes in handy.

 

[Phelonious Ponk]

OK, so if I'm getting this right, a good pre-out/amp-in impedance match, and enough voltage out of the pre (or DAC, in this case) to drive the amp to listening levels and you're good. Everything else is just the usual stuff, as long as you have impedance compatibility and enough voltage gain, nothing else is lost by skipping the pre?

Tim

 

[microstrip]

With my comment there, I actually partially disagreed with Amir (as I disagreed with him in similar situations with certain amps being able - or not - to drive tough speakers like MLs, for example). In cases like the one we are discussing here, a source component may simply not be able to deliver adequate current to the amp (if we are dealing with an amp requiring unusual input amperage to be driven properly), and the sound is fairly thin and lacking a bit in dynamic contrast. This is the case with Spectral amps, for example, where even the Berkeley Alpha DAC (which was developed driving them directly) isn't as capable as with their preamps in the middle, which put out up to 2A peak (1A RMS) to the amplifier (whose manual claims 100mA min required)- it's only then that the whole system's dynamic capabilities are at full swing. In those cases, the issue can be the DAC's limited output (from said amp's perspective), and certainly the impedance on either end can be a contributing factor. This all goes back to system matching.

Ack,

If there is any current limitation in a source, it will reveal as distortion in the signal and even a very crude measurement at 0 dB will immediately show it.
Spectral amplifiers have a low input impedance - it is a case similar to the DartZeel 108 50 ohm input impedance. Only appropriate sources can drive them. We should remember that these are exceptional cases.

Most of the time people - myself included - do not use the preamplifier for electrical system matching, but for sound optimization.

 

[microstrip]

(...) In fact I suspect that the amplifier end is more of the problem rather than source or pre because this is where we typically see greater variance (input sensitivities). The fix isn't too difficult usually requiring a few resistors soldered on the ICs or attenuators like those by Scott Endler if the problem is an overly excitable amplifier. (...)

Jack,

High gain amplifiers can be a nuisance with some tube preamplifiers and in the past I have tried using attenuators in several amplifiers to solve this problem, and every time it spoiled the sound of the system. We tried several pairs of resistors with different values in order to get different loads and found that this is one of the best ways of ruining the sound of an amplifier. Dynamics and bass slam were always affected.

Otherwise we would not be debating the problem of the low impedance of the Burmester 911 mk3 in another thread - we would just fit a 20 kohm resistor in series with the input. It would just reduce the high gain of this amplifier and increase the input impedance.

 

[rbbert]

Of course, the last time I checked car batteries actually delivered 14.4 volts.

 

[JackD201]

Jack,

High gain amplifiers can be a nuisance with some tube preamplifiers and in the past I have tried using attenuators in several amplifiers to solve this problem, and every time it spoiled the sound of the system. We tried several pairs of resistors with different values in order to get different loads and found that this is one of the best ways of ruining the sound of an amplifier. Dynamics and bass slam were always affected.

Otherwise we would not be debating the problem of the low impedance of the Burmester 911 mk3 in another thread - we would just fit a 20 kohm resistor in series with the input. It would just reduce the high gain of this amplifier and increase the input impedance.

Shhhhhh my friend, some people don't like us talking about SQ

 

[Phelonious Ponk]

Shhhhhh my friend, some people don't like us talking about SQ

I think SQ is what we are talking about. If I put together a larger single source system (that's a given for me), how do I choose a source, with gain and volume control, that is a good match for my amps, or is there something else to it, some other factors? If not; if impedance matching and enough voltage to drive the amp is it (other than the quality of the components - noise, distortion, etc.), then how can an additional boxful of components in the signal chain (called a preamp), regardless of the quality of those components, do anything but degrade SQ?

Tim

 

[garylkoh]

then how can an additional boxful of components in the signal chain (called a preamp), regardless of the quality of those components, do anything but degrade SQ?

Tim

.... plus an additional pair of interconnects - which means two more imperfect connections. I wish I knew.... but in all my testing, paradoxically, a good preamp always added transparency.

When I was setting up the demo with the "Dragons" I initially set up the fabulous Burmester 089 CD player with volume control directly driving the power amps (mine - with 75k ohm input impedance - but with 7m XLR balanced cables) and the sound, while good, was not great. When I added the 077 preamp, transparency went up a notch, and macro- and micro- dynamics was better.

I didn't have to pay for the 077 preamp, so it's definitely not buyer's expectations.

 

[AudioExplorations]

a good preamp always added transparency.

How could this technically be possible, what could cause this? Is it possible the filtration/coloration is giving the perception of transparency? Like wilson speakers with their mid-bass hump are giving the perception of life music? I am sure the 089 can adequately drive such a high impedance load.

Just interested in how this could be explained.

 

[Phelonious Ponk]

How could this technically be possible, what could cause this? Is it possible the filtration/coloration is giving the perception of transparency? Like wilson speakers with their mid-bass hump are giving the perception of life music? I am sure the 089 can adequately drive such a high impedance load.

Just interested in how this could be explained.

Me too. Volume variations? Coloration? Expectations? There has to be a perception explanation.

Tim